Erector spinae plane block for postoperative pain

Abstract

Background

Acute and chronic postoperative pain are important healthcare problems, which can be treated with a combination of opioids and regional anaesthesia. The erector spinae plane block (ESPB) is a new regional anaesthesia technique, which might be able to reduce opioid consumption and related side effects.

Objectives

To compare the analgesic effects and side effect profile of ESPB against no block, placebo block or other regional anaesthetic techniques.

Search methods

We searched the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, Embase and Web of Science on 4 January 2021 and updated the search on 3 January 2022.

Selection criteria

Randomised controlled trials (RCTs) investigating adults undergoing surgery with general anaesthesia were included. We included ESPB in comparison with no block, placebo blocks or other regional anaesthesia techniques irrespective of language, publication year, publication status or technique of regional anaesthesia used (ultrasound, landmarks or peripheral nerve stimulator).

Quasi‐RCTs, cluster‐RCTs, cross‐over trials and studies investigating co‐interventions in either arm were excluded.

Data collection and analysis

Two review authors independently assessed all trials for inclusion and exclusion criteria, and risk of bias (RoB), and extracted data.

We assessed risk of bias using the Cochrane RoB 2 tool, and we used GRADE to rate the certainty of evidence for the primary outcomes.

The primary outcomes were postoperative pain at rest at 24 hours and block‐related adverse events. Secondary outcomes were postoperative pain at rest (2, 48 hours) and during activity (2, 24 and 48 hours after surgery), chronic pain after three and six months, as well as cumulative oral morphine requirements at 2, 24 and 48 hours after surgery and rates of opioid‐related side effects.

Main results

We identified 69 RCTs in the first search and included these in the systematic review. We included 64 RCTs (3973 participants) in the meta‐analysis. The outcome postoperative pain was reported in 38 out of 64 studies; block‐related adverse events were reported in 40 out of 64 studies. We assessed RoB as low in 44 (56%), some concerns in 24 (31%) and high in 10 (13%) of the study results. Overall, 57 studies reported one or both primary outcomes. Only one study reported results on chronic pain after surgery.

In the updated literature search on 3 January 2022 we found 37 new studies and categorised these as awaiting classification.

ESPB compared to no block

There is probably a slight but not clinically relevant reduction in pain intensity at rest 24 hours after surgery in patients treated with ESPB compared to no block (visual analogue scale (VAS), 0 to 10 points) (mean difference (MD) ‐0.77 points, 95% confidence interval (CI) ‐1.08 to ‐0.46; 17 trials, 958 participants; moderate‐certainty evidence). There may be no difference in block‐related adverse events between the groups treated with ESPB and those receiving no block (no events in 18 trials reported, 1045 participants, low‐certainty evidence).

ESPB compared to placebo block

ESPB probably has no effect on postoperative pain intensity at rest 24 hours after surgery compared to placebo block (MD ‐0.14 points, 95% CI ‐0.29 to 0.00; 8 trials, 499 participants; moderate‐certainty evidence). There may be no difference in block‐related adverse events between ESPB and placebo blocks (no events in 10 trials reported; 592 participants; low‐certainty evidence).

ESPB compared to other regional anaesthetic techniques

Paravertebral block (PVB)

ESPB may not have any additional effect on postoperative pain intensity at rest 24 hours after surgery compared to PVB (MD 0.23 points, 95% CI ‐0.06 to 0.52; 7 trials, 478 participants; low‐certainty evidence). There is probably no difference in block‐related adverse events (risk ratio (RR) 0.27, 95% CI 0.08 to 0.95; 7 trials, 522 participants; moderate‐certainty evidence).

Transversus abdominis plane block (TAPB)

ESPB may not have any additional effect on postoperative pain intensity at rest 24 hours after surgery compared to TAPB (MD ‐0.16 points, 95% CI ‐0.46 to 0.14; 3 trials, 160 participants; low‐certainty evidence). There may be no difference in block‐related adverse events (RR 1.00, 95% CI 0.21 to 4.83; 4 trials, 202 participants; low‐certainty evidence).

Serratus anterior plane block (SAPB)

The effect on postoperative pain could not be assessed because no studies reported this outcome. There may be no difference in block‐related adverse events (RR 1.00, 95% CI 0.06 to 15.59; 2 trials, 110 participants; low‐certainty evidence).

Pectoralis plane block (PECSB)

ESPB may not have any additional effect on postoperative pain intensity at rest 24 hours after surgery compared to PECSB (MD 0.24 points, 95% CI ‐0.11 to 0.58; 2 trials, 98 participants; low‐certainty evidence). The effect on block‐related adverse events could not be assessed.

Quadratus lumborum block (QLB)

Only one study reported on each of the primary outcomes.

Intercostal nerve block (ICNB)

ESPB may not have any additional effect on postoperative pain intensity at rest 24 hours after surgery compared to ICNB, but this is uncertain (MD ‐0.33 points, 95% CI ‐3.02 to 2.35; 2 trials, 131 participants; very low‐certainty evidence). There may be no difference in block‐related adverse events, but this is uncertain (RR 0.09, 95% CI 0.04 to 2.28; 3 trials, 181 participants; very low‐certainty evidence).

Epidural analgesia (EA)

We are uncertain whether ESPB has an effect on postoperative pain intensity at rest 24 hours after surgery compared to EA (MD 1.20 points, 95% CI ‐2.52 to 4.93; 2 trials, 81 participants; very low‐certainty evidence). A risk ratio for block‐related adverse events was not estimable because only one study reported this outcome.

Authors' conclusions

ESPB in addition to standard care probably does not improve postoperative pain intensity 24 hours after surgery compared to no block. The number of block‐related adverse events following ESPB was low. Further research is required to study the possibility of extending the duration of analgesia.

We identified 37 new studies in the updated search and there are three ongoing studies, suggesting possible changes to the effect estimates and the certainty of the evidence in the future.

Plain language summary

Pain relief and side effects of a new regional anaesthetic technique (ESPB)

Key messages

Erector spinae plane block (ESPB) probably does not improve postoperative pain intensity 24 hours after surgery, but it may be performed without serious adverse events and may decrease requirements for oral pain relief as well as side effects associated with oral pain relief.

What is regional anaesthesia?

Acute postoperative pain is commonly treated with a combination of medications like opioids and regional anaesthesia. Regional anaesthesia is a way of specifically numbing nerves that carry pain signals in the area where the operation takes place. ESPB is a new technique, first described in 2016, which may reduce the use of opioids in the postoperative period as well as opioid‐associated adverse effects. When applying ESPB, local anaesthetics are injected in the fascial plane underneath the erector spinae muscle and this can be done at different spinal levels. The erector spinae muscles are a group of muscles running along the back on each side of the spine.

What did we want to find out?

We investigated the effectiveness in relieving pain as well as the potential side effects and risks of the ESPB compared to no block, placebo block (patients received saline injection instead of active drugs) and other regional anaesthesia in adults undergoing surgery with general anaesthesia.

What did we do?

We included 64 studies that randomly assigned patients to one of the treatment groups (randomised controlled trials) with 3973 patients in total. In these studies all patients received standard opioid care as well as the different regional anaesthesia according to their group allocation.

What did we find?

Postoperative pain at rest 24 hours after surgery did not show a clinically relevant difference, with no additional benefit on postoperative pain intensity.

Block‐related adverse events were very rare, and we did not observe a significant difference between groups.

Oral opioid consumption as well as opioid‐related side effects (such as nausea and itching) were decreased in the groups of patients receiving ESPB treatment.

Limitations of the evidence

We rated the certainty of evidence for the primary outcomes using the GRADE score (Grades of Recommendation, Assessment, Development and Evaluation). The certainty of evidence was graded mostly low, ranging from moderate to very low. This was in part due to many small studies with very diverse outcomes.

Conclusion

The current evidence suggests that ESPB probably does not improve postoperative pain intensity one day after surgery beyond standard pain‐reducing medications. However, ESPB may not have a high risk of serious block‐related adverse events. Furthermore, ESPB may help with reducing the need for opioid medication and thereby reducing opioid‐related side effects. Further research is needed to better define the role of ESPB in postoperative pain treatment.

How up‐to‐date is this evidence?

The evidence is up‐to‐date to 4 January 2021.

Summary of findings

Summary of findings 1. Summary of findings ‐ Erector spinae plane block vs no block in adults undergoing surgery with general anaesthesia.

Patient or population: adults undergoing surgery with general anaesthesia Setting: operating theatre, PACU, inpatient ward Intervention: ESPB Comparison: no block
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	N° of participants (studies)	Certainty in the evidence (GRADE)	Comment
	Assumed risk with standard of care	Corresponding risk with ESPB
Mean difference in postoperative pain intensity at rest 24 hours postoperatively (VAS 0 to 10)	The mean VAS scores ranged across the control group from 0.6 to 4.8	The mean postoperative pain intensity at rest (24 hours postoperatively) was on average 0.77 points lower (95% CI ‐1.08 to ‐0.46) in the intervention group	—	958 (17)	Moderatea	There is probably no clinically relevant reduction in pain intensity at rest 24 hours after surgery in patients treated with ESPB compared to opioid treatment.
Rates of block‐related adverse events	No patients out of 522 in the control group suffered from block‐related adverse events	No patients out of 523 in the ESPB group suffered from block‐related adverse events	Not estimable in RevMan In R: RR 1.00 (0.40 to 2.49)	1045 (18)	Lowb	There is probably no difference in block‐related adverse events between the groups treated with an ESPB and those receiving opioid treatment.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk on the comparison group and the relative effect of the intervention (and its 95% confidence interval). CI: confidence interval; ESPB: erector spinae plane block; MD: mean difference; NA: not applicable; PACU: post‐anaesthesia care unit; RCT: randomised controlled trial; RR: risk ratio; VAS: visual analogue scale
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is the possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

^aNo change of result in sensitivity analysis excluding studies with a high risk of bias or some concern (MD ‐0.62, 95% CI‐0.95 to ‐0.30; 14 studies, 798 participants). Therefore, no downgrading of evidence for risk of bias. Downgraded by one level due to imprecision.
^bDowngraded by two levels due to imprecision.

*Clinically relevant assumptions for pain: MD ‐1 point on VAS 0 to 10, SD = based on average SD of all included low risk of bias trials.

*Clinically relevant assumptions for block‐related adverse effects: RR > 1.1 or < 0.9.

Summary of findings 2. Summary of findings ‐ Erector spinae plane block vs placebo block in adults undergoing surgery with general anaesthesia.

Patient or population: adults undergoing surgery with general anaesthesia Setting: operating theatre, PACU, inpatient ward Intervention: ESPB Comparison: placebo block
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	N° of participants (studies)	Certainty in the evidence (GRADE)	Comment
	Assumed risk with standard of care	Corresponding risk with ESPB
Mean difference in postoperative pain intensity at rest 24 hours postoperatively (VAS 0 to 10)	The mean VAS scores ranged across the control group from 1.3 to 4.4	The mean postoperative pain at rest (24 hours postoperatively) in the intervention group was not different (MD ‐0.14 points, 95% CI ‐0.29 to 0.00)	—	499 (8)	Moderatea	There is probably no clinically relevant reduction in pain intensity at rest 24 hours after surgery in patients treated with ESPB compared to placebo block.
Rates of block‐related adverse events	No patients out of 296 in the control group suffered from block‐related adverse events	No patients out of 296 in the ESPB group suffered from block‐related adverse events	Not estimable in RevMan In R: RR 1.00 (0.29 to 3.41)	592 (10)	Lowb	There is probably no difference in block‐related adverse events.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk on the comparison group and the relative effect of the intervention (and its 95% confidence interval). CI: confidence interval; ESPB: erector spinae plane block; MD: mean difference; NA: not applicable; PACU: post‐anaesthesia care unit; RCT: randomised controlled trial; RR: risk ratio; VAS: visual analogue scale
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is the possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

^aNo change of result in sensitivity analysis excluding studies with a high risk of bias or some concerns (MD ‐0.16, 95% CI ‐0.35 to 0.03; 7 studies, 439 participants). Therefore, no downgrading for risk of bias. Downgraded due to imprecision.
^bDowngraded by two levels due to imprecision.

*Clinically relevant assumptions for pain: MD ‐1 point on VAS 0 to 10, SD = based on average SD of all included low risk of bias trials.

*Clinically relevant assumptions for pain: RR > 1.1 or < 0.9.

Summary of findings 3. Summary of findings ‐ Erector spinae plane block vs paravertebral block in adults undergoing surgery with general anaesthesia.

Patient or population: adults undergoing surgery with general anaesthesia Setting: operating theatre, PACU, inpatient ward Intervention: ESPB Comparison: PVB
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	N° of participants (studies)	Certainty in the evidence (GRADE)	Comment
	Assumed risk with standard of care	Corresponding risk with ESPB
Mean difference in postoperative pain intensity at rest 24 hours postoperatively (VAS 0 to 10)	The mean VAS scores ranged across the control group from 0.1 to 3.5	The mean postoperative pain at rest (24 hours postoperatively) in the intervention group was not different (MD 0.23 points, 95% CI ‐0.06 to 0.52)	—	478 (7)	Lowa	An ESPB may not have an effect on postoperative pain intensity at rest 24 hours after surgery compared to a PVB.
Rates of block‐related adverse events	13 patients out of 261 in the control group suffered from block‐related adverse events	No patients out of 261 in the ESPB group suffered from block‐related adverse events	RR 0.10 (95% CI 0.02 to 0.55) In R: RR 0.27 (0.08 to 0.95)	522 (7)	Moderateb	There may be no difference in block‐related adverse events between the groups treated with an ESPB and those receiving a PVB.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk on the comparison group and the relative effect of the intervention (and its 95% confidence interval). CI: confidence interval; ESPB: erector spinae plane block; MD: mean difference; NA: not applicable; PACU: post‐anaesthesia care unit; PVB: paravertebral block; RCT: randomised controlled trial; RR: risk ratio; VAS: visual analogue scale
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is the possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

^aNo change of result in sensitivity analysis (MD 0.22, 95% CI ‐0.12 to ‐0.55); 387 participants, 6 studies, excluding studies with a high risk of bias or some concerns. Downgraded by two levels due to inconsistency and imprecision.
^bDowngraded by one level due to risk of bias.

* Clinically relevant assumptions for pain: MD ‐1 point on VAS 0 to 10, SD = based on average SD of all included low risk of bias trials.
* Clinically relevant assumptions for pain: RR > 1.1 or < 0.9.

Summary of findings 4. Summary of findings ‐ Erector spinae plane block vs transversus abdominis plane block in adults undergoing surgery with general anaesthesia.

Patient or population: adults undergoing surgery with general anaesthesia Setting: operating theatre, PACU, inpatient ward Intervention: ESPB Comparison: TAPB
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	N° of participants (studies)	Certainty in the evidence (GRADE)	Comment
	Assumed risk with standard of care	Corresponding risk with ESPB
Mean difference in postoperative pain intensity at rest 24 hours postoperatively (VAS 0 to 10)	The mean VAS scores ranged across the control group from 1 to 2.7	The mean postoperative pain at rest (24 hours postoperatively) in the intervention group was not different (MD ‐0.16, 95% CI ‐0.46 to 0.14)	—	160 (3)	Lowa	There is probably no clinically relevant reduction in pain intensity at rest 24 hours after surgery in patients treated with ESPB compared to TAPB.
Rates of block‐related adverse events	2 patients out of 101 in the control group suffered from block‐related adverse events	2 patients out of 101 in the ESPB group suffered from block‐related adverse events	In RevMan:RR 1.00 (95% CI 0.15 to 6.84) In R: RR 1.00 (0.21 to 4.83)	202 (4 RCTs)	Lowb	There may be no difference in block‐related adverse events between patients treated with an ESPB and those receiving a TAPB.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk on the comparison group and the relative effect of the intervention (and its 95% confidence interval). CI: confidence interval; ESPB: erector spinae plane block; MD: mean difference; NA: not applicable; PACU: post‐anaesthesia care unit; RCT: randomised controlled trial; RR: risk ratio; TAPB: transversus abdominis plane block; VAS: visual analogue scale
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is the possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

^aNo change of result in sensitivity analysis excluding studies with a high risk of bias or some concerns (MD ‐0.16, 95% CI ‐0.46 to 0.14; 3 studies, 160 participants). Downgraded by two levels due to imprecision and inconsistency.
^bDowngraded by two levels due to imprecision.

* Clinically relevant assumptions for pain: MD ‐1 point on VAS 0 to 10, SD = based on average SD of all included low risk of bias trials.

* Clinically relevant assumptions for pain: RR > 1.1 or < 0.9.

Summary of findings 5. Summary of findings ‐ Erector spinae plane block vs serratus anterior plane block in adults undergoing surgery with general anaesthesia.

Patient or population: adults undergoing surgery with general anaesthesia Setting: operating theatre, PACU, inpatient ward Intervention: ESPB Comparison: SAPB
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	N° of participants (studies)	Certainty in the evidence (GRADE)	Comment
	Assumed risk with standard of care	Corresponding risk with ESPB
Mean difference in postoperative pain intensity at rest 24 hours postoperatively (VAS 0 to 10)	—	—	—	0 (0)	—	—
Rates of block‐related adverse events	No patients out of 55 in the control group suffered from block‐related adverse events	No patients out of 55 in the ESPB group suffered from block‐related adverse events	Not estimable in RevMan In R: RR 1.00 (0.06 to 15.59)	110 (2)	Lowa	There is probably no difference in block‐related adverse events between the groups treated with an ESPB and those receiving SAPB treatment.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk on the comparison group and the relative effect of the intervention (and its 95% confidence interval). CI: confidence interval; ESPB: erector spinae plane block; MD: mean difference; NA: not applicable; PACU: post‐anaesthesia care unit; RCT: randomised controlled trial; RR: risk ratio; SAPB: serratus anterior plane block; VAS: visual analogue scale
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is the possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

^aDowngraded by two levels due to imprecision.

*Clinically relevant assumptions for pain: RR > 1.1 or < 0.9.

Summary of findings 6. Summary of findings ‐ Erector spinae plane block vs pectoralis plane block in adults undergoing surgery with general anaesthesia.

Patient or population: adults undergoing surgery with general anaesthesia Setting: operating theatre, PACU, inpatient ward Intervention: ESPB Comparison: PECSB
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	N° of participants (studies)	Certainty in the evidence (GRADE)	Comment
	Assumed risk with standard of care	Corresponding risk with ESPB
Mean difference in postoperative pain intensity at rest 24 hours postoperatively (VAS 0 to 10)	The mean VAS scores ranged across the control group from 1 to 2.33	The mean postoperative pain at rest (24 hours postoperatively) in the intervention group was not different (MD 0.24 points, 95% CI ‐0.11 to 0.58)	—	98 (2)	Lowa	An ESPB may not have an effect on postoperative pain intensity at rest 24 hours after surgery compared to a PECSB.
Rates of block‐related adverse events	No patients out of 23 in the control group suffered from block‐related adverse events	1 patient out of 24 in the ESPB group suffered from block‐related adverse events	NA	47 (1)	NA	Unable to calculate further statistics due to an insufficient number of trials
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk on the comparison group and the relative effect of the intervention (and its 95% confidence interval). CI: confidence interval; ESPB: erector spinae plane block; MD: mean difference; NA: not applicable; PACU: post‐anaesthesia care unit; PECSB: pectoralis plane block; RCT: randomised controlled trial; RR: risk ratio; VAS: visual analogue scale
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is the possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

^aNo change of result in sensitivity analysis because both studies included were of low risk of bias. Therefore, no downgrading of evidence for risk of bias. Downgraded by two levels due to imprecision and inconsistency.

*Clinically relevant assumptions for pain: MD ‐1 point on VAS 0 to 10, SD = based on average SD of all included low risk of bias trials.

*Clinically relevant assumptions for pain: RR > 1.1 or < 0.9.

Summary of findings 7. Summary of findings ‐ Erector spinae plane block vs quadratus lumborum block in adults undergoing surgery with general anaesthesia.

Patient or population: adults undergoing surgery with general anaesthesia Setting: operating theatre, PACU, inpatient ward Intervention: ESPB Comparison: QLB
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	N° of participants (studies)	Certainty in the evidence (GRADE)	Comment
	Assumed risk with standard of care	Corresponding risk with ESPB
Mean difference in postoperative pain intensity at rest 24 hours postoperatively (VAS 0 to 10)	The mean VAS scores was 2.1	The mean postoperative pain at rest (24 hours postoperatively) in the intervention group was not different (MD ‐0.05 points, 95% CI ‐0.21 to 0.31)	—	40 (1)	NA	Unable to calculate further statistics due to an insufficient number of trials
Rates of block‐related adverse events	No patients out of 20 in the control group suffered from block‐related adverse events	No patients out of 20 in the ESPB group suffered from block‐related adverse events	—	40 (1)	NA	Unable to calculate further statistics due to an insufficient number of trials
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk on the comparison group and the relative effect of the intervention (and its 95% confidence interval). CI: confidence interval; ESPB: erector spinae plane block; MD: mean difference; NA: not applicable; PACU: post‐anaesthesia care unit; QLB: quadratus lumborum block; RCT: randomised controlled trial; RR: risk ratio; VAS: visual analogue scale
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is the possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

*Clinically relevant assumptions for pain: MD ‐1 point on VAS 0 to 10, SD = based on average SD of all included low risk of bias trials.

*Clinically relevant assumptions for pain: RR > 1.1 or < 0.9.

Summary of findings 8. Summary of findings ‐ Erector spinae plane block vs intercostal nerve block in adults undergoing surgery with general anaesthesia.

Patient or population: adults undergoing surgery with general anaesthesia Setting: operating theatre, PACU, inpatient ward Intervention: ESPB Comparison: ICNB
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	N° of participants (studies)	Certainty in the evidence (GRADE)	Comment
	Assumed risk with standard of care	Corresponding risk with ESPB
Mean difference in postoperative pain intensity at rest 24 hours postoperatively (VAS 0 to 10)	The mean VAS scores ranged across the control group from 1.0 to 2.26	The mean postoperative pain at rest (24 hours postoperatively) in the intervention group was not different (MD ‐0.33 points, 95% CI ‐3.02 to 2.35)	—	131 (2)	Very lowa	An ESPB may not have an effect on postoperative pain intensity at rest 24 hours after surgery compared to a ICNB.
Rates of block‐related adverse events	5 patients out of 91 in the control group suffered from block‐related adverse events	No patients out of 90 in the ESPB group suffered from block‐related adverse events	In RevMan: RR 0.09 0.01 to 1.56 In R: RR 0.09 (0.04 to 2.28)	181 (3)	Very lowb	There may be no difference in block‐related adverse events between patients treated with ESPB and ICNB.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk on the comparison group and the relative effect of the intervention (and its 95% confidence interval). CI: confidence interval; ESPB: erector spinae plane block; ICNB: intercostal nerve block; MD: mean difference; NA: not applicable; PACU: post‐anaesthesia care unit; RCT: randomised controlled trial; RR: risk ratio; VAS: visual analogue scale
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is the possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

^aDowngraded by three levels due to inconsistency, imprecision and risk of bias.
^bDowngraded by three levels due to risk of bias and imprecision.

*Clinically relevant assumptions for pain: MD ‐1 point on VAS 0 to 10, SD = based on average SD of all included low risk of bias trials.

*Clinically relevant assumptions for pain: RR > 1.1 or < 0.9.

Summary of findings 9. Summary of findings ‐ Erector spinae plane block vs epidural analgesia in adults undergoing surgery with general anaesthesia.

Patient or population: adults undergoing surgery with general anaesthesia Setting: operating theatre, PACU, inpatient ward Intervention: ESPB Comparison: EA
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	N° of participants (studies)	Certainty in the evidence (GRADE)	Comment
	Assumed risk with standard of care	Corresponding risk with ESPB
Mean difference in postoperative pain intensity at rest 24 hours postoperatively (VAS 0 to 10)	The mean VAS scores ranged across the control group from 0.53 to 2.08	The mean postoperative pain at rest (24 hours postoperatively) in the intervention group was not different (MD 1.2, 95% CI ‐2.52 to 4.93)	—	81 (2)	Very lowa	We are uncertain whether an ESPB (compared to EA) has an effect on postoperative pain intensity at rest 24 hours after surgery
Rates of block‐related adverse events	No patients out of 25 in the control group suffered from block‐related adverse events	No patients out of 25 in the ESPB group suffered from block‐related adverse events	NA	50 (1)	NA	Unable to calculate further statistics due to an insufficient number of trials
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk on the comparison group and the relative effect of the intervention (and its 95% confidence interval). CI: confidence interval; EA: epidural analgesia; ESPB: erector spinae plane block; MD: mean difference; NA: not applicable; PACU: post‐anaesthesia care unit; RCT: randomised controlled trial; RR: risk ratio; VAS: visual analogue scale
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is the possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

^aDowngraded by three levels due to inconsistency, risk of bias and imprecision.

*Clinically relevant assumptions for pain: MD ‐1 point on VAS 0 to 10 scale, SD = based on average SD of all included low risk of bias trials.

*Clinically relevant assumptions for pain: RR > 1.1 or < 0.9.

Background

The erector spinae plane block (ESPB) is a new regional anaesthetic technique, first described by Forero and colleagues in 2016 (Forero 2016). ESPB is an ultrasound‐guided technique of applying local anaesthetic in the fascial plane deep to the erector spinae muscle and it can be used at different spinal levels. This makes ESPB a potentially relevant regional anaesthetic technique for many different surgical procedures from thoracic surgery to cholecystectomy. The ESPB also has the benefits of being relatively easy to perform as well as having fewer serious side effects when compared to other regional anaesthetic techniques. Given the current interest in this new block we carried out a Cochrane systematic review of studies using ESPB for many different surgical interventions and investigated its effects using meta‐analysis.

Description of the condition

Postoperative pain and chronic pain are relevant healthcare problems following surgery. According to a recently published survey, 54% of patients still reported moderate‐to‐extreme pain at discharge following broad‐spectrum surgical procedures (Buvanendran 2015). Additionally, the European project euCPSP showed that the overall incidence of relevant (moderate‐to‐severe) chronic pain was 11.8% one year after surgery (Fletcher 2015). Enhanced Recovery After Surgery (ERAS) programmes have been described for many routine surgical procedures (e.g. thoracotomy). One of the main goals of ERAS programmes is sufficient postoperative pain control with as small a dose of opioids as possible (Cavallaro 2019; Kennedy 2020; Noss 2018; Persing 2020; Semenkovich 2018; Soffin 2016). Opioids for perioperative pain therapy have been criticised for several reasons; as a consequence, opioid‐free anaesthesia and analgesia, and more specifically regional anaesthesia techniques as an important cornerstone, are gaining popularity (Fiore 2019).

Description of the intervention

Epidural catheters are currently the gold standard for postoperative pain therapy following major thoracic (e.g. thoracotomy) (Yeung 2016) and open abdominal surgical procedures (e.g. laparotomies) (Guay 2014; Guay 2016; Guay 2016a; Roeb 2017). However, they are criticised because of a possibly higher than currently noted risk of severe adverse events (e.g. epidural haematoma) (Popping 2008; Su 2019; Volk 2012), increased intraoperative need for volume replacement (as a consequence of a higher sympathetic blockade) (Chen 2020; Holte 2004), and a possibly large number of failed blocks (Hermanides 2012; Pan 2004). Furthermore, the increased use of new oral anticoagulants in an ageing population does not allow epidural catheter placement due to an increased bleeding risk (Volk 2015).

Many different peripheral fascial plane blocks (e.g. transversus abdominal plane block, pectoralis plane block) have been described as possibly less invasive and much easier to learn than alternative regional anaesthetic techniques. One of these new and interesting fascial blocks is the erector spinae plane block, which was initially described by Forero and colleagues as a treatment for thoracic neuropathic pain (Forero 2016). The possible advantages of this block are its simplicity and fast learning curve, the wide spectrum of possible implementation over the whole spine (thoracic, abdominal surgery), and the potential for low numbers of puncture‐related adverse events (El‐Boghdadly 2017). Furthermore, it has been advocated that erector spinae plane blocks might be safely applied even in anticoagulated patients, patients undergoing cardiopulmonary bypass surgery and patients with impaired coagulation (De Cassai 2018; Maddineni 2020). The erector spinae plane block is described as an ultrasound‐guided local anaesthetic application into the plane between the erector spinae muscle and the transverse process at the spine by using an in‐plane or out‐of‐plane approach. This block can be performed either as single shot or continuous catheter‐based regional anaesthetic technique.

How the intervention might work

To date, it is unclear exactly how an erector spinae plane block works as there are several conflicting mechanisms described in the literature.

Several anatomic trials primarily revealed a fascial distribution of the injected volume in a craniocaudal direction (3 to 4 levels) and a spread to the lateral posterior back muscles. More recently it has been shown that there is a volume‐dependent spread to the paravertebral space via the inter‐foraminal foramen (~3 spinal levels), but this spread did not significantly increase following larger injection volumes beyond 20 mL (Choi 2019). In contrast, 10 mL volumes did not lead to penetration of dye to the paravertebral space (Choi 2019). However, there are also conflicting data in other cadaver trials reporting no spread to the paravertebral space following an erector spinae plane block at the same thoracic level with the same volume (Ivanusic 2018; Yang 2018). Other authors, by contrast, reported that the spread to the paravertebral space might be different due to a different length of the transverse process at the lumbar and thoracic spine level (Tulgar 2019). They described the effect of an erector spinae plane block following a lumbar injection like a lumbar plexus block. Furthermore, a limited spread to the epidural space has also been described following an application of 20 mL local anaesthetic volume (Schwartzmann 2020). Additionally, even gaps between the superior costotransverse ligament layers are described as being relevant for the efficacy of an erector spinae plane block.

These different anatomical routes via which the local anaesthetic might spread after ESPB injection have, as a consequence, potentially different types of analgesia. For example, somatic analgesia after ESPB is provided by local anaesthetic reaching the dorsal rami but ESPB might also cause visceral analgesia via epidural or paravertebral spread of local anaesthetic.

Taken together, there is evidence that the erector spinae plane block can result in a sensory block (possibly visceral and somatic analgesia) involving both the dorsal and ventral rami, although the extent of blockade is highly variable. There is evidence that there is wide variation in the extent of sensory block with the erector spinae plane block (Schwartzmann 2020).

Why it is important to do this review

Although it is currently not definitively clear from the literature how exactly the erector spinae block works, several randomised controlled trials (RCTs) and case reports have already been published. However, the clinical efficacy of this block for postoperative pain control and treatment following a wide range of surgical procedures (e.g. breast, thoracic, abdominal, spine, hip surgery) is currently unclear (El‐Boghdadly 2017).

Additionally, the exact role of this block, especially in comparison to other well‐established blocks (e.g. epidural, paravertebral block), must be defined (El‐Boghdadly 2017). In addition, this less invasive block might provide regional analgesia for many patients — including those with contraindications to other more invasive blocks (like epidural catheters) — undergoing various surgical procedures (even with effective anticoagulation) (Smith 2020).

Many patients might benefit from an erector spinae plane block through its addition to a multimodal postoperative pain management regimen (e.g. reduced pain intensity scores, opioid sparing, enhanced recovery).

Objectives

To compare the analgesic efficacy and safety profile of erector spinae plane blocks versus no block, placebo block or other regional anaesthetic techniques in adults undergoing surgery with general anaesthesia.

Methods

Criteria for considering studies for this review

Types of studies

We only included randomised controlled trials (RCTs). We excluded cluster‐randomised, cross‐over and quasi‐randomised trials. We applied no restrictions on language, publication year or publication status.

Types of participants

We included adult participants (≥ 18 years) undergoing any surgery with general anaesthesia.

Types of interventions

We included all RCTs comparing single or continuous erector spinae plane block versus no block, placebo block (patients received saline instead of local anaesthetics) or any other regional anaesthetic technique. We excluded studies investigating co‐interventions (e.g. erector spinae plane block in combination with any other regional anaesthetic technique) either in the treatment or control group.

Inclusion was independent of the technique used (i.e. landmarks, peripheral nerve stimulator or ultrasound), the level of training of providers and the amount or concentration of local anaesthetics used.

Types of outcome measures

The following outcomes were defined as primary and secondary outcomes.

Primary outcomes

1. Mean difference in postoperative pain intensity at rest (24 hours postoperatively)

2. Rates of block‐related adverse events (any kind of neurological complication, accidental vascular puncture, motor blockade, local anaesthetic toxicity)

Secondary outcomes

1. Mean difference in postoperative pain intensity during activity (24 hours postoperatively)

2. Mean difference in postoperative pain intensity at rest/during activity (two hours (within the postoperative care unit), 48 hours)

3. Rate of chronic post‐surgical pain (after three and six months)

4. Cumulative mean oral morphine requirement in mg (two hours (within the postoperative care unit), 24 hours, 48 hours postoperatively)

5. Rates of opioid‐related adverse events (e.g. postoperative vomiting (POV), postoperative nausea and vomiting (PONV), respiratory depression)

6. Any other side effect. We planned to collect all side effects reported by the included studies. We planned to summarise the results narratively (no meta‐analysis was to be performed). Side effects are presented in an additional table format. We aimed to provide an overview of all side effects that the included studies had investigated and reported.

We applied no restrictions regarding the scales that were used to measure pain intensity.

Search methods for identification of studies

Generally, we searched all databases using a combination of indexed and free‐text terms.

Electronic searches

We combined the MEDLINE search with the Cochrane Highly Sensitive Search Strategy for identifying randomised controlled trials in MEDLINE (Lefebre 2019). Where appropriate, we used similar search strategies for identifying RCTs in the other databases. We applied no restrictions on language, publication year or publication status. The search strategy for MEDLINE, including a search for systematic reviews, can be found in Appendix 1.

We searched the following databases on 4 January 2021 and updated the search on 3 January 2022:

1. Cochrane Central Register of Controlled Trials (CENTRAL), in the Cochrane Library;

2. MEDLINE (Ovid SP, 1946 to 3 January 2022);

3. Embase (Ovid SP, 1974 to 3 January 2022);

4. Web of Science (1945 to 3 January 2022).

Searching other resources

We checked the bibliographic references and citations of relevant studies and reviews for further references to trials. We searched ClinicalTrials.gov (www.clinicaltrials.gov) and the World Health Organization International Clinical Trials Registry Platform (WHO ICTRP) (apps.who.int/trialsearch/) for unpublished and ongoing studies. When necessary, we contacted trial authors for additional information. We developed the search strategy in consultation with the Cochrane Anaesthesia Information Specialist.

Data collection and analysis

Two review authors (LO, AS) independently scanned titles, abstracts and full texts to exclude irrelevant studies.

Selection of studies

Two review authors (LO, AS) independently identified studies that might be included in this review. For disagreements, we consulted a third review author (EPZ) and resolved all differences by discussion. The selection process was recorded in sufficient detail to complete a PRISMA flow diagram. We listed all studies that were excluded during full‐text screening along with the reasons for exclusion.

Data extraction and management

Two review authors (LO, AS) independently extracted data using a standardised data extraction form (e.g. including type of anaesthesia, local anaesthetic dose, use of multimodal analgesia, use of PONV prophylaxis, type of surgery) developed by the review authors. We tried to obtain missing data by contacting the lead authors of the relevant articles. At each step of data extraction, we resolved differences by discussion within the group of review authors.

Assessment of risk of bias in included studies

Two review authors (LO, AS) independently assessed risk of bias using the Cochrane risk of bias tool 2 for randomised trials (RoB 2) (Higgins 2019; Sterne 2019). The review authors resolved disagreements by discussion with a third review author (EPZ). The effect of interest was the effect of assignment at baseline, regardless of whether the interventions were received as intended (the ‘intention‐to‐treat effect’). We assessed the risk of bias for all results reported in the included studies that are specified as one of the primary outcomes of the current review and contribute to the review’s summary of findings tables:

• Mean difference in postoperative pain intensity at rest (24 hours postoperatively)

• Rates of block‐related adverse events (any kind of neurological complication, accidental vascular puncture, motor blockade, local anaesthetic toxicity)

We assessed five domains by answering signalling questions and judged them as ‘low risk of bias’, ‘some concerns’ or ‘high risk of bias’:

• Bias arising from the randomisation process

• Bias due to deviations from intended interventions

• Bias due to missing outcome data

• Bias in measurement of the outcome

• Bias in selection of the reported result

We used the RoB 2 Excel tool to implement RoB 2. We summarised the overall risk of bias for each study result as described in Higgins 2019. We judged the result to be at ‘overall low risk of bias’ if all domains were assessed as ‘low risk of bias’. We judged the result to be ‘overall some concerns’ if at least one domain for this result was assessed as ‘some concerns’, but not to be at high risk of bias for any domain. We judged the result to be at ‘overall high risk of bias’ if at least one domain for this result was assessed as ‘high risk of bias’.

We presented this information in the review and reflected this in the analyses and conclusions (see Sensitivity analysis; Results).

Measures of treatment effect

For proportions (dichotomous outcomes), we calculated the risk ratio (RR) with 95% confidence interval (CI). For continuous outcomes, we estimated the mean difference (MD) with 95% CI.

For the outcome 'postoperative pain', we calculated the MD because all studies used the same visual analogue scale or numeric rating scale (from 0 to 10 points).

For the outcome 'cumulative postoperative morphine consumption', we converted all reported opioids into oral morphine equivalents using the opioid calculator app of the Faculty of Pain Medicine of the Australian and New Zealand College of Anaesthetists (https://apps.apple.com/de/app/opioid-calculator/id1039219870).

We assessed the number needed to treat for an additional beneficial outcome (NNTB) for efficacy outcomes, and the number needed to treat for an additional harmful outcome (NNTH) for adverse events if enough trials could be pooled (more than four trials per outcome).

Size of effects

We used clinically relevant size of effects rather than statistical significance to describe meta‐analysis estimates (Nikolakopoulou 2019; Wasserstein 2019). For the efficacy outcome postoperative pain intensity, we defined the clinically important size of benefit as a MD of ‐1 (pain intensity reduction) and the opposite as clinically important harm (MD = 1). We considered effects lower than ‐1 and larger than 1 as clinically important. The range between ‐1 and 1 is termed the 'range of equivalence', which includes unimportant treatment effects (Nikolakopoulou 2019). The relative position of the point estimates indicates the direction of the effect as 'benefit', 'no or minimal effect' or 'harm' (Figure 1). The extent of the 95% CI indicates the certainty of the estimated effect as 'some concerns' or as 'major concerns' (Figure 1).

1.

Magnitude‐based inferences to describe precision and clinical relevance of the effect estimates for postoperative pain within 24 hours

For safety outcomes such as opioid‐related and block‐related adverse events, we defined the clinically important size of benefit as a RR of 0.90 and the reciprocal as clinically important harm (RR = 1.11) with a 'range of equivalence' of 0.9 to 1.11 (Figure 2). We considered effects lower than 0.90 and larger than 1.11 as clinically important.

2.

Magnitude‐based inferences to describe precision and clinical relevance of the effect estimates for adverse events within 24 hours

Unit of analysis issues

We included only RCTs. In a simple parallel‐group design for a clinical trial, participants are individually randomised to one of two intervention groups, and a single measurement for each outcome from each participant is collected and analysed (Higgins 2019b). There is no unit of analysis issue with two‐armed RCTs.

We also included multi‐armed studies. We planned to overcome a unit of analysis error for studies that contributed multiple comparisons by either (Higgins 2019b):

• combining groups (by using the appropriate formula for adding SDs when required) to create a single pair‐wise comparison;

• selecting one pair of interventions and excluding the others;

• splitting the ‘shared’ group into two or more groups with smaller sample sizes, and including two or more (reasonably independent) comparisons, if the presented data in the trials allowed us to do so.

Dealing with missing data

For included studies, we noted the level of attrition. We contacted review authors whenever possible to ask about missing data or other potentially unreported data in the included studies, however we only received two replies. In the case of missing outcome data (missing participants), we included data in the analysis only for those participants whose results were known. We performed a complete‐case analysis. We subsequently excluded studies with incomplete reporting of their study flow or disputable exclusions in a sensitivity analysis to assess their impact on the results. We considered the potential impact of the missing data on the results in the interpretation of the results of the review.

In the case of poor reporting of outcome data and missing statistics from the data summaries presented, we calculated missing statistics, as described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2019b). If data were reported as median values with interquartile ranges, we assumed that the median is very similar to the mean when the distribution of data is symmetrical, and we used the median directly in the meta‐analysis and calculated the standard deviation from the interquartile range in accordance with Higgins 2019b. We did not pool asymmetrical data.

Assessment of heterogeneity

We assessed clinical and methodological differences within included studies to decide whether studies were sufficiently homogeneous to be combined in meta‐analysis.

We measured statistical heterogeneity using the Chi² test, the I² statistic and the 95% prediction interval (PI) for random‐effects meta‐analysis. The 95% PI helps us to understand the impact of heterogeneity on the range of true treatment effects in future studies. In case of heterogeneity, the 95% PI covers a wider range with a different clinical conclusion than the 95% CI (IntHout 2016). We restricted the calculation of a 95% PI to meta‐analyses with four or more studies (≥ 200 participants), since the interval would be imprecise when a summary estimate was based on only a few small studies (IntHout 2016). We used the R package meta to calculate 95% PIs.

We declared statistical heterogeneity if P < 0.1 for the Chi² statistic, or I² ≥ 40% (40% to 60%: may represent moderate heterogeneity; 50% to 90%: may represent substantial heterogeneity; 75% to 100%: considerable heterogeneity), or the range of the 95% PI revealed a different clinical interpretation of the effect estimate compared to the 95% CI with respect to the range of equivalence.

Assessment of reporting biases

We created funnel plots for outcomes including more than 10 trials by plotting effect estimates of included trials versus their precision (inverse of the standard error of the point estimate). We used these plots only as a guiding technique to detect possible reporting bias and small‐study effects.

If asymmetry is suggested by visual assessment, we investigated it by performing exploratory analyses (e.g. Arcsine test for binary data (Rucker 2008), Egger's test for continuous data (Egger 1997)). To adjust for small‐study effects, we used Duval and Tweedie's trim and fill method (Duval 2000). We performed all statistical tests for publication bias by using the R package meta (Schwarzer 2007).

Data synthesis

For dichotomous data, we used the Mantel‐Haenszel method. For continuous data, we used the inverse variance method in RevMan Web. We used the fixed‐effect model for meta‐analysis when it was reasonable to assume that studies were estimating the same underlying treatment effect (i.e. when trials were examining the same intervention and trial populations, and when the methods were judged sufficiently similar). When clinical or methodological heterogeneity was sufficient to suggest that underlying treatment effects might differ between trials, we used random‐effects meta‐analysis to produce an average estimate of the treatment effect. However, after taking into account that study weights are more balanced under the random‐effects than the fixed‐effect model (assigning large studies less relative weight and small studies more relative weight), we reported summary statistics in conjunction with results of a sensitivity analysis (obtained via both models).

Computational problems can occur when no events are observed in one or both groups in an individual study (Higgins 2019b). RevMan ignores zero/zero event trials and uses a constant continuity correction of 0.5 for studies with zero events in one arm. We used RevMan as our primary analysis method. Additionally, we conducted meta‐analyses using a constant continuity correction of 0.5 for all trials with zero cell frequencies using the R package meta (Schwarzer 2007; Sweeting 2004). We presented the latter results as sensitivity analyses.

If a meta‐analysis was not possible, we reported primary outcomes for each included study within a table combined with the RoB 2 results. We restricted 95% prediction intervals and zero event analyses, which we only performed for outcomes with more than four trials.

Subgroup analysis and investigation of heterogeneity

We investigated statistical heterogeneity as described (Assessment of heterogeneity). We performed subgroup analyses to calculate RR or MD in conjunction with the corresponding CI for each subgroup if statistical heterogeneity was present (P < 0.01 for the Chi² test of heterogeneity or a different clinical conclusion of 95% CI versus 95% PI (see 'range of equivalence')) and 10 or more studies per outcome were available. We restricted subgroup analyses to GRADE‐relevant primary outcomes and considered non‐overlapping subgroup CIs as a statistically significant difference. Mixed‐effects meta‐regression models were not used due to an insufficient number of trials. We used the R packages meta and metafor for mixed‐effects meta‐regression (Schwarzer 2007b; Viechtbauer 2010). We analysed data pertaining to the following subgroups, if possible:

1. surgical procedures (e.g. breast, cardiac, thoracic, abdominal and hip surgery);

2. local anaesthetic volume (more versus less than 20 mL of local anaesthetic);

3. use of a fixed multimodal analgesia (combination of opioids and non‐opioid);

4. use of a baseline analgesia with non‐opioid.

Sensitivity analysis

We performed sensitivity analyses that focussed on the following issues in the primary outcomes.

1. Including only ‘overall low risk of bias results’ and excluding results with ‘some concerns’ or ‘high risk of bias’.

2. Excluding studies with high levels of attrition (> 15% missing outcome data).

3. Effect estimate using fixed‐effect versus random‐effects model.

4. Influence of inclusion of randomised trials with zero events.

Summary of findings and assessment of the certainty of the evidence

We used the GRADE approach to assess the certainty of evidence associated with the following (primary) outcomes in our review (Schünemann 2019b; Schünemann 2019).

1. Mean difference in postoperative pain intensity at rest (24 hours postoperatively)

2. Rates of block‐related adverse events (any kind of neurological complication, accidental vascular puncture, motor blockade, local anaesthetic toxicity)

We constructed summary of findings tables for each comparison. Through the GRADE approach, we appraised the certainty of evidence on the basis of the extent to which one can be confident that the estimated effect reflects the true effect. The certainty of the body of evidence was based on decisions about within‐study risk of bias (study limitations), indirectness, imprecision of the effect estimate, heterogeneity (inconsistency) and risk of publication bias.

Within‐study risk of bias: we judged the certainty of evidence as adequate for within‐study risk of bias if most information was derived from studies at 'overall low risk of bias'. We downgraded the certainty of evidence when the proportion of data from studies at high risk or some concern of bias was sufficient to affect interpretation of results (determined using sensitivity analysis).

Indirectness: we judged the certainty of evidence as adequate if outcome data were based on direct comparisons of interest, on the population of interest and on the outcome of interest.

Imprecision: the 95% CI expresses the uncertainty of the estimated treatment effect. Imprecise treatment effects, with a wide confidence interval, may be associated with high uncertainty regarding the clinical decision. We considered the range of treatment effects included in the confidence interval and compared it with the respective range of equivalence. The inclusion of different clinical implications in the confidence interval indicates an imprecise treatment effect, and we downgraded once (some concerns) or twice (major concerns) for imprecision as depicted in Figure 1 and Figure 2 (Nikolakopoulou 2019).

Inconsistency: we downgraded the certainty of evidence for inconsistency by one level when statistical heterogeneity was present (P < 0.01 for the Chi² test of heterogeneity or a different clinical conclusion of 95% CI versus 95% PI existed (see 'range of equivalence')) without a satisfactory explanation by subgroup analysis.

Publication bias: for publication bias, we downgraded the certainty of evidence by one level when the statistical test for funnel plot asymmetry suggested publication bias, and when the adjustment for small‐study effects as assessed by Duval and Tweedie’s fill and trim analysis changed the conclusion.

The GRADE assessment resulted in one of four levels of 'certainty'. These expressed our confidence in the estimate of effect.

• High certainty evidence: further research is very unlikely to change our confidence in the estimated effect.

• Moderate certainty evidence: further research is likely to have an important impact on our confidence in the estimated effect and may change the estimate.

• Low certainty evidence: further research is very likely to have an important impact on our confidence in the estimated effect and is likely to change the estimate.

• Very low certainty evidence: any estimate of effect is very uncertain.

Results

Description of studies

Results of the search

We ran the literature search on 4 January 2021 using a predetermined search strategy (Appendix 1) and updated the search on 3 January 2022 to identify new relevant studies. This search identified 939 results, of which 59 were duplicates. We used Rayyanto screen and organise the results of the literature search. After deduplication, 880 studies remained and, of those, 806 studies were excluded for being irrelevant after title and abstract screening (Figure 3). Full‐text screening of 74 studies identified a further five studies for exclusion. Reasons for exclusion were wrong format, wrong outcome, lack of translation and one study that had been retracted (Tulgar 2019). We used the PICO criteria as shown in Table 10 to determine inclusion and exclusion criteria.

3.

PRISMA flow chart of the search process for this meta‐analysis

1. PICO inclusion and exclusion criteria for literature search.

Variable	Inclusion criteria	Exclusion criteria
Type of study	Randomised controlled trials (RCTs)	Cluster‐randomised trials, cross‐over trials, quasi‐randomised trials, non‐randomised trials
Publication year	Any	None
Publication status	Published, in progress	Retracted
Language	All	None
Participants	Adults (≥ 18 years) undergoing surgery under general anaesthesia	Children (< 18 years), surgeries that did not include general anaesthesia
Intervention	Single or continuous ESPB versus no treatment, sham treatment or any other regional anaesthetic technique following an operation with general anaesthetic	Co‐interventions (i.e. ESPB in combination with other regional anaesthetic techniques)
Block technique	Landmarks, peripheral nerve stimulator or ultrasound	None
Training level of provider	Any	None
Concentration of local anaesthetic	Any	None
Database	Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, Embase and Web of Science	None

ESPB: erector spinae plane block

Five included trials could not be used for meta‐analysis, because there was only one trial each available for this comparison (tumescent analgesia, retrolaminar block, modified thoracolumbar interfacial plane block, wound infiltration and intrathecal morphine). These results were reported only qualitatively.

The PRISMA flow diagram visualises the flow of the literature search and screening process for this systematic review (Figure 3). Of the 64 trials included, 28 studies with 1611 participants compared erector spinae plane block (ESPB) against no block; 13 studies with 807 participants compared ESPB against placebo blocks; 11 studies with 859 participants compared ESPB against paravertebral block (PVB); four studies with 224 participants compared ESPB against transversus abdominis plane block (TAPB); three studies with 255 participants compared ESPB against serratus anterior plane block (SAPB); four studies with 212 participants compared ESPB against pectoralis plane block (PECSB); two studies with 140 participants compared ESPB against quadratus lumborum block (QLB); three studies with 241 participants compared ESPB against intercostal nerve block (ICNB) and two studies with 81 participants compared ESPB against epidurals (Table 11).

2. Overview of comparisons in included studies.

Comparison	Studies	Participants
ESPB vs no block (control)	28	1611
ESPB vs placebo (NaCl block)	13	807
ESPB vs PVB (paravertebral block)	11	859
ESPB vs TAP (transversus abdominis plane block)	4	224
ESPB vs PECS (pectoralis plane block)	4	212
ESPB vs SAPB (serratus anterior plane block)	4	255
ESPB vs ICNB (intercostal nerve block)	3	241
ESPB vs QLB (quadratus lumborum block)	2	140
ESPB vs epidural	2	81
ESPB vs ITM (intrathecal morphine)	1	54
ESPB vs mTLIP (modified‐thoracolumbar interfascial plane block)	1	90
ESPB vs MTPB (mid‐transverse process to pleura block)	1	40
ESPB vs TA (tumescent anaesthesia)	1	43
ESPB vs WI (wound infiltration)	1	50
ESPB vs RLB (retrolaminar block)	1	45

Summary view of the comparisons in all included studies. Those comparisons with only one study were excluded from the meta‐analysis, but results were reported narratively.

ESPB: erector spinae plane block

During the updated literature search on 3 January 2022, 37 new studies were found and subsequently categorised as awaiting classification.

Included studies

Study design and publication status

Sixty‐four RCTs, published between 2018 and 2021, with a total of 3973 randomised participants were eligible for our meta‐analyses; another five trials were included in the qualitative review (Figure 3).

The number of participants per study varied from 29 (Calia 2019) to 106 (Turhan 2020), with the mean and median study population being 63.3 and 60 respectively. Most studies were equally divided into two arms, while nine trials were divided into three arms (Abdelhamid 2020; Chen 2020; Ciftci 2020b; Ciftci 2020c; Elsabeeny 2020; Eskin 2020; Gürkan 2020; Ibrahim 2020; Turhan 2020). For these three‐armed studies we included each comparison separately. Forty‐eight trials published study protocols in different databases, whereas 16 trials did not publish any study protocol (Anushree 2020; Calia 2019; Chen 2020; Ciftci 2020c; Ciftci 2020d; Fu 2020; Gad 2019; Gultekin 2020; Guo 2019; Gürkan 2018; Nagaraja 2018; Qiang 2018; Sharma 2020; Sinha 2019; Turhan 2020; Zheng 2019). Almost all studies were published in journal format, but four trials were only published as conference abstracts at the time of this review (Anushree 2020; Arora 2019; Calia 2019; Yayik 2018).

Participants

All included trials investigated adult participants undergoing surgery with an American Society of Anesthesiology (ASA) class of I‐III. There was only one trial that also included ASA IV patients (Sobhy 2020).

The exclusion criteria in almost all trials were allergies to local anaesthetics, hepatic or renal disorders, current use of opioids, coagulation problems, infection at the injection site and obesity with a body mass index (BMI) of 30 or more. Two studies specifically investigated patients with a BMI above 40 kg/m² undergoing laparoscopic gastric sleeve surgery (Mostafa 2020; Abdelhamid 2020).

There were seven trials that did not publish any specific comments regarding their patients (Anushree 2020; Arora 2019; Calia 2019; Khorasanizadeh 2020; Singh 2019; Yang 2019; Yayik 2018).

Patients in the included trials underwent abdominal surgery (Abdelhamid 2020; Abu 2019; Aksu 2019a; Altıparmak 2019a; Anushree 2020; Aygun 2020; Fu 2020; Gultekin 2020; Hamed 2019; Ibrahim 2020; Kamel 2020; Kang 2019; Mostafa 2020; Sakae 2020; Prasad 2020; Wang 2019b), breast surgery (Aksu 2019a; Altıparmak 2019b; Arora 2019; El Ghamry 2019a; Elsabeeny 2020; Gad 2019; Gürkan 2018; Gürkan 2020; Khorasanizadeh 2020; Oksuz 2019; Park 2021; Seelam 2020; Sharma 2020; Singh 2019; Sinha 2019; Sotome 2021; Swisher 2020; Yao 2020b), cardiac surgery (Krishna 2019; Nagaraja 2018), hip surgery (Tulgar 2018), shoulder surgery (Ciftci 2020a), spine surgery (Calia 2019; Ciftci 2020b; El Ghamry 2019b; Eskin 2020; Singh 2020; Yayik 2018; Zhang 2020; Zhang 2021) and thoracic surgery (Chen 2020; Ciftci 2020c; Ciftci 2020d; Ekinci 2020; Fang 2019; Finnerty 2020; Fiorelli 2020; Gaballah 2019; Guo 2019; Liu 2021; Qiang 2018; Shim 2020; Sobhy 2020; Taketa 2019; Turhan 2020; Yang 2019; Yao 2020a; Yaoping 2019; Yuzhong 2018a; Yuzhong 2018b; Zhao 2020; Zheng 2019).

Intervention

These studies compared EPSB to no block, placebo blocks and a wide variety of other regional anaesthesia techniques (Table 11).

Twenty‐eight studies compared ESPB to no block (Abdelhamid 2020; Aksu 2019b; Aksu 2019a; Anushree 2020; Arora 2019; Calia 2019; Ciftci 2020d; Ciftci 2020c; Elsabeeny 2020; Eskin 2020; Fu 2020; Gultekin 2020; Gürkan 2018; Gürkan 2020; Krishna 2019; Liu 2021; Park 2021; Prasad 2020; Qiang 2018; Seelam 2020; Sharma 2020; Singh 2019; Singh 2020; Sobhy 2020; Yaoping 2019; Yayik 2018; Zhang 2020; Zheng 2019). Thirteen studies compared ESPB to placebo block (NaCl injection without local anaesthetic) as the comparison (Abu 2019; Ciftci 2020a; El Ghamry 2019b; Hamed 2019; Ibrahim 2019; Ibrahim 2020; Mostafa 2020; Shim 2020; Yang 2019; Yao 2020b; Yao 2020a; Yuzhong 2018a; Zhang 2021); 11 trials compared ESPB to paravertebral block (PVB) (Chen 2020; Ciftci 2020c; El Ghamry 2019a; Fang 2019; Guo 2019; Gürkan 2020; Swisher 2020; Taketa 2019; Turhan 2020; Yuzhong 2018b; Zhao 2020); four trials compared ESPB to transversus abdominis plane block (TAPB) (Abdelhamid 2020; Altıparmak 2019a; Ibrahim 2020; Kamel 2020); four trials compared ESPB to pectoralis plane block (PECSB) (Altıparmak 2019b; Gad 2019; Khorasanizadeh 2020; Sinha 2019) and serratus anterior plane block (SAPB) (Elsabeeny 2020; Ekinci 2020; Finnerty 2020; Gaballah 2019); three studies compared ESPB to intercostal nerve block (ICNB) (Chen 2020; Fiorelli 2020; Turhan 2020); two compared ESPB to quadratus lumborum block (QLB) (Aygun 2020; Tulgar 2018), while two compared ESPB to epidural techniques (Nagaraja 2018; Sakae 2020).

The following regional anaesthetic and neuraxial anaesthetic techniques were only used in one study each: modified thoracolumbar interfascial plane block; mid‐transverse process to pleura block; intrathecal morphine; tumescent analgesia; retrolaminar block and wound infiltration. They were not included in the meta‐analyses, but we reported the results narratively (Table 12).

3. Narrative reporting of primary results from studies that were not included in a meta‐analysis due to limited data.

Study and reason for exclusion from meta‐analysis	1.1 ‐ Pain intensity at rest 24 hours after surgery: mean (SD)	1.2 ‐ Block‐related adverse events
	Results	Results
Ciftci 2020b Only 1 RCT in this comparison	ESPB 0 (0.25), mTLIP 0 (0.25), opiates 0 (0.25).	0 in all groups
Kang 2019 Only 1 RCT in this comparison	ESPB 2.5 (1.0), ITM 1.3 (1.1)	0 in all groups
Oksuz 2019 Only 1 RCT in this comparison	ESPB 0.95 (0.49), TA 2.09 (0.61)	0 in all groups
Sotome 2021 Only 1 RCT in this comparison	ESPB 1 (0), RLB 1 (0)	0 in all groups
Wang 2019b Only 1 RCT in this comparison	—	0 in all groups

Abbreviations:

ESPB: erector spinae plane block
ITM: intrathecal morphine
mTLIP: modified thoracolumbar interfascial plane block
RCT: randomised controlled trial
RLB: retrolaminar block
TA: tumescent anaesthesia

Most studies used ultrasound for the application of a single‐shot ESP block. There were only two trials, which applied an ESP catheter (Taketa 2019; Nagaraja 2018). One study used a landmark technique for the epidural catheter placement, but an ultrasound‐guided approach for the ESPB (Nagaraja 2018); one study used fluoroscopy for the application of an ESPB (Prasad 2020), while another study compared an intraoperatively placed ICNB with an ESPB (Turhan 2020).

Depending on the surgical procedure an ESPB was placed either unilaterally or bilaterally with a local anaesthetic volume on each side varying between 15 (e.g. Abdelhamid 2020) and 30 mL (e.g. Ciftci 2020a). However, most trials applied 20 mL of a long‐lasting local anaesthetic. The type of local anaesthetic varied between bupivacaine, levobupivacaine and ropivacaine at different concentrations (0.2% (e.g. Anushree 2020) to 0.75% (e.g. Fiorelli 2020)). Only one included trial applied a mixture of long‐lasting and short‐lasting local anaesthetic (Aygun 2020). A mixture of 0.5 μg/kg dexmedetomidine in combination with ropivacaine 0.75% was applied only in one trial (Fiorelli 2020). Furthermore, nine included trials applied postoperative nausea and vomiting (PONV) prophylaxis with intravenous dexamethasone, which might have prolonged the analgesic duration of local analgesia (Abdelhamid 2020; Altıparmak 2019a; Altıparmak 2019b; Elsabeeny 2020; Eskin 2020; Fiorelli 2020; Kang 2019; Oksuz 2019; Sotome 2021). ESP blocks were either placed in awake or sedated patients preoperatively or after anaesthetic induction during anaesthesia.

Outcome measures

The included trials used a visual analogue scale (VAS), numerical rating scale (NRS) or a verbal rating scale (VRS) (all using scales ranging from 0 to 10 points) for measuring pain intensity. For simplicity, we chose to use the most commonly used scale (VAS), as predetermined in our protocol.

For the measurement of opioid requirement we converted all opioids used in the included studies to oral morphine equivalents in milligrams, in order to make comparison easier. For this, we used the opioid calculator app of the Faculty of Pain Medicine of the Australian and New Zealand College of Anaesthetists (http://www.opioidcalculator.com.au), as stated in our protocol.

Ongoing studies

During our literature search three ongoing studies were found (Sarkar 2022; Singh 2022; Zengin 2022). Sarkar 2022 is a RCT, which investigates the effects of ESPB compared to QLB on patients undergoing laparoscopic nephrectomy. Singh 2022 is a placebo‐controlled RCT, which investigates the effects of ESPB on patients undergoing lumbar spine surgery. Zengin 2022 is a RCT, which investigates the effects of ESPB compared to SAPB on patients undergoing video‐assisted thoracic surgery (VATS).

Studies awaiting classification

During the updated literature search on 3 January 2022, 37 new studies were found and subsequently categorised as awaiting classification. These 37 RCTs compare the efficacy of ESPB with other regional anaesthetic procedures or placebo on outcomes such as postoperative pain intensity and opioid consumption in patients undergoing a variety of surgical procedures with general anaesthesia. Seventeen trials compared ESPB with no block (Asar 2022; Finnerty 2021; Genc 2021; Iqbal 2021; Jin 2021; Kim 2021; Park 2021a; Piskin 2021; Sifaki 2021; Taskaldiran 2021; Wahdan 2021; Wang 2021a; Yildiz 2021; Zengin 2021; Zhang 2021a; Zhang 2021b; Zhang 2021c), nine trials compared ESPB with a placebo block (Athar 2021; Dost 2021; Sarkar 2021; Shanthanna 2021; Theodoraki 2021; Verma 2020; Yörükoğlu 2021; Zhu 2021; Zimmerer 2021), two trials compared ESPB with PECSB (Abraham 2021; Genc 2021), one trial compared ESPB with PVB (Agarwal 2021), one trial compared ESPB with RLB (Zhao 2021), three trials compared ESPB with TEA (Elsabeeny 2021; Khan 2021; Zhang 2021c), two trials compared ESPB with TAPB (Ozdemir 2021; Shen 2021), one trial compared ESPB with SAPB (Elsabeeny 2021), one trial compared ESPB with TLIP (Wang 2021), two trials compared ESPB with subcutaneous infiltration (Ramachandran 2021; Rao 2021) and one trial compared ESPB with QLB (Abd 2021).

Excluded studies

After full‐text screening, we excluded five studies. Tao 2019 was excluded due to its retrospective study design. Qiulan 2019 was excluded because they did not investigate outcomes relevant to our meta‐analysis. Sajna 2020 was excluded because they used the wrong procedure. Wang 2019a was published in Chinese and we were unable to translate it to English or German in order to include it in our results. One study was excluded as the study has been retracted (Tulgar 2019).

Furthermore, five included studies were excluded from the quantitative meta‐analysis.

Risk of bias in included studies

We assessed the risk of bias (RoB) of results from 64 studies included in our meta‐analysis (Abdelhamid 2020; Abu 2019; Aksu 2019a; Aksu 2019b; Altıparmak 2019a; Altıparmak 2019b; Anushree 2020; Arora 2019; Aygun 2020; Calia 2019; Chen 2020; Ciftci 2020a; Ciftci 2020c; Ciftci 2020d; Ekinci 2020; El Ghamry 2019a; El Ghamry 2019b; Elsabeeny 2020; Eskin 2020; Fang 2019; Finnerty 2020; Fiorelli 2020; Fu 2020; Gaballah 2019; Gad 2019; Gultekin 2020; Guo 2019; Gürkan 2018; Gürkan 2020; Hamed 2019; Ibrahim 2019; Ibrahim 2020; Kamel 2020; Khorasanizadeh 2020; Krishna 2019; Liu 2021; Mostafa 2020; Nagaraja 2018; Park 2021; Prasad 2020; Qiang 2018; Sakae 2020; Seelam 2020; Sharma 2020; Shim 2020; Singh 2019; Singh 2020; Sinha 2019; Sobhy 2020; Swisher 2020; Taketa 2019; Tulgar 2018; Turhan 2020; Yang 2019; Yao 2020a; Yao 2020b; Yaoping 2019; Yayik 2018; Yuzhong 2018a; Yuzhong 2018b; Zhang 2020; Zhang 2021; Zhao 2020; Zheng 2019), using the RoB 2 tool as recommended in the Cochrane Handbook for Systematic Reviews of Interventions.

The 64 studies meta‐analysed contributed to 78 results for two primary outcomes in the respective summary of findings tables. Overall, 57 studies reported one or both primary outcomes. Only one study reported results on chronic pain after surgery.

Out of the 78 study results, we rated 44 (56%) as low risk of bias, 24 (31%) as some concerns of risk of bias and 10 (13%) as high risk of bias.

An overview of the risk of bias assessments for the primary outcomes can be seen in the figures section (Figure 4; Figure 5). All outcomes included a mixture of low, some concerns and high risk of bias studies.

4.

Risk of bias summary for primary outcome of postoperative pain at 24 hours

5.

Risk of bias summary for primary outcome of block‐related adverse events

Effects of interventions

See: Table 1; Table 2; Table 3; Table 4; Table 5; Table 6; Table 7; Table 8; Table 9

We compared ESPB with no block, placebo blocks and other regional anaesthetic techniques in 64 RCTs, including 3973 patients in total. We separated the data into nine comparisons (ESPB versus no blocks, placebo, PVB, TAPB, SAPB, PECSB, QLB, ICNB and epidural). For each comparison we investigated 16 outcomes as described in the methodology section (Primary outcomes; Secondary outcomes). We performed meta‐analysis for each outcome in each comparison whenever the data permitted.

Our findings for the different comparisons are summarised in: Table 1; Table 2; Table 3; Table 4; Table 5; Table 6; Table 7; Table 8; Table 9. Other meta‐analysis results for the secondary outcomes can be found in: Table 13; Table 14; Table 15; Table 16. We reported outcomes that could not be analysed within meta‐analyses narratively in table format (Table 12; Table 17). Sensitivity analyses focusing on the influence of high risk of bias or some concerns, and using a fixed‐effect model, can be found in Table 18. There is no sensitivity analysis focusing on the influence of a high level of attrition because there was only one trial with a high attrition level (Shim 2020).

4. Cumulative mean oral morphine requirement in mg at different times postoperatively (mg oral equivalents of morphine).

Comparisons	Number of studies (participants)	Effect estimate (95% CI) – random‐effects meta‐analysis	Inconsistency (95% PI), heterogeneity (I2)
1. Erector spinae plane block vs no block
1.10 Cumulative mean oral morphine requirement in mg ‐ 2 hours	1 (60)	MD ‐13.47 (‐15.49 to ‐11.44)	NA
1.11 Cumulative mean oral morphine requirement in mg ‐ 24 hours	23 (1341)	MD ‐33.59 (‐56.37 to ‐10.81)	I² = 100% 95% PI (−107.13 to 40.49)
1.12 Cumulative mean oral morphine requirement in mg ‐ 48 hours	4 (260)	MD ‐54.14 (‐69.09 to ‐39.20)	I² = 100% 95% PI (−330.80 to 211.74)
2. Erector spinae plane block vs placebo
Cumulative mean oral morphine requirement in mg ‐ 2 hours	0	—	—
2.8 Cumulative mean oral morphine requirement in mg ‐ 24 hours	11 (677)	MD ‐21.05 (‐33.17 to ‐8.93)	I² = 98% 95% PI (−54.39 to 11.79)
2.9 Cumulative mean oral morphine requirement in mg ‐ 48 hours	3 (195)	MD ‐26.62 mg (‐46.41, ‐6.84)	I² = 75%
3. Erector spinae plane block vs paravertebral block
3.10 Cumulative mean oral morphine requirement in mg ‐ 2 hours	1 (90)	—	—
3.11 Cumulative mean oral morphine requirement in mg ‐ 24 hours	8 (496)	MD 6.94 (‐2.65 to 16.53)	I² = 95% 95% PI −27.37 to 41.27)
3.12 Cumulative mean oral morphine requirement in mg ‐ 48 hours	5 (336)	MD 9.57 (‐4.33 to 23.47)	I² = 87% 95% PI (−45.94 to 64.86)
4. Erector spinae plane block vs transversus abdominis plane block
Cumulative mean oral morphine requirement in mg ‐ 2 hours	0	—	—
4.4 Cumulative mean oral morphine requirement in mg ‐ 24 hours	3 (158)	MD ‐17.31 (‐41.42 to 6.79)	I² = 99%
Cumulative mean oral morphine requirement in mg ‐ 48 hours	0	—	—
5. Erector spinae plane block vs serratus anterior plane block
Cumulative mean oral morphine requirement in mg ‐ 2 hours	0	—	—
5.3 Cumulative mean oral morphine requirement in mg ‐ 24 hours	2 (110)	MD ‐3.36 (‐17.50 to 10.78)	I² = 45%
Cumulative mean oral morphine requirement in mg ‐ 48 hours	0	—	—
6. Erector spinae plane block vs pectoralis plane block
Cumulative mean oral morphine requirement in mg ‐ 2 hours	0	—	—
6.3 Cumulative mean oral morphine requirement in mg ‐ 24 hours	3 (162)	MD 0.21 (‐0.16 to 0.58)	I² = 45%
Cumulative mean oral morphine requirement in mg ‐ 48 hours	0	—	—
7. Erector spinae plane block vs quadratus lumborum block
Cumulative mean oral morphine requirement in mg ‐ 2 hours	0	—	—
7.4 Cumulative mean oral morphine requirement in mg ‐ 24 hours	2 (120)	MD ‐0.31 (‐2.22 to 1.61)	I² = 0%
Cumulative mean oral morphine requirement in mg ‐ 48 hours	0	—	—
8. Erector spinae plane block vs intercostal nerve block
Cumulative mean oral morphine requirement in mg ‐ 2 hours	0	—	—
8.6 Cumulative mean oral morphine requirement in mg ‐ 24 hours	2 (121)	MD 9.02 (3.05 to 14.99)	I² = 75%
8.7 Cumulative mean oral morphine requirement in mg ‐ 48 hours	2 (121)	MD 7.98 (2.60 to 13.36)	I² = 0%
9. Erector spinae plane block vs epidural analgesia
Cumulative mean oral morphine requirement in mg ‐ 2 hours	0	—	—
9.7 Cumulative mean oral morphine requirement in mg ‐ 24 hours	1 (31)	MD 1.75 (‐1.44 to 4.94)	NA
Cumulative mean oral morphine requirement in mg ‐ 48 hours	0	—	—

All opioids used in the trials were converted to oral morphine requirements.

Abbreviations:

CI: confidence interval
MD: mean difference
NA: not applicable
PI: prediction interval

5. Mean difference in postoperative pain at rest at different time points postoperatively (VAS).

Comparisons	Number of studies (participants)	Effect estimate (95% CI) – random‐effects meta‐analysis	Inconsistency (95% PI), heterogeneity (I2)
1. Erector spinae plane block vs no block
1.4 Mean difference in postoperative pain intensity at rest ‐ 2 hours	10 (611)	MD ‐2.08 (‐2.59 to ‐1.58)	I² = 91% 95% PI (−4.34 to 0.20)
1.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	17 (958)	MD ‐0.77 (‐1.08 to ‐0.46)	I² = 89% 95% PI (−1.98 to 0.44)
1.5 Mean difference in postoperative pain intensity at rest ‐ 48 hours	6 (360)	MD ‐0.47 (‐0.79 to ‐0.14)	I² = 89% 95% PI (−2.07 to 1.09)
2. Erector spinae plane block vs placebo
2.4 Mean difference in postoperative pain intensity at rest ‐ 2 hours	7 (439)	MD ‐1.46 (‐2.44 to ‐0.48)	I² = 97% 95% PI (−6.42 to 3.47)
2.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	8 (499)	MD ‐0.14 (‐0.29 to 0.00)	I² = 20% 95% PI (−1.22 to 1.07)
2.5 Mean difference in postoperative pain intensity at rest ‐ 48 hours	2 (120)	MD ‐0.49 (‐0.84 to ‐0.14)	I² = 9%
3. Erector spinae plane block vs paravertebral block
3.4 Mean difference in postoperative pain intensity at rest ‐ 2 hours	3 (211)	MD 0.26 (‐0.24 to 0.76)	I² = 59%
3.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	7 (478)	MD 0.23 (‐0.06 to 0.52)	I² = 84% 95% PI (−0.80 to 1.27)
3.5 Mean difference in postoperative pain intensity at rest ‐ 48 hours	4 (298)	MD ‐0.01 (‐0.14 to 0.13)	I² = 0% 95% PI (−0.30 to 0.29)
4. Erector spinae plane block vs transversus abdominis plane block
4.3 Mean difference in postoperative pain intensity at rest ‐ 2 hours	3 (160)	MD ‐0.33 (‐1.74 to 1.08)	I² = 98%
4.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	3 (160)	MD ‐0.16 (‐0.46 to 0.14)	I² = 69%
Mean difference in postoperative pain intensity at rest ‐ 48 hours	0	—	—
5. Erector spinae plane block vs serratus anterior plane block
5.2 Mean difference in postoperative pain intensity at rest ‐ 2 hours	1 (60)	MD ‐0.10 (‐0.40 to 0.20)	NA
Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	0	—	—
Mean difference in postoperative pain intensity at rest ‐ 48 hours	0	—	—
6. Erector spinae plane block vs pectoralis plane block
6.3 Mean difference in postoperative pain intensity at rest ‐ 2 hours	3 (162)	MD 0.21 (‐0.16 to 0.58)	I² = 45%
6.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	2 (98)	MD 0.24 (‐0.11 to 0.58)	I² = 41%
Mean difference in postoperative pain intensity at rest ‐ 48 hours	0	—	—
7. Erector spinae plane block vs quadratus lumborum block
Mean difference in postoperative pain intensity at rest ‐ 2 hours	0	—	—
7.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	1 (40)	MD 0.05 (‐0.21 to 0.31)	NA
Mean difference in postoperative pain intensity at rest ‐ 48 hours	0	—	—
8. Erector spinae plane block vs intercostal nerve block
Mean difference in postoperative pain intensity at rest ‐ 2 hours	0	—	—
8.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	2 (131)	MD ‐0.33 (‐3.02 to 2.35)	I² = 96%
8.4 Mean difference in postoperative pain intensity at rest ‐ 48 hours	1 (60)	MD ‐1.37 (‐2.47 to ‐0.27)	NA
9. Erector spinae plane block vs epidural analgesia
9.4 Mean difference in postoperative pain intensity at rest ‐ 2 hours	1 (31)	MD 3.00 (1.52 to 4.48)	NA
9.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	2 (81)	MD 1.20 (‐2.52 to 4.93)	I² = 96%
9.6 Mean difference in postoperative pain intensity at rest ‐ 48 hours	1 (50)	MD ‐1.20(‐1.78 to ‐0.62)	NA

*Clinically relevant assumptions for pain: MD ‐1 point on VAS 0 to 10.

Abbreviations:

CI: confidence interval
MD: mean difference
NA: not applicable
PI: prediction interval

6. Mean difference in postoperative pain during activity at different time points postoperatively (VAS).

Comparisons	Number of studies (participants)	Effect estimate (95% CI) – random‐effects meta‐analysis	Inconsistency (95% PI), heterogeneity (I2)
1. Erector spinae plane block vs no block
1.6 Mean difference in postoperative pain intensity during activity ‐ 2 hours	4 (240)	MD ‐3.03 (‐4.29 to ‐1.77)	I² = 96% 95% PI (−9.11 to 3.05)
1.3 Mean difference in postoperative pain intensity during activity (24 hours postoperatively)	9 (518)	MD ‐1.31 (‐1.69 to ‐0.93)	I² = 84% 95% PI (−2.70 to 0.07)
1.7 Mean difference in postoperative pain intensity during activity ‐ 48 hours	5 (280)	MD ‐0.48 (‐0.81 to ‐0.16)	I² = 72% 95% PI (−1.51 to 0.55)
2. Erector spinae plane block vs placebo
2.6 Mean difference in postoperative pain intensity during activity ‐ 2 hours	—	—	—
2.3 Mean difference in postoperative pain intensity during activity (24 hours postoperatively)	3 (199)	MD ‐0.10 (‐0.19 to ‐0.02)	I² = 0%
2.7 Mean difference in postoperative pain intensity during activity ‐ 48 hours	—	—	—
3. Erector spinae plane block vs paravertebral block
3.6 Mean difference in postoperative pain intensity during activity ‐ 2 hours	2 (141)	MD 0.23 (‐1.14 to 1.60)	I² = 76%
3.3 Mean difference in postoperative pain intensity during activity (24 hours postoperatively)	6 (408)	MD 0.24 (‐0.06 to 0.53)	I² = 79% 95% PI (−0.88 to 1.41)
3.7 Mean difference in postoperative pain intensity during activity ‐ 48 hours	4 (298)	MD ‐0.02 (‐0.31 to 0.27)	I² = 48% 95% PI (−1.19 to 1.15)
4. Erector spinae plane block vs transversus abdominis plane block
Mean difference in postoperative pain intensity during activity ‐ 2 hours	0	—	—
Mean difference in postoperative pain intensity during activity (24 hours postoperatively)	0	—	—
Mean difference in postoperative pain intensity during activity ‐ 48 hours	0	—	—
5. Erector spinae plane block vs serratus anterior plane block
Mean difference in postoperative pain intensity during activity ‐ 2 hours	0	—	—
Mean difference in postoperative pain intensity during activity (24 hours postoperatively)	0	—	—
Mean difference in postoperative pain intensity during activity ‐ 48 hours	0	—	—
6. Erector spinae plane block vs pectoralis plane block
Mean difference in postoperative pain intensity during activity ‐ 2 hours	0	—	—
Mean difference in postoperative pain intensity during activity (24 hours postoperatively)	0	—	—
Mean difference in postoperative pain intensity during activity ‐ 48 hours	0	—	—
7. Erector spinae plane block vs quadratus lumborum block
Mean difference in postoperative pain intensity during activity ‐ 2 hours	0	—	—
7.3 Mean difference in postoperative pain intensity during activity (24 hours postoperatively)	1 (80)	MD 0.00 (‐0.65 to 0.65)	NA
Mean difference in postoperative pain intensity during activity ‐ 48 hours	0	—	—
8. Erector spinae plane block vs intercostal nerve block
Mean difference in postoperative pain intensity during activity ‐ 2 hours	0	—	—
8.3 Mean difference in postoperative pain intensity during activity (24 hours postoperatively)	2 (131)	MD ‐1.52 (‐4.61 to 1.56)	I² = 95%
8.5 Mean difference in postoperative pain intensity during activity ‐ 48 hours	1 (60)	MD ‐1.73 (‐3.02 to ‐0.44)	NA
9. Erector spinae plane block vs epidural analgesia
Mean difference in postoperative pain intensity during activity ‐ 2 hours	0	—	—
9.3 Mean difference in postoperative pain intensity during activity (24 hours postoperatively)	1 (50)	MD ‐0.72 (‐1.14 to ‐0.30)	NA
9.6 Mean difference in postoperative pain intensity during activity ‐ 48 hours	1 (50)	MD ‐1.36 (‐1.96 to ‐0.76)	NA

*Clinically relevant assumptions for pain: MD ‐1 point on VAS 0 to 10.

Abbreviations:

CI: confidence interval
MD: mean difference
NA: not applicable
PI: prediction interval

7. Rates of opioid‐related adverse events (PONV and pruritus).

Comparisons	Number of studies (participants)	Effect estimate (95% CI) – random‐effects meta‐analysis	Inconsistency (95% PI), heterogeneity (I2)	Zero event analysis in R	NNTB/NNTH (1/RR) (for outcomes with > 4 studies)
1. Erector spinae plane block vs no block
1.13 Rates of opioid‐related adverse events (PONV)	21 (1111)	RR 0.46 (0.36 to 0.60)	I² = 19%, 95% PI (0.23 to 0.90)	—	2.17
1.14 Rates of opioid‐related adverse events (itching)	7 (440)	RR 0.41 (0.26 to 0.63)	I² = 0%, 95% PI (0.18 to 0.92)	—	2.44
2. Erector spinae plane block vs placebo
2.10 Rates of opioid‐related adverse events (PONV)	11 (695)	RR 0.39 (0.28 to 0.56)	I² = 0%, 95% PI (0.26 to 0.59)	—	2.56
2.11 Rates of opioid‐related adverse events (itching)	2 (110)	RR 0.51 (0.21 to 1.28)	I² = 51%	—	NA
3. Erector spinae plane block vs paravertebral block
3.13 Rates of opioid‐related adverse events (PONV)	10 (668)	RR 1.13 (0.81 to 1.58)	I² = 0%, 95% PI (0.65 to 1.98)	RR 1.14 (0.80 to 1.62) I2 = 0%	0.88
3.14 Rates of opioid‐related adverse events (itching)	1 (60)	—	—	—	NA
4. Erector spinae plane block vs transversus abdominis plane block
4.5 Rates of opioid‐related adverse events (PONV)	3 (160)	RR 0.71 (0.44 to 1.16)	I² = 0%	—	NA
Rates of opioid‐related adverse events (itching)	0	—	—	—	—
5. Erector spinae plane block vs serratus anterior plane block
5.4 Rates of opioid‐related adverse events (PONV)	3 (170)	RR 0.70 (0.29 to 1.71)	NA	RR 0.71 (0.30 to 1.70) I² = 0%	NA
5.5 Rates of opioid‐related adverse events (itching)	2 (120)	RR 0.57 (0.19 to 1.75)	NA	RR 0.60 (0.20 to 1.75) I² = 0%	NA
6. Erector spinae plane block vs pectoralis plane block
Rates of opioid‐related adverse events (PONV)	0	—	—	—	—
Rates of opioid‐related adverse events (itching)	0	—	—	—	—
7. Erector spinae plane block vs quadratus lumborum block
7.5 Rates of opioid‐related adverse events (PONV)	2 (120)	RR 1.00 (0.35 to 2.84)	NA	RR 1.00 (0.37 to 2.74) I² = 0%	NA
Rates of opioid‐related adverse events (itching)	0	—	—	—	—
8. Erector spinae plane block vs intercostal nerve block
8.8 Rates of opioid‐related adverse events (PONV)	2 (121)	RR 1.26 (0.40, 3.98)	I² = 0%	—	NA
Rates of opioid‐related adverse events (itching)	0	—	—	—	—
9. Erector spinae plane block vs epidural analgesia
Rates of opioid‐related adverse events (PONV)	0	—	—	—	—
9.8 Rates of opioid‐related adverse events (itching)	1 (31)	—	—	—	—

Abbreviations:

CI: confidence interval
ESPB: erector spinae plane block
ICNB: intercostal nerve block
NA: not applicable
NE: not estimable
NNTB: number needed to treat for an additional beneficial outcome
NNTH: number needed to treat for an additional harmful outcome
PECSB: pectoralis block
PI: prediction intervals
PONV: postoperative nausea and vomiting
PVB: paravertebral block
QLB: quadratus lumborum block
RR: risk ratio
SAPB: serratus anterior plane block
TAP: transversus abdominis plane block

Footnotes:

*Clinically relevant assumptions for pain: if RR > 1.1 or RR < 0.9.

*NNTB/NNTH only calculated if > 4 trials included.

8. Side effects ‐ summary of all studies.

Study ID	Comparison	PONV (%)		Itching (%)		Other side effects (%)
		ESPB	Comparison	EPSB	Comparison	ESPB	Comparison
1	Abdelhamid 2020	36.4%	Opioids: 63.6% TAPB: 40.9%	—	—	—	—
2	Aksu 2019	13%	Opioids: 39.1%	—	—	—	—
3	Aksu 2019a	28%	Opioids: 24%	—	—	—	—
4	Abu 2019	6.66%	Sham: 10%	—	—	Hypotension: ESPB: 6.66%	Sham: 3.33%
5	Altiparmak 2019	—	—	—	—	—	—
6	Altiparmak 2019a	26.47%	TAPB: 38.24%	—	—	Bradycardia: ESPB: 5.9%	TAPB: 2.9%
7	Anushree 2020	23.8%	Sham: 66.6%	—	—	—	—
8	Arora 2019	—	—	—	—	—	—
9	Aygun 2020	ESPB: 15%	QLB: 15%	—	—	Shoulder pain: ESPB: 7.5%	QLB: 12.5%
10	Calia 2019	—	—	—	—	—	—
11	Chen 2020	20.83%	PVB: 16.67% ICNB: 12.5%	—	—	—	—
12	Ciftci 2020	16.67%	Opioids: 73.33%	10%	Opioids: 56.67%	—	—
13	Ciftci 2020a	26.67%	Sham: 60%	13.33%	Opioids: 43.33%	—	—
14	Ciftci 2020b	10%	mTLIP: 10%, Opiates: 43.33%	23.33%	mTLIP: 13.33% Opioids: 13.33%	—	—
15	Ciftci 2020c	16.67%	PVB: 20% Opiates: 60%	20%	PVB: 23.33% Opioids: 56.67%	—	—
16	Ekinci 2020	13.33%	SAPB: 20%	13.33%	Opioids: 23.34%	—	—
17	El 2019	6.7%	Sham: 16.7%	—	—	Somnolence: ESPB 20%	Sham: 0%
18	El 2019a	28.6%	PVB 34.3%	—	—	—	—
19	Elsabeeny 2020	0%	SAPB: 0% Opiates: 0%	—	—	—	—
20	Eskin 2020	2.5%	MTPB 7.5% Opiates 17.5%	2.5%	TMPB 7.5% Opioids: 15%	—	—
21	Fang 2019	17.8%	PVB 23.9%	—	—	Hypotension: ESPB: 6.7% Bradycardia: ESPB 0%	Hypotension: PVB 21.7% Bradycardia: PVB 8.7%
22	Finnerty 2020	5%	SAPB: 6.67%	0%	SAPB 0%	—	—
23	Fiorelli 2020	—	—	—	—	0%	0%
24	Fu 2020	6.7%	Opioids: 26.7%	10.0%	Opioids: 13.4%	Hypotension: ESPB: 6.7% Headache: ESPB 6.7% Respiratory depression: ESPB 0%	Hypotension: Opiates: 10% Headache: Opiates: 20.0% Respiratory depression: Opiates 3.3%
25	Gaballah 2019	—	—	—	—	—	—
26	Gad 2019	—	—	—	—	—	—
27	Gultekin 2020	—	—	—	—	—	—
28	Guo 2019	5%	PVB 2.5%	—	—	Hypotension: ESPB 2.5%	Hypotension: PVB 0%
29	Gürkan 2018	0.16%	Opioids: 0.2%	—	—	—	—
30	Gürkan 2020	0.14%	PVB: 4% Opiates: 24%	—	—	—	—
31	Hamed 2019	—	—	—	—	—	—
32	Ibrahim 2019	24%	Sham 19 (76%)	28%	Sham: 36%	Headache: ESPB: 48% Dizziness: 20%	Headache: Sham 40% Dizziness: Sham 80%
33	Ibrahim 2020	—	—	—	—	—	—
34	Kamel 2020	8.33%	TAPB 24%	—	—	—	—
35	Kang 2019	44.4%	ITM 59.3%	7.4%	ITM: 92.6%	Hypotension ESPB 14.8%	Hypotension: ITM 25.9%
36	Khorasanizadeh 2020	—	—	—	—	—	—
37	Krishna 2019	—	—	—	—	—	—
38	Liu 2021	17.5%	Opioids: 27.5%	5%	Opioids: 12.5%	Respiratory depression: 2.5% Hypotension: 2.5%	Respiratory depression: Opiates 2.5% Hypotension: Opiates 7.5%
39	Mostafa 2020	3.33%	Sham 6.67%	—	—	Hypotension: 1.67%	Hypotension: Sham 3.33%
40	Nagaraja 2018	—	—	—	—	—	—
41	Oksuz 2019	0%	Tumescent analgesia 18.18%	—	—	Hypotension: ESPB 4.76%	Hypotension: TA 4.76%
42	Park 2021	20.7%	37.9%	—	—	—	—
43	Prasad 2020	0%	Opioids: 3.28%	—	—	—	—
44	Qiang 2018	25%	Opioids: 60%	—	—	—	—
45	Sakae 2020	—	—	6.45%	Epidural: 9.68%	Urinary retention: ESPB 3.23%	Urinary retention: Epidural:12.9%
46	Seelam 2020	26%	30%	—	—	—	—
47	Sharma 2020	6.67%	Opioids: 13.33%	—	—	—	—
48	Shim 2020	2.17%	Sham 0%	—	—	—	—
49	Singh 2019	0%	Opioids: 12.5%	—	—	—	—
50	Singh 2020	0	Opioids: 5%	—	—	—	—
51	Sinha 2019	—	—	—	—	—	—
52	Sobhy 2020	3.33%	Opioids: 15%	Itching: ESPB 3.33%	Itching: Opioids: 8.33%	Hypotension: ESPB 3.33% Bradycardia: ESPB 3.33%	Hypotension: Control 8.33% Bradycardia: Control 8.33%
53	Sotome 2021	5%	RLB 13%	—	—	—	—
54	Swisher 2020	—	—	—	—	—	—
55	Taketa 2019	39%	PVB 20%	—	—	—	—
56	Tulgar 2018	0%	QLB 0% Opioids: 5%	—	—	—	—
57	Turhan 2020	0.94%	PVB 0% ICNB 1.89%	—	—	—	—
58	Wang 2019	6.7%	WI 33.3%	—	—	—	—
59	Yang 2019	10%	Sham 13.33%	—	—	—	—
60	Zheng 2019	22.5%	Opioids: 25%	15%	Opioids: 20%	Urinary retention: ESPB: 5%	Urinary retention: Opiates: 5%
61	Yao 2020	7.7%	Sham 22.5%	—	—	—	—
62	Yao 2020a	5.4%	Sham 18.4%	—	—	—	—
63	Yaoping 2019	—	—	—	—	—	—
64	Yayik 2018	—	—	—	—	—	—
65	Yuzhong 2018	16%	PVB 18%	—	—	—	—
66	Yuzhong 2018a	10%	Sham 28%	—	—	—	—
67	Zhang 2020	—	—	—	—	—	—
68	Zhang 2021	6.7%	Sham: 20%	—	—	Dizziness 6.7%	Dizziness: Sham: 10%
69	Zhao 2020	ESPB 0%	PVB 0%	—	—	—	—

All side effects reported in included studies. Percentages reported to 2 decimal figures.

Abbreviations:

ESPB: erector spinae plane block
ICNB: intercostal nerve block
mTLIP: modified thoracolumbar interfascial plane block
PECSB: pectoralis block
PVB: paravertebral block
QLB: quadratus lumborum block
SAPB: serratus anterior plane block
TAPB: transversus abdominis plane block
TMPB: mid‐transverse to pleura block
WI: wound infiltration

9. Sensitivity analyses.

Comparisons	Sensitivity analysis excluding studies that do not report study flow (95% CI)	Sensitivity analysis including only low risk of bias studies (95% CI)	Sensitivity analysis using fixed‐effect meta‐analysis (95% CI)	Original results using random‐effects meta‐analysis (95% CI)
1. Erector spinae plane block vs no block
1.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	MD ‐0.67 (‐1.00 to ‐0.34)	MD ‐0.67 (‐1.22 to ‐0.12)	MD ‐0.76 (‐0.85 to ‐0.67)	MD ‐0.77 (‐1.08 to ‐0.46)
1.2 Rates of block‐related adverse events	NA ‐ no sensitivity analysis possible in RevMan	NA ‐ no sensitivity analysis possible in RevMan	NA ‐ no sensitivity analysis possible in RevMan	No analysis possible in RevMan due to zero events In R: RR 1.00 (0.40 to 2.49)
2. Erector spinae plane block vs placebo
1.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	MD ‐0.15 (‐0.37 to 0.06)	MD ‐0.18 (‐0.42 to 0.05)	MD ‐0.12 (‐0.22 to ‐0.02)	MD ‐0.14 (‐0.29 to 0.00)
1.2 Rates of block‐related adverse events	NA ‐ no sensitivity analysis possible in RevMan	NA ‐ no sensitivity analysis possible in RevMan	NA ‐ no sensitivity analysis possible in RevMan	No analysis possible in RevMan due to zero events In R: RR 1.00 (0.29 to 3.41)
3. Erector spinae plane block vs paravertebral block
1.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	MD 0.14 (‐0.11 to 0.40)	MD 0.28 (‐0.06 to 0.62)	MD ‐0.05 (‐0.11 to 0.00)	MD 0.23 (‐0.06 to 0.52)
1.2 Rates of block‐related adverse events	RR 0.10 (0.02 to 0.55)	RR 0.11 (0.01 to 1.99)	RR 0.10 (0.02 to 0.55)	RR 0.10 (0.02 to 0.55) In R: RR 0.27 (0.08 to 0.95)
4. Erector spinae plane block vs transversus abdominis plane block
1.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	NA ‐ all studies reported study flow	MD ‐0.25 (‐0.63 to 0.13)	MD ‐0.21 (‐0.36 to ‐0.05)	MD ‐0.16 (‐0.46 to 0.14)
1.2 Rates of block‐related adverse events	NA ‐ all studies reported study flow	RR 1.00 (0.15 to 6.84)	RR 1.00 (0.07 to 15.00)	RR 1.00 (0.15 to 6.84) In R: RR 1.00 (0.21 to 4.83)
5. Erector spinae plane block vs serratus anterior plane block
1.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	—	—	—	—
1.2 Rates of block‐related adverse events	NA ‐ no sensitivity analysis possible in RevMan	NA ‐ no sensitivity analysis possible in RevMan	NA ‐ no sensitivity analysis possible in RevMan	Not estimable in RevMan In R: R 1.00 (0.06 to 15.59)
6. Erector spinae plane block vs pectoralis plane block
1.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	NA ‐ all studies reported study flow	NA ‐ not enough studies left	MD 0.27 (0.03 to 0.52)	MD 0.24 (‐0.11 to 0.58)
1.2 Rates of block‐related adverse events	—	—	—	—
7. Erector spinae plane block vs quadratus lumborum block
1.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	—	—	—	—
1.2 Rates of block‐related adverse events	—	—	—	—
8. Erector spinae plane block vs intercostal nerve block
1.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	NA ‐ not enough studies left	MD ‐1.74 (‐2.75 to ‐0.73)	MD 0.49 (0.06 to 0.93)	MD ‐0.33 (‐3.02 to 2.35)
1.2 Rates of block‐related adverse events	RR 0.09 (0.01 to 1.56)	NA	RR 0.09 (0.01 to 1.56)	RR 0.09 (0.01 to 1.56) In R: 0.32 (0.04 to 1.56)
9. Erector spinae plane block vs epidural analgesia
1.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively)	NA ‐ not enough studies left	MD 3.16 (1.81 to 4.51)	MD ‐0.30 (‐0.70 to 0.11)	MD 1.20 (‐2.52 to 4.93)
1.2 Rates of block‐related adverse events	—	—	—	—

95% PI and zero events analysis was only calculated for comparisons that included > 4 trials.

*Clinically relevant assumptions for pain: MD ‐1 point on VAS 0 to 10, SD = based on average SD of all included low risk of bias trials.

*Clinically relevant assumptions for pain: RR > 1.1 or < 0.9.

Abbreviations:

MD: mean difference
NA: not applicable
RR: risk ratio

Comparison 1: Erector spinae plane block versus no block

This comparison included 28 studies with 1611 participants in total comparing the analgesic effect of ESPB against no block (Abdelhamid 2020; Aksu 2019b; Aksu 2019a; Anushree 2020; Arora 2019; Calia 2019; Ciftci 2020d; Ciftci 2020c; Elsabeeny 2020; Eskin 2020; Fu 2020; Gultekin 2020; Gürkan 2018; Gürkan 2020; Krishna 2019; Liu 2021; Park 2021; Prasad 2020; Qiang 2018; Seelam 2020; Sharma 2020; Singh 2019; Singh 2020; Sobhy 2020; Yaoping 2019; Yayik 2018; Zhang 2020; Zheng 2019).

Primary outcomes

Mean difference in postoperative pain intensity at rest at 24 hours

Seventeen trials investigated postoperative pain intensity at rest at 24 hours (Abdelhamid 2020; Ciftci 2020d; Ciftci 2020c; Elsabeeny 2020; Eskin 2020; Fu 2020; Gultekin 2020; Park 2021; Prasad 2020; Singh 2019; Singh 2020; Sobhy 2020; Tulgar 2018; Yaoping 2019; Yayik 2018; Zhang 2020; Zheng 2019). There is probably no clinically relevant reduction in pain intensity at rest 24 hours after surgery in patients treated with ESPB compared to no block (mean difference (MD) ‐0.77 points, 95% confidence interval (CI) ‐1.08 to ‐0.46; 17 trials, 958 participants; Analysis 1.1). The 95% prediction interval (95% PI) ranges from ‐1.98 to 0.44. We downgraded the certainty of evidence by one level to moderate, due to imprecision (Table 1).

1.1. Analysis.

Comparison 1: Erector spinae plane block vs no block, Outcome 1: Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS)

Sensitivity and subgroup analyses

Sensitivity analyses focusing on the influence of studies with some concerns and high risk of bias (MD ‐0.67 points, 95% CI ‐1.22 to ‐0.12) or studies that did not declare the study flow (MD ‐0.67 points, 95% CI ‐1.00 to ‐0.34) did not show a significant difference. Similarly, the sensitivity analysis using a fixed‐effect meta‐analysis did not show a different result (MD ‐0.76, 95% CI ‐0.85 to ‐0.67) (Table 18). There is no sensitivity analysis focusing on the influence of a high level of attrition due to lack of data.

Due to high heterogeneity, we further analysed this outcome using subgroup analyses, accounting for type of surgery, type of local anaesthetic (LA), volume of LA, use of a fixed multimodal analgesia scheme and baseline non‐opioid analgesia. The subgroup analysis focusing on the influence of surgery showed that patients undergoing abdominal (MD ‐1.09 points, 95% CI ‐1.26 to ‐0.92) or thoracic surgery (MD ‐1.20 points, 95% CI ‐1.62 to ‐0.79) reported clinically relevant reduction in pain intensity at rest 24 hours after surgery in patients treated with ESPB compared to no block. This was not reported in trials focusing on breast, spine or hip surgery (Analysis 10.1). Within the other subgroup analyses the 95% CI overlapped and the subgroup analysis could therefore not explain the heterogeneity in the meta‐analysis results (Analysis 10.2; Analysis 10.3; Analysis 10.4; Analysis 10.5).

10.1. Analysis.

Comparison 10: Subgroup analysis: erector spinae plane block vs no block, Outcome 1: Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS) ‐ Type of surgery

10.2. Analysis.

Comparison 10: Subgroup analysis: erector spinae plane block vs no block, Outcome 2: Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS) ‐ Type of local anaesthetic

10.3. Analysis.

Comparison 10: Subgroup analysis: erector spinae plane block vs no block, Outcome 3: Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS) ‐ Volume of local anaesthetic

10.4. Analysis.

Comparison 10: Subgroup analysis: erector spinae plane block vs no block, Outcome 4: Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS) ‐ Fixed multimodal analgesia

10.5. Analysis.

Comparison 10: Subgroup analysis: erector spinae plane block vs no block, Outcome 5: Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS) ‐ Baseline non‐opioid analgesia

Rates of block‐related adverse events

Eighteen trials investigated rates of block‐related adverse events (Abdelhamid 2020; Aksu 2019a; Aksu 2019b; Ciftci 2020c; Elsabeeny 2020; Gürkan 2020; Krishna 2019; Liu 2021; Park 2021; Prasad 2020; Qiang 2018; Seelam 2020; Sharma 2020; Singh 2019; Singh 2020; Tulgar 2018; Yaoping 2019; Zhang 2020). There were no adverse events in either group in any of the studies included (Analysis 1.2). The 95% PI ranges from 0.37 to 2.69. We downgraded the certainty of evidence to low due to imprecision (double downgrade) (Table 1).

1.2. Analysis.

Comparison 1: Erector spinae plane block vs no block, Outcome 2: Rates of block‐related adverse events

Sensitivity and subgroup analyses

Further zero event sensitivity analysis in R showed that there is probably no difference in block‐related adverse events between the groups treated with an ESPB and those receiving no block (risk ratio (RR) 1.00, 95% CI 0.40 to 2.49). Other sensitivity analyses were not possible due to zero events. Subgroup analysis was not performed due to low heterogeneity (I² = 0%).

Secondary outcomes

Postoperative pain at rest/during activity at two hours, 24 hours and 48 hours

Ten trials investigated postoperative pain at rest after two hours and showed a clinically relevant decrease in pain intensity in the ESPB group (MD ‐2.08 points, 95% CI ‐2.59 to ‐1.58; 10 trials, 611 participants; Analysis 1.4) (Abdelhamid 2020; Ciftci 2020d; Ciftci 2020c; Eskin 2020; Fu 2020; Krishna 2019; Prasad 2020; Singh 2019; Singh 2020; Yayik 2018). Four trials investigated postoperative pain during activity at two hours and again showed a clinically relevant decrease in pain intensity in the ESPB group (MD ‐3.03 points, 95% CI ‐4.29 to ‐1.77; 4 trials, 240 participants; Analysis 1.6) (Ciftci 2020d; Ciftci 2020c; Fu 2020; Yayik 2018). Nine trials investigated postoperative pain during activity at 24 hours and showed a clinically relevant decrease in pain intensity in the ESPB group (MD ‐1.31 points, 95% CI‐1.69 to ‐0.93; 9 trials, 518 participants; Analysis 1.3) (Ciftci 2020d; Ciftci 2020c; Fu 2020; Park 2021; Sobhy 2020; Yaoping 2019; Yayik 2018; Zhang 2020; Zheng 2019). Six trials investigated postoperative pain at rest after 48 hours, but the difference was not clinically relevant (MD ‐0.47 points, CI ‐0.79 to ‐0.14; 6 trials, 360 participants; Analysis 1.5) (Ciftci 2020c; Eskin 2020; Fu 2020; Yaoping 2019; Zhang 2020; Zheng 2019). Five trials investigated postoperative pain during activity at 48 hours, but the mean difference was not clinically relevant (MD ‐0.48 points, 95% CI ‐0.81 to ‐0.16; 5 trials, 280 participants; Analysis 1.7) (Ciftci 2020c; Fu 2020; Yaoping 2019; Zhang 2020; Zheng 2019).

1.4. Analysis.

Comparison 1: Erector spinae plane block vs no block, Outcome 4: Mean difference in postoperative pain intensity at rest ‐ 2 hours (VAS)

1.6. Analysis.

Comparison 1: Erector spinae plane block vs no block, Outcome 6: Mean difference in postoperative pain intensity during activity ‐ 2 hours (VAS)

1.3. Analysis.

Comparison 1: Erector spinae plane block vs no block, Outcome 3: Mean difference in postoperative pain intensity during activity (24 hours postoperatively) (VAS)

1.5. Analysis.

Comparison 1: Erector spinae plane block vs no block, Outcome 5: Mean difference in postoperative pain intensity at rest ‐ 48 hours (VAS)

1.7. Analysis.

Comparison 1: Erector spinae plane block vs no block, Outcome 7: Mean difference in postoperative pain intensity during activity ‐ 48 hours (VAS)

Rate of chronic postsurgical pain (after three months, six months, one year)

One trial reported on chronic postsurgical pain after three months (Park 2021), so meta‐analysis was not possible. However, in this one trial there was no difference in the relative risk of chronic postsurgical pain after three months between the group treated with ESPB and the group receiving no block (RR 0.80, 95% CI 0.46 to 1.40; 1 trial, 58 participants; Analysis 1.8). One trial reported on chronic postsurgical pain after six months (Park 2021), so meta‐analysis was not possible. However, again there was no difference in the relative risk of chronic postsurgical pain between the group treated with ESPB and the group receiving no block in this study (RR 0.53, 95% CI 0.27 to 1.06; 1 trial, 58 participants; Analysis 1.9).

1.8. Analysis.

Comparison 1: Erector spinae plane block vs no block, Outcome 8: Rate of chronic post‐surgical pain after 3 months

1.9. Analysis.

Comparison 1: Erector spinae plane block vs no block, Outcome 9: Rate of chronic post‐surgical pain after 6 months

Cumulative oral morphine consumption at two hours, 24 hours and 48 hours postoperatively

One trial investigated cumulative oral morphine consumption at two hours postoperatively (Ciftci 2020d), 23 trials at 24 hours (Aksu 2019a; Aksu 2019b; Anushree 2020; Arora 2019; Calia 2019; Ciftci 2020d; Ciftci 2020c; Elsabeeny 2020; Eskin 2020; Gultekin 2020; Gürkan 2018; Gürkan 2020; Krishna 2019; Liu 2021; Park 2021; Seelam 2020; Sharma 2020; Singh 2019; Singh 2020; Sobhy 2020; Tulgar 2018; Yayik 2018; Zhang 2020), and four trials at 48 hours (Ciftci 2020c; Eskin 2020; Fu 2020; Zhang 2020). One trial showed decreased oral morphine consumption in patients treated with an ESPB compared to no block (MD ‐13.47 mg, 95% CI ‐15.49 to ‐11.44; 1 trial, 60 participants; Analysis 1.10). Patients treated with an ESPB showed a decrease in oral morphine consumption 24 hours after surgery compared to those receiving no block (MD ‐33.59 mg, 95% CI ‐56.37 to ‐10.81; 23 trials, 1341 participants; Analysis 1.11). Patients treated with an ESPB also showed a decrease in oral morphine consumption 48 hours after surgery compared to those receiving no block (MD ‐54.14 mg, 95% CI ‐69.09 to ‐39.20; 4 trials, 260 participants; Analysis 1.12).

1.10. Analysis.

Comparison 1: Erector spinae plane block vs no block, Outcome 10: Cumulative mean oral morphine requirement in mg ‐ 2 hours

1.11. Analysis.

Comparison 1: Erector spinae plane block vs no block, Outcome 11: Cumulative mean oral morphine requirement in mg ‐ 24 hours

1.12. Analysis.

Comparison 1: Erector spinae plane block vs no block, Outcome 12: Cumulative mean oral morphine requirement in mg ‐ 48 hours

Rates of opioid‐related adverse events

Twenty‐one trials investigated the rate of postoperative nausea and vomiting (PONV) (Abdelhamid 2020; Aksu 2019a; Aksu 2019b; Anushree 2020; Ciftci 2020d; Ciftci 2020c; Elsabeeny 2020; Eskin 2020; Fu 2020; Gürkan 2018; Gürkan 2020; Liu 2021; Park 2021; Prasad 2020; Qiang 2018; Sharma 2020; Singh 2019; Singh 2020; Sobhy 2020; Tulgar 2018; Zheng 2019). There was a reduction in the RR for PONV in the ESPB group compared to the group treated with opioids only (RR 0.46, 95% CI 0.36 to 0.60; 21 trials, 1111 participants; NNT = 2.17; Analysis 1.13). Seven trials investigated the incidence of pruritus postoperatively (Ciftci 2020d; Ciftci 2020c; Eskin 2020; Fu 2020; Liu 2021; Sobhy 2020; Zheng 2019). The meta‐analysis showed a reduction in pruritus in the ESPB group compared to the opioid group (RR 0.41, 95% CI 0.26 to 0.63; 440 participants, 7 trials; NNT = 2.44; Analysis 1.14).

1.13. Analysis.

Comparison 1: Erector spinae plane block vs no block, Outcome 13: Rates of opioid‐related adverse events (PONV)

1.14. Analysis.

Comparison 1: Erector spinae plane block vs no block, Outcome 14: Rates of opioid‐related adverse events (pruritus)

Comparison 2: Erector spinae plane block versus placebo treatment

This comparison included 13 studies with 807 participants in total comparing the analgesic effect of ESPB versus placebo treatment (Abu 2019; Ciftci 2020a; El Ghamry 2019b; Hamed 2019; Ibrahim 2019; Ibrahim 2020; Mostafa 2020; Shim 2020; Yang 2019; Yao 2020b; Yao 2020a; Yuzhong 2018a; Zhang 2021).

Primary outcomes

Mean difference in postoperative pain intensity at rest at 24 hours

Eight trials investigated postoperative pain intensity at rest at 24 hours (Abu 2019; Ciftci 2020a; El Ghamry 2019b; Hamed 2019; Mostafa 2020; Yang 2019; Yao 2020b; Zhang 2021). An ESPB has no effect on pain intensity at rest 24 hours after surgery compared to placebo block (MD ‐0.14 points, 95% CI ‐0.29 to 0.00; 8 trials, 499 participants; Analysis 2.1). The 95% prediction interval ranges from ‐1.22 to 1.07. We downgraded the certainty of evidence to moderate due to imprecision (Table 2).

2.1. Analysis.

Comparison 2: Erector spinae plane block vs placebo treatment, Outcome 1: Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS)

Sensitivity and subgroup analyses

Sensitivity analyses focusing on the influence of studies with some concerns and high risk of bias (MD ‐0.18 points, 95% CI ‐0.42 to 0.05) and those that did not declare the study flow (MD ‐0.15 points, 95% CI ‐0.37 to 0.06) did not show a difference. Similarly, using a fixed‐effect model did not show a significantly different result (MD ‐0.12 points, 95% CI ‐0.22 to ‐0.02). There is no sensitivity analysis focusing on the influence of a high level of attrition due to lack of data (Table 18). We did not carry out subgroup analysis due to low heterogeneity (I² = 20%).

Rates of block‐related adverse events

Ten trials investigated rates of block‐related adverse events (Abu 2019; Ciftci 2020a; Hamed 2019; Ibrahim 2019; Ibrahim 2020; Shim 2020; Yang 2019; Yao 2020a; Yao 2020b; Zhang 2021). There were no adverse events in either group in any of the included studies (Analysis 2.2). The 95% prediction interval ranges from 0.23 to 4.23. We downgraded the certainty of evidence to low due to imprecision (double downgrade) (Table 2).

2.2. Analysis.

Comparison 2: Erector spinae plane block vs placebo treatment, Outcome 2: Rates of block‐related adverse events

Sensitivity and subgroup analyses

Further zero event sensitivity analysis in R showed that there is probably no difference in block‐related adverse events between the groups treated with an ESPB and those receiving placebo block (RR 1.00, 95% CI 0.29 to 3.41) (Table 18). Other sensitivity analyses were not possible due to zero events. We did not perform subgroup analysis due to low heterogeneity (I² = 0%).

Secondary outcomes

Postoperative pain at rest/during activity at two hours, 24 hours and 48 hours

Seven trials investigated postoperative pain at rest after two hours (Abu 2019; Ciftci 2020a; El Ghamry 2019b; Hamed 2019; Mostafa 2020; Yang 2019; Yao 2020b). The results showed a clinically relevant reduction in pain intensity in the ESPB group (MD ‐1.46 points, 95% CI ‐2.44 to ‐0.48; 7 trials, 439 participants; Analysis 2.4). Only one trial investigated postoperative pain during activity at two hours (Yao 2020b), therefore meta‐analysis was not possible. Three trials investigated postoperative pain during activity at 24 hours, which showed no clinically relevant difference (MD ‐0.10 points, 95% CI ‐0.19 to ‐0.02; 3 trials, 199 participants; Analysis 2.3) (El Ghamry 2019b; Yao 2020b; Zhang 2021). Two trials investigated postoperative pain at rest after 48 hours and showed no clinically relevant difference (MD ‐0.49 points, 95% CI ‐0.84 to ‐0.14; 2 trials, 120 participants; Analysis 2.5) (Ciftci 2020a; Zhang 2021). Only one trial investigated postoperative pain during activity at 48 hours (Zhang 2021).

2.4. Analysis.

Comparison 2: Erector spinae plane block vs placebo treatment, Outcome 4: Mean difference in postoperative pain intensity at rest ‐ 2 hours (VAS)

2.3. Analysis.

Comparison 2: Erector spinae plane block vs placebo treatment, Outcome 3: Mean difference in postoperative pain intensity during activity (24 hours postoperatively) (VAS)

2.5. Analysis.

Comparison 2: Erector spinae plane block vs placebo treatment, Outcome 5: Mean difference in postoperative pain intensity at rest ‐ 48 hours (VAS)

Rate of chronic postsurgical pain (after three months, six months, one year)

None of the included trials reported data on this outcome.

Cumulative oral morphine consumption at two hours, 24 hours and 48 hours postoperatively

No trial investigated cumulative oral morphine consumption at two hours postoperatively, but 11 trials assessed this at 24 hours (Ciftci 2020a; El Ghamry 2019b; Hamed 2019; Ibrahim 2019; Ibrahim 2020; Mostafa 2020; Shim 2020; Yao 2020a; Yao 2020b; Yuzhong 2018a; Zhang 2021) and two trials at 48 hours (Ciftci 2020a; Yao 2020a; Zhang 2021). Patients treated with an ESPB required less morphine compared to those treated with a placebo block 24 hours after surgery (MD ‐21.05 mg, 95% CI ‐33.17 to ‐8.93; 11 trials, 677 participants; Analysis 2.6). The same holds true for 48 hours after surgery (MD ‐26.62 mg, 95% CI ‐46.41 to ‐6.84; 3 trials, 195 participants; Analysis 2.7).

2.6. Analysis.

Comparison 2: Erector spinae plane block vs placebo treatment, Outcome 6: Cumulative mean oral morphine requirement in mg ‐ 24 hours

2.7. Analysis.

Comparison 2: Erector spinae plane block vs placebo treatment, Outcome 7: Cumulative mean oral morphine requirement in mg ‐ 48 hours

Rates of opioid‐related adverse events

Eleven trials investigated the rate of PONV (Abu 2019; Ciftci 2020a; El Ghamry 2019b; Ibrahim 2019; Mostafa 2020; Shim 2020; Yang 2019; Yao 2020a; Yao 2020b; Yuzhong 2018a; Zhang 2021). The meta‐analysis showed a reduction in PONV in the ESPB group compared to patients treated with placebo treatment (RR 0.39, 95% CI 0.28 to 0.56; 11 trials, 695 participants; NNT = 2.56; Analysis 2.8). Two trials investigated the incidence of pruritus postoperatively (Ciftci 2020a; Ibrahim 2019). In contrast, a reduction in pruritus in the ESPB group compared to patients treated with placebo block was not found (RR 0.51, 95% CI 0.21 to 1.28; 2 trials, 110 participants; Analysis 2.9).

2.8. Analysis.

Comparison 2: Erector spinae plane block vs placebo treatment, Outcome 8: Rates of opioid‐related adverse events (PONV)

2.9. Analysis.

Comparison 2: Erector spinae plane block vs placebo treatment, Outcome 9: Rates of opioid‐related adverse events (pruritus)

Comparison 3: Erector spinae plane block versus paravertebral block (PVB)

This comparison included 11 studies with 859 participants in total comparing the analgesic effect of ESPB against PVB (Chen 2020; Ciftci 2020c; El Ghamry 2019a; Fang 2019; Guo 2019; Gürkan 2020; Swisher 2020; Taketa 2019; Turhan 2020; Yuzhong 2018b; Zhao 2020).

Mean difference in postoperative pain intensity at rest at 24 hours

Seven trials investigated postoperative pain intensity at rest at 24 hours (Ciftci 2020c; El Ghamry 2019a; Fang 2019; Guo 2019; Taketa 2019; Turhan 2020; Zhao 2020). An ESPB may not have an effect on postoperative pain intensity at rest 24 hours after surgery compared to a PVB (MD 0.23 points, 95% CI ‐0.06 to 0.52; 7 trials, 478 participants; Analysis 3.1). The 95% prediction interval ranges from ‐0.80 to 1.27. We downgraded the certainty of evidence to low due to inconsistency and imprecision (Table 3).

3.1. Analysis.

Comparison 3: Erector spinae plane block vs paravertebral block, Outcome 1: Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS)

Sensitivity and subgroup analyses

Sensitivity analyses focusing on the influence of studies with some concerns and high risk of bias (MD 0.28 points, 95% CI ‐0.06 to 0.62) or studies that did not declare the study flow (MD 0.14 points, 95% CI ‐0.11 to 0.40) did not show an influence. Similarly, using a fixed‐effect meta‐analysis did not show a different result (MD ‐0.05 points, 95% CI ‐0.11 to 0.00) (Table 18). There is no sensitivity analysis focusing on the influence of a high level of attrition due to lack of data. Subgroup analysis due to high heterogeneity (I² = 84%) was not possible because of limited data (fewer than 10 trials).

Rates of block‐related adverse events

Seven trials investigated rates of block‐related adverse events (Chen 2020; Ciftci 2020c; El Ghamry 2019a; Fang 2019; Swisher 2020; Taketa 2019; Turhan 2020). The meta‐analysis showed a clinically relevant reduction in the RR of block‐related adverse events in the ESPB group compared with the PVB group (RR 0.10, 95% CI 0.02 to 0.55; 7 trials, 522 participants, NNT = 10; Analysis 3.2). The 95% prediction interval was 0.05 to 1.41. We downgraded the certainty of evidence to moderate due to risk of bias (Table 3).

3.2. Analysis.

Comparison 3: Erector spinae plane block vs paravertebral block, Outcome 2: Rates of block‐related adverse events

Sensitivity and subgroup analyses

Sensitivity analyses focusing on the influence of studies with some concerns and high risk of bias (RR 0.11, 95% CI 0.01 to 1.99) showed a different result, whereas studies that did not declare the study flow (RR 0.1, 95% CI 0.02 to 0.55) did not show an influence. Similarly, using a fixed‐effect meta‐analysis did not show a different result (RR 0.1 points, 95% CI 0.02 to 0.55) (Table 18). There is no sensitivity analysis focusing on the influence of a high level of attrition due to lack of data. Due to many zero events in both the ESPB and PVB groups, we performed further sensitivity analysis in R. This also showed that there may be a clinically relevant reduction in block‐related adverse events between the groups treated with an ESPB and those receiving a PVB (RR 0.27, 95% CI 0.08 to 0.95; 7 trials, 522 participants; NNT = 3.70). Subgroup analysis was not performed due to low heterogeneity (I² = 0%).

Secondary outcomes

Postoperative pain at rest/during activity at two hours, 24 hours and 48 hours

Three trials investigated postoperative pain at rest after two hours (Ciftci 2020c; El Ghamry 2019a; Taketa 2019). The results showed no difference in pain intensity at rest two hours after surgery in the ESPB group compared to those receiving a PVB (MD 0.26 points, 95% CI ‐0.24 to 0.76; 3 trials, 211 participants; Analysis 3.4). Two trials investigated postoperative pain intensity during activity at two hours (Ciftci 2020c; Taketa 2019). The results did not show a difference between the ESPB and PVB groups (MD 0.23 points, 95% CI ‐1.14 to 1.60; 2 trials, 141 participants; Analysis 3.6). Six trials investigated postoperative pain intensity during activity at 24 hours (Ciftci 2020c; Fang 2019; Guo 2019; Taketa 2019; Turhan 2020; Zhao 2020). The results did not show a difference between the ESPB and PVB groups (MD 0.24 points, 95% CI ‐0.06 to 0.53; 6 trials, 408 participants; Analysis 3.3). Four trials investigated postoperative pain intensity at rest after 48 hours (Ciftci 2020c; Fang 2019; Taketa 2019; Zhao 2020). The results did not show a difference between the ESPB and PVB groups (MD ‐0.01 points, 95% CI ‐0.14 to 0.13; 4 trials, 298 participants; Analysis 3.5). Four trials investigated postoperative pain during activity at 48 hours (Ciftci 2020c; Fang 2019; Taketa 2019; Zhao 2020). The results did not show a difference between the ESPB and PVB groups (MD ‐0.02 points, 95% CI ‐0.31 to 0.27; 4 trials, 298 participants; Analysis 3.7).

3.4. Analysis.

Comparison 3: Erector spinae plane block vs paravertebral block, Outcome 4: Mean difference in postoperative pain intensity at rest ‐ 2 hours (VAS)

3.6. Analysis.

Comparison 3: Erector spinae plane block vs paravertebral block, Outcome 6: Mean difference in postoperative pain intensity during activity ‐ 2 hours

3.3. Analysis.

Comparison 3: Erector spinae plane block vs paravertebral block, Outcome 3: Mean difference in postoperative pain intensity during activity (24 hours postoperatively) (VAS)

3.5. Analysis.

Comparison 3: Erector spinae plane block vs paravertebral block, Outcome 5: Mean difference in postoperative pain intensity at rest ‐ 48 hours

3.7. Analysis.

Comparison 3: Erector spinae plane block vs paravertebral block, Outcome 7: Mean difference in postoperative pain intensity during activity ‐ 48 hours (VAS)

Rate of chronic postsurgical pain (after three months, six months, one year)

None of the included trials reported data on this outcome.

Cumulative oral morphine consumption at two hours, 24 hours and 48 hours postoperatively

One trial investigated cumulative oral morphine consumption at two hours postoperatively (Yuzhong 2018b), therefore no meta‐analysis was performed. Eight trials investigated cumulative oral morphine consumption at 24 hours (Chen 2020; Ciftci 2020c; El Ghamry 2019a; Guo 2019; Gürkan 2020; Turhan 2020; Yuzhong 2018b; Zhao 2020). The meta‐analysis showed no difference in cumulative oral morphine consumption at 24 hours after surgery between the ESPB and PVB treatment groups (MD 6.94 mg, 95% CI ‐2.65 to 16.53; 8 trials, 496 participants; Analysis 3.9). Five trials investigated cumulative oral morphine consumption at 48 hours (Chen 2020; Ciftci 2020c; Turhan 2020; Yuzhong 2018b; Zhao 2020). The meta‐analysis did not show a difference in cumulative oral morphine consumption at 48 hours after surgery between the ESPB and PVB treatment groups (MD 9.57 mg, 95% CI ‐4.33 to 23.47; 5 trials, 336 participants; Analysis 3.10).

3.9. Analysis.

Comparison 3: Erector spinae plane block vs paravertebral block, Outcome 9: Cumulative mean oral morphine requirement in mg ‐ 24 hours

3.10. Analysis.

Comparison 3: Erector spinae plane block vs paravertebral block, Outcome 10: Cumulative mean oral morphine requirement in mg ‐ 48 hours

Rates of opioid‐related adverse events

Ten trials investigated the rate of PONV (Chen 2020; Ciftci 2020c; El Ghamry 2019a; Fang 2019; Guo 2019; Gürkan 2020; Taketa 2019; Turhan 2020; Yuzhong 2018b; Zhao 2020). The meta‐analysis results did not show a difference in PONV between the ESPB and PVB treated groups (RR 1.13, 95% CI 0.81 to 1.58; 10 trials, 668 participants; NNT = 0.88; Analysis 3.11). Only one trial investigated the incidence of pruritus postoperatively (Ciftci 2020c), so that no meta‐analysis was possible.

3.11. Analysis.

Comparison 3: Erector spinae plane block vs paravertebral block, Outcome 11: Rates of opioid‐related adverse events (PONV)

Comparison 4: Erector spinae plane block versus transversus abdominis plane block (TAPB)

This comparison included four studies with 224 participants in total comparing the analgesic effect of ESPB against TAPB blocks (Abdelhamid 2020; Altıparmak 2019a; Kamel 2020; Ibrahim 2020).

Primary outcomes

Mean difference in postoperative pain intensity at rest at 24 hours

Three trials investigated postoperative pain intensity at rest at 24 hours (Abdelhamid 2020; Altıparmak 2019a; Kamel 2020). An ESPB may have no effect on pain intensity at rest 24 hours after surgery compared to a TAPB (MD ‐0.16 points, 95% CI‐0.46 to 0.14; 3 trials, 160 participants; Analysis 4.1). No 95% prediction interval was calculated due to limited data. We downgraded the certainty of evidence to low due to imprecision and inconsistency.

4.1. Analysis.

Comparison 4: Erector spinae plane block vs transversus abdominis plane block, Outcome 1: Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS)

Sensitivity and subgroup analyses

Sensitivity analyses focusing on the influence of studies with risk of bias did not show a difference (MD ‐0.25, 95% CI ‐0.63 to 0.13). Similarly, using a fixed‐effect meta‐analysis did not show a significantly different result (MD ‐0.21, CI ‐0.36 to ‐0.05). We did not perform sensitivity analysis focusing on study flow because all studies reported this (Table 18). There is no sensitivity analysis focusing on the influence of a high level of attrition due to lack of data. The subgroup analysis due to high heterogeneity (I² = 64%) was not possible because of lack of data (fewer than 10 trials).

Rates of block‐related adverse events

Four trials investigated rates of block‐related adverse events (Abdelhamid 2020; Altıparmak 2019a; Ibrahim 2020; Kamel 2020). There may be no difference in block‐related adverse events between ESPB and TAPB (RR 1.00, 95% CI 0.15 to 6.84; 4 trials, 202 participants; Analysis 4.2). The 95% prediction interval was 0.03 to 31.68. We downgraded the certainty of evidence to low due to imprecision (double downgrade).

4.2. Analysis.

Comparison 4: Erector spinae plane block vs transversus abdominis plane block, Outcome 2: Rates of block‐related adverse events

Sensitivity and subgroup analyses

Due to zero events in both groups, we performed further sensitivity analysis in R. This showed that there is probably no difference in block‐related adverse events between the groups treated with an ESPB and TAPB (RR 1.00, 95% CI 0.21 to 4.83; 3 trials, 202 participants). Sensitivity analysis focusing on the influence of studies with risk of bias or some concerns did not show a significant difference (RR 1.00, 95% CI 0.15 to 6.84). We did not perform sensitivity analysis focusing on study flow because all studies reported this. Similarly, using a fixed‐effect meta‐analysis did not show a significantly different result (RR 1.00, 95% CI 0.07 to 15.00). We did not perform sensitivity analysis focusing on attrition due to lack of data. We did not perform subgroup analysis due to low heterogeneity (I² = 0%).

Secondary outcomes

Postoperative pain at rest/during activity at two hours, 24 hours and 48 hours

Three trials investigated postoperative pain at rest after two hours (Abdelhamid 2020; Altıparmak 2019a; Kamel 2020). Results showed no difference in pain intensity at rest after two hours (MD ‐0.33 points, 95% CI ‐1.74 to 1.08; 3 trials, 160 participants; Analysis 4.3). No trial investigated postoperative pain at rest after 48 hours or postoperative pain during activity at two hours, 24 hours or at 48 hours.

4.3. Analysis.

Comparison 4: Erector spinae plane block vs transversus abdominis plane block, Outcome 3: Mean difference in postoperative pain intensity at rest ‐ 2 hours (VAS)

Rate of chronic postsurgical pain (after three months, six months, one year)

None of the included trials reported data focusing on this outcome.

Cumulative oral morphine consumption at two hours, 24 hours and 48 hours postoperatively

Three trials investigated cumulative oral morphine consumption at 24 hours postoperatively (Altıparmak 2019a; Ibrahim 2020; Kamel 2020). The results showed no difference in oral morphine consumption at 24 hours in the group treated with TAPB compared to the group treated with ESPB (MD ‐17.31, 95% CI ‐41.42 to 6.79; 3 trials, 158 participants; Analysis 4.4). No trial investigated cumulative oral morphine consumption at two hours or at 48 hours postoperatively.

4.4. Analysis.

Comparison 4: Erector spinae plane block vs transversus abdominis plane block, Outcome 4: Cumulative mean oral morphine requirement in mg ‐ 24 hours

Rates of opioid‐related adverse events

Three trials investigated the rate of PONV (Abdelhamid 2020; Altıparmak 2019a; Kamel 2020). The meta‐analysis showed no difference in the RR for PONV in the ESPB group compared with the TAPB group (RR 0.71, 95% CI 0.44 to 1.16; 3 trials, 160 participants; Analysis 4.5). No trials investigated pruritus as a postoperative outcome.

4.5. Analysis.

Comparison 4: Erector spinae plane block vs transversus abdominis plane block, Outcome 5: Rates of opioid‐related adverse events (PONV)

Comparison 5: Erector spinae plane block versus serratus anterior plane block (SAPB)

This comparison included four studies with 255 participants in total comparing the analgesic effect of ESPB against SAPB block (Elsabeeny 2020; Ekinci 2020; Finnerty 2020; Gaballah 2019).

Primary outcomes

Mean difference in postoperative pain intensity at rest at 24 hours

No studies reported on this outcome.

Rates of block‐related adverse events

Two trials investigated rates of block‐related adverse events (Elsabeeny 2020; Finnerty 2020). There were no adverse events in either group in any of the studies included (Analysis 5.1). No 95% prediction interval was calculated due to limited data. We downgraded the certainty of evidence to low due to imprecision (double downgrade).

5.1. Analysis.

Comparison 5: Erector spinae plane block vs serratus anterior plane block, Outcome 1: Rates of block‐related adverse events

Sensitivity and subgroup analyses

Further sensitivity analysis in R focusing on zero events showed that there is probably no difference in block‐related adverse events between the groups treated with an ESPB and SAPB (RR 1.00, 95% CI 0.06 to 15.59). Due to zero events we did not perform other sensitivity analyses. We did not perform subgroup analysis due to low heterogeneity (I² = 0%) and insufficient data.

Secondary outcomes

Postoperative pain at rest/during activity at two hours, 24 hours and 48 hours

Only one trial investigated postoperative pain at rest after two hours (Gaballah 2019), therefore no meta‐analysis could be performed (Analysis 5.2). No trials investigated postoperative pain during activity at two hours, 24 hours or 48 hours, or postoperative pain at rest after 24 hours or 48 hours.

5.2. Analysis.

Comparison 5: Erector spinae plane block vs serratus anterior plane block, Outcome 2: Mean difference in postoperative pain intensity at rest ‐ 2 hours (VAS)

Rate of chronic postsurgical pain (after three months, six months, one year)

None of the included trials reported data focusing on this outcome.

Cumulative oral morphine consumption at two hours, 24 hours and 48 hours postoperatively

Two trials investigated cumulative oral morphine consumption at 24 hours postoperatively (Elsabeeny 2020; Finnerty 2020), and none at two hours or 48 hours. The results showed no difference in oral morphine consumption (MD ‐3.36 mg, 95% CI ‐17.50 to 10.78; 2 trials, 110 participants; Analysis 5.3).

5.3. Analysis.

Comparison 5: Erector spinae plane block vs serratus anterior plane block, Outcome 3: Cumulative mean oral morphine requirement in mg ‐ 24 hours

Rates of opioid‐related adverse events

Three trials investigated the rate of PONV and showed that ESPB has no effect on PONV compared with SAPB treatment (RR 0.70, 95% CI 0.29 to 1.71; 3 trials, 170 participants; Analysis 5.4) (Ekinci 2020; Elsabeeny 2020; Finnerty 2020). No trial investigated the incidence of pruritus postoperatively.

5.4. Analysis.

Comparison 5: Erector spinae plane block vs serratus anterior plane block, Outcome 4: Rates of opioid‐related adverse events (PONV)

Comparison 6: Erector spinae plane block versus pectoralis plane block (PECSB)

This comparison included four studies with 212 participants that investigated the analgesic effect of ESPB against PECSB (Altıparmak 2019b; Gad 2019; Khorasanizadeh 2020; Sinha 2019).

Primary outcomes

Mean difference in postoperative pain intensity at rest at 24 hours

Two trials investigated postoperative pain intensity at rest at 24 hours (Altıparmak 2019b; Sinha 2019). There may be no difference in postoperative pain intensity at rest 24 hours after surgery between the groups treated with an ESPB and those receiving PECSB (MD 0.24 points, 95% CI ‐0.11 to 0.58; 2 trials, 98 participants; Analysis 6.1). No 95% prediction interval was calculated due to limited data. We downgraded the certainty of evidence to low due to inconsistency and imprecision (Table 6).

6.1. Analysis.

Comparison 6: Erector spinae plane block vs pectoralis plane block, Outcome 1: Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS)

Sensitivity and subgroup analyses

We did not perform sensitivity analyses focusing on the influence of studies with risk of bias and some concerns because both trials were low‐risk. Using a fixed‐effect model the result was not different (MD 0.27 points, 95% CI 0.03 to 0.52). We did not perform sensitivity analysis focusing on study flow due to all studies reporting this. We did not perform sensitivity analysis focusing on attrition bias due to limited data. We did not perform subgroup analysis due to low heterogeneity (I² = 41%).

Rates of block‐related adverse events

Only one trial investigated rates of block‐related adverse events in this comparison (Gad 2019), so meta‐analysis was not possible (Analysis 6.2).

6.2. Analysis.

Comparison 6: Erector spinae plane block vs pectoralis plane block, Outcome 2: Rates of block‐related adverse outcomes

Secondary outcomes

Postoperative pain at rest/during activity at two hours, 24 hours and 48 hours

Three trials investigated postoperative pain at rest after two hours (Altıparmak 2019b; Khorasanizadeh 2020; Sinha 2019). The results showed no difference in pain intensity at rest between the groups (MD 0.21 points, 95% CI ‐0.16 to 0.58; 3 trials, 162 participants; Analysis 6.3). No trials investigated postoperative pain during activity at two hours, 24 hours or 48 hours, or postoperative pain at rest after 48 hours.

6.3. Analysis.

Comparison 6: Erector spinae plane block vs pectoralis plane block, Outcome 3: Mean difference in postoperative pain intensity at rest ‐ 2 hours (VAS)

Rate of chronic postsurgical pain (after three months, six months, one year)

None of the included trials reported data focusing on this outcome.

Cumulative oral morphine consumption at two hours, 24 hours and 48 hours postoperatively

Three trials investigated cumulative oral morphine consumption at 24 hours postoperatively (Altıparmak 2019b; Gad 2019; Sinha 2019). The results showed a significant increase in morphine consumption in the ESPB group compared to the group treated with PECSB (MD 14.08 mg, 95% CI 4.21 to 23.94; 3 trials, 145 participants; Analysis 6.4). No trials investigated cumulative oral morphine consumption at two hours or 48 hours postoperatively.

6.4. Analysis.

Comparison 6: Erector spinae plane block vs pectoralis plane block, Outcome 4: Cumulative mean oral morphine requirement in mg ‐ 24 hours

Rates of opioid‐related adverse events

No trials investigated these outcomes.

Comparison 7: Erector spinae plane block versus quadratus lumborum block (QLB)

This comparison included two studies with 140 participants in total comparing the analgesic effect of ESPB against QLB (Aygun 2020; Tulgar 2018).

Primary outcomes

Mean difference in postoperative pain intensity at rest at 24 hours

Only one study reported this outcome (Tulgar 2018), so no meta‐analysis was possible (Analysis 7.1) (Table 7).

7.1. Analysis.

Comparison 7: Erector spinae plane block vs quadratus lumborum block, Outcome 1: Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS)

Rates of block‐related adverse events

Only one trial investigated rates of block‐related adverse events in this comparison (Tulgar 2018), so meta‐analysis was not possible. In this study, there were no block‐related adverse events in either of the groups (Analysis 7.2) (Table 7).

7.2. Analysis.

Comparison 7: Erector spinae plane block vs quadratus lumborum block, Outcome 2: Rates of block‐related adverse events

Secondary outcomes

Postoperative pain at rest/during activity at two hours, 24 hours and 48 hours

One trial reported pain during activity after 24 hours (Aygun 2020), therefore no meta‐analysis was possible (Analysis 7.3). No trials investigated postoperative pain during activity at two hours and 48 hours, or postoperative pain at rest after two hours and 48 hours.

7.3. Analysis.

Comparison 7: Erector spinae plane block vs quadratus lumborum block, Outcome 3: Mean difference in postoperative pain intensity during activity (24 hours postoperatively)

Rate of chronic postsurgical pain (after three months, six months, one year)

None of the included trials reported data on this outcome.

Cumulative oral morphine consumption at two hours, 24 hours and 48 hours postoperatively

Two trials investigated oral morphine consumption at 24 hours postoperatively (Aygun 2020; Tulgar 2018), and none at two hours or 48 hours. The results did not show a difference in cumulative oral morphine consumption after 24 hours in the ESPB and QLB treated groups (MD ‐0.31 mg, 95% CI ‐2.22 to 1.61; 2 trials, 120 participants; Analysis 7.4).

7.4. Analysis.

Comparison 7: Erector spinae plane block vs quadratus lumborum block, Outcome 4: Cumulative mean oral morphine requirement in mg ‐ 24 hours

Rates of opioid‐related adverse events

Two trials investigated the rate of PONV (Aygun 2020; Tulgar 2018). The meta‐analysis results did not show a difference in PONV between the two groups (RR 1.00, 95% CI 0.35 to 2.84; 2 trials, 120 participants; Analysis 7.5). No trial investigated the incidence of pruritus postoperatively.

7.5. Analysis.

Comparison 7: Erector spinae plane block vs quadratus lumborum block, Outcome 5: Rates of opioid‐related adverse events (PONV)

Comparison 8: Erector spinae plane block versus intercostal nerve block (ICNB)

This comparison included three studies with 241 participants in total comparing the analgesic effect of ESPB against ICNB (Chen 2020; Fiorelli 2020; Turhan 2020).

Primary outcomes

Mean difference in postoperative pain intensity at rest at 24 hours

Two trials investigated postoperative pain intensity at rest at 24 hours (Fiorelli 2020; Turhan 2020). We are uncertain whether there is a difference in postoperative pain intensity at rest 24 hours after surgery between the groups treated with an ESPB and those receiving ICNB (MD ‐0.33 points, 95% CI ‐3.02 to 2.35; 2 trials, 131 participants; Analysis 8.1). No 95% prediction interval was calculated due to limited data. We downgraded the certainty of evidence to very low due to inconsistency, imprecision (double downgrade) and risk of bias (Table 8).

8.1. Analysis.

Comparison 8: Erector spinae plane block vs intercostal nerve block, Outcome 1: Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS)

Sensitivity and subgroup analyses

Sensitivity analysis focusing on studies with risk of bias and some concerns showed a significantly different result (MD ‐1.74 points, 95% CI ‐2.75 to ‐0.73). Sensitivity analyses using a fixed‐effect model did not show a significantly different result (MD 0.49 points, 95% CI 0.06 to 0.93). We did not perform sensitivity analysis focusing on studies that did not report study flow and focusing on attrition bias due to insufficient data (Table 18). We did not perform subgroup analysis due to high heterogeneity (I² = 96%) and lack of data.

Rates of block‐related adverse events

Three trials investigated rates of block‐related adverse events (Chen 2020; Fiorelli 2020; Turhan 2020). There may be no difference in block‐related adverse events (RR 0.09, 95% CI 0.01 to 1.56; 3 trials, 181 participants; Analysis 8.2). The 95% prediction interval was not calculated due to limited data. We downgraded the certainty of evidence to very low due to risk of bias and imprecision (double downgrade).

8.2. Analysis.

Comparison 8: Erector spinae plane block vs intercostal nerve block, Outcome 2: Rates of block‐related adverse events

Sensitivity and subgroup analyses

Due to zero events, we performed further sensitivity analysis in R. This also showed no difference in the RR for block‐related adverse events in the ESPB group (RR 0.32, 95% CI 0.04 to 2.28; 181 participants; 3 trials). Sensitivity analysis focusing on studies that did not report study flow showed no difference in the results (RR 0.09, 95% CI 0.01 to 1.56). Sensitivity analysis using a fixed‐effect meta‐analysis did not show a significantly different result (RR 0.09, 95% CI 0.01 to 1.56). We did not perform sensitivity analysis focusing on risk of bias because there was only one study at low risk of bias, so no meta‐analysis was possible. We did not perform subgroup analysis due to insufficient data.

Secondary outcomes

Postoperative pain at rest/during activity at two hours, 24 hours and 48 hours

No trials investigated postoperative pain at rest or during activity after two hours. Two trials investigated postoperative pain intensity during activity at 24 hours (Fiorelli 2020; Turhan 2020). Meta‐analysis showed no difference in pain intensity in the ESPB group compared to the ICNB group (MD ‐1.52 points, 95% CI ‐4.61 to 1.56; 2 trials, 131 participants; Analysis 8.3). Only one trial investigated postoperative pain at rest and during activity after 48 hours (Fiorelli 2020), therefore no meta‐analysis was possible.

8.3. Analysis.

Comparison 8: Erector spinae plane block vs intercostal nerve block, Outcome 3: Mean difference in postoperative pain intensity during activity (24 hours postoperatively) (VAS)

Rate of chronic postsurgical pain (after three months, six months, one year)

None of the included trials reported data on this outcome.

Cumulative oral morphine consumption at two hours, 24 hours and 48 hours postoperatively

No trial investigated cumulative oral morphine consumption at two hours postoperatively. Two trials investigated cumulative oral morphine consumption at 24 hours (Chen 2020; Turhan 2020). The results showed higher oral morphine consumption in the ESPB group compared to the ICNB group (MD 9.02 mg, 95% CI 3.05 to 14.99; 2 trials, 121 participants; Analysis 8.6). Two trials investigated cumulative oral morphine consumption at 48 hours postoperatively (Chen 2020; Turhan 2020). The results again showed higher oral morphine consumption in the ESPB group (MD 7.98 mg, 95% CI 2.60 to 13.36; 2 trials, 121 participants; Analysis 8.7).

8.6. Analysis.

Comparison 8: Erector spinae plane block vs intercostal nerve block, Outcome 6: Cumulative mean oral morphine requirement in mg ‐ 24 hours

8.7. Analysis.

Comparison 8: Erector spinae plane block vs intercostal nerve block, Outcome 7: Cumulative mean oral morphine requirement in mg ‐ 48 hours

Rates of opioid‐related adverse events

Two trials investigated the rate of PONV (Chen 2020; Turhan 2020). The meta‐analysis results showed no difference in the RR for PONV (RR 1.26, 95% CI 0.40 to 3.98; 2 trials, 121 participants; Analysis 8.8). No trial investigated the incidence of pruritus postoperatively.

8.8. Analysis.

Comparison 8: Erector spinae plane block vs intercostal nerve block, Outcome 8: Rates of opioid‐related adverse events (PONV)

Comparison 9: Erector spinae plane block versus epidural analgesia (EA)

This comparison included two studies with 81 participants in total comparing the analgesic effect of ESPB against epidural analgesia (Nagaraja 2018; Sakae 2020).

Primary outcomes

Mean difference in postoperative pain intensity at rest at 24 hours

Two trials investigated postoperative pain intensity at rest at 24 hours (Nagaraja 2018; Sakae 2020). We are uncertain whether there is a difference in postoperative pain intensity at rest 24 hours after surgery (MD 1.20 points, 95% CI ‐2.52 to 4.93; 2 trials, 81 participants; Analysis 9.1). No 95% prediction interval was calculated due to limited data. We downgraded the certainty of evidence to very low due to inconsistency, risk of bias and imprecision (double downgrade) (Table 9).

9.1. Analysis.

Comparison 9: Erector spinae plane block vs epidural analgesia, Outcome 1: Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS)

Sensitivity and subgroup analyses

Sensitivity analyses using a fixed‐effect meta‐analysis did not show a significantly different result (MD ‐0.30, 95% CI ‐0.70 to 0.11). We did not perform sensitivity analyses focusing on study flow and risk of bias, because all studies reported study flow and the comparison included only two trials ‐ one was a high‐risk trial (Nagaraja 2018). We did not perform subgroup analysis due to the low number of trials (fewer than 10).

Rates of block‐related adverse events

Only one trial investigated rates of block‐related adverse events in this comparison (Nagaraja 2018), so a meta‐analysis was not possible. In this trial there were no adverse events related to ESPB or EA in either group (Table 9).

Secondary outcomes

Postoperative pain at rest/during activity at two hours, 24 hours and 48 hours

Only one trial investigated postoperative pain at rest after two hours (Sakae 2020), therefore no meta‐analysis was possible. No trials investigated postoperative pain during activity at two hours. Only one trial investigated postoperative pain during activity at 24 hours, after 48 hours and at rest after 48 hours (Nagaraja 2018), therefore no meta‐analysis was possible.

Rate of chronic postsurgical pain (after three months, six months, one year)

None of the included trials reported data focusing on this outcome.

Cumulative oral morphine consumption at two hours, 24 hours and 48 hours postoperatively

Only one trial investigated the cumulative morphine consumption 24 hours after surgery (Sakae 2020), so a meta‐analysis was not possible for this outcome.

Rates of opioid‐related adverse events

No trial investigated the rate of postoperative nausea and vomiting (PONV). Only one trial investigated the incidence of pruritus postoperatively (Sakae 2020), therefore no meta‐analysis was possible.

Narrative description of study results not included in meta‐analysis

Study results that could not be analysed within meta‐analyses are reported in Table 12 and Table 17. One trial investigated ESPB compared to modified‐thoracolumbar interfascial plane block (mTLIP) and showed no difference in any of the investigated outcomes (Ciftci 2020b). Mid‐transverse process to pleura block (MTPB) was also compared to ESPB and showed a decrease in morphine consumption in the ESPB group at 24 and 48 hours, but no difference in pain intensity between the groups (Eskin 2020). Pruritus and PONV were also reduced in the ESPB group (Eskin 2020). The comparison of intrathecal morphine (ITM) and ESPB showed increased pain intensities and morphine consumption in the ESPB group, but the risk for PONV and pruritus in the ITM group was increased (Kang 2019). Tumescent analgesia was also compared to ESPB and results showed decreased pain intensities two and 24 hours after surgery, decreased morphine consumption at 24 hours and decreased PONV rates in the ESPB group (Oksuz 2019). The comparison of retrolaminar block versus ESPB showed no difference in pain intensity after 24 hours, but a reduced risk for PONV in the ESPB group (Sotome 2021). One trial comparing ESPB versus wound infiltration showed a reduction in morphine consumption and the risk for PONV in the ESPB group 24 hours after surgery, but pain intensity was not investigated (Wang 2019b). All other opioid‐related side effects, which could not be analysed within meta‐analyses, are summarised narratively in Table 17.

Discussion

This meta‐analysis includes 64 RCTs (3973 participants), published between 2018 and 2021. It investigates the effects of ESPB versus no block, placebo block and other regional anaesthetic techniques in adults undergoing surgery under general anaesthesia. Outcomes assessed in this meta‐analysis were postoperative pain intensity at rest and during activity at different time points (measured on a visual analogue scale from 0 to 10 points), adverse effects of regional anaesthesia and of opioids, cumulative oral morphine requirements (in mg oral morphine equivalents) at different time points and risk for chronic postsurgical pain.

In the following paragraphs we will discuss the results of each comparison, the risk of bias assessment, the certainty of the evidence and the limitations of this meta‐analysis.

Summary of main results

This review includes 64 RCTs with a total of 3973 patients investigating ESPB compared to no block, sham block and other regional anaesthetic techniques. It is important to note again that all groups also received standard care, including opioids. The main findings of this review are summarised in Table 1 (ESPB compared to no block); Table 2 (ESPB compared to placebo treatment); Table 3 (ESPB compared to PVB); Table 4 (ESPB compared to TAPB), Table 5 (ESPB compared to SAPB); Table 6 (ESPB compared to PECSB); Table 7 (ESPB compared to QLB); Table 8 (ESPB compared to ICNB); and Table 9 (ESPB compared to EA).

Patients treated with an ESPB compared to no block probably do not report clinically relevant reduced pain intensity at rest 24 hours after surgery. However, subgroup analysis showed that there was clinically relevant reduced pain intensity at rest following thoracic and abdominal surgery. There may be no difference in block‐related adverse events. However, there was a clinically relevant reduced pain intensity during activity 24 hours after surgery. Furthermore, clinically relevant reduced pain intensities at rest and during activity at two hours after surgery were evident in the ESPB group compared to no block. Chronic postsurgical pain at three and six months was only reported in one trial, which showed no difference. Cumulative oral morphine consumption was decreased in the ESPB group after 24 and 48 hours, which was associated with a reduced risk for PONV and pruritus. No meta‐analysis could be performed for the outcome of cumulative oral morphine consumption after two hours because only one trial was included.

Compared to placebo blocks, ESPB probably has no clinically relevant effect on postoperative pain intensity at rest 24 hours after surgery and there may be no difference in block‐related adverse events. Only at two hours after surgery is there clinically relevant reduced pain intensity at rest in the ESPB group. At the other time points differences in pain intensities were clinically not relevant or absent. Meta‐analyses were not possible for the outcomes of postoperative pain during activity at two hours and 48 hours, because only one trial was included. Chronic postsurgical pain was not reported. In contrast, cumulative oral morphine consumption was decreased in the ESPB group at 24 hours and 48 hours compared to placebo treatment, which was associated with a reduced risk for PONV. No meta‐analysis could be performed for the outcome of cumulative oral morphine consumption after two hours because no trial was included.

Compared to PVB, ESPB may not have an effect on postoperative pain intensity at rest after 24 hours, but there is probably a reduced risk for block‐related adverse events compared to PVB. Furthermore, there was no difference in pain intensity at rest after two and 48 hours, as well as no difference in postoperative pain during activity after two, 24 and 48 hours. Chronic postsurgical pain was not reported. Cumulative oral morphine consumption was not different between the EPSB and PVB groups at 24 or 48 hours and meta‐analysis was not possible for the outcome after two hours. There was also no difference in PONV rates between the ESPB and PVB groups and no meta‐analysis could be performed for opioid‐related itching, because only one study was included.

Compared to TAPB, ESPB may have no effect on postoperative pain intensity at rest after 24 hours and there may be no difference in block‐related adverse events. There was no difference in pain intensity at rest after two hours. For the outcomes of postoperative pain intensity at rest after 48 hours as well as postoperative pain during activity at two hours, 24 hours and 48 hours, no meta‐analyses could be performed. Chronic postsurgical pain was not reported. In contrast, cumulative oral morphine consumption was not decreased in the ESPB group at 24 hours compared to TAPB treatment and the risk for PONV was not different. No meta‐analysis could be performed for cumulative oral morphine consumption at two and 48 hours, or for pruritus.

Compared to SAPB, ESPB may have no effect on block‐related adverse events. Pain intensity was only reported by one study at rest after two hours and therefore no meta‐analyses could be performed for these outcomes. Chronic postsurgical pain was not reported. Cumulative oral morphine consumption was not significantly different between the EPSB and SAPB groups at 24 hours and was not reported at two or 48 hours. There may be no difference in PONV rates between the ESPB and SAPB groups. No trials investigated pruritus as a postoperative outcome.

Compared to PECSB, ESPB may have no effect on postoperative pain intensity at rest after two hours or 24 hours. Meta‐analysis was not possible for the outcome of block‐related adverse events because only one study was included (Gad 2019), which reported only one adverse event in the ESPB group and none in the PECSB group. No trials investigated postoperative pain during activity at two hours, 24 hours or 48 hours, or postoperative pain at rest after 48 hours. Chronic postsurgical pain was not reported. The results for cumulative oral morphine consumption at 24 hours showed an increase in consumption in the ESPB group compared to the PECSB group. Morphine consumption at two hours or 48 hours was not investigated by any of the trials in this comparison. Similarly, opioid‐related side effects (PONV and pruritus) were not investigated.

For the comparison of QLB and ESPB treatments both primary outcomes were only reported in one study, so no meta‐analysis was performed. Similarly, no meta‐analysis was possible for pain intensity outcomes at any time point. Chronic postsurgical pain was not reported. Cumulative oral morphine consumption was not significantly different between the EPSB and QLB groups at 24 hours and was not investigated at two hours or 48 hours. The incidence of PONV was also not different. No trial investigated the incidence of pruritus postoperatively.

Compared to ICNB, ESPB may not have any additional effect on postoperative pain intensity at rest 24 hours after surgery, but this is uncertain. Furthermore, there may be no difference in block‐related adverse events, but this is uncertain. For the outcomes of postoperative pain intensity at rest or on activity after two hours or 48 hours, no meta‐analysis was performed. Chronic postsurgical pain was not reported. Cumulative oral morphine consumption was higher in the EPSB group at 24 hours and 48 hours, compared to the ICNB group, and it was not investigated at two hours. There was no difference in PONV. No trial investigated the incidence of pruritus postoperatively.

Compared to the epidural group, we are uncertain whether ESPB has an effect on postoperative pain intensity at rest 24 hours after surgery. Meta‐analysis was not possible for the outcome of block‐related adverse events because only one study was included (Nagaraja 2018), which reported only one adverse event in the ESPB group and nothing in the PECSB group. No meta‐analysis was possible for postoperative pain intensities at other time points. Chronic postsurgical pain was not reported. Meta‐analysis was also not possible for cumulative morphine consumption at any time point or for opioid‐related adverse events (PONV and itching).

Overall completeness and applicability of evidence

This review included 64 randomised controlled trials, but the sample size for each trial was rather small (range of 29 to 106 participants per study), which increases the risk of heterogeneity and limits external validity.

Quality of the evidence

We only graded the certainty of evidence for the two primary outcomes in each comparison (postoperative pain intensity at rest after 24 hours and block‐related adverse events). We used the five areas of the GRADE system to assess the certainty of the evidence (heterogeneity, indirectness, imprecision, publication bias and the risk of bias assessment tool).

An important limitation of this review is that fact that most studies had a relatively small number of participants and many comparisons had only a small number of trials for each outcome. This meant that most subgroup analyses could not be performed and therefore heterogeneity, which was often high, could not be further explored. The certainty of evidence was often downgraded because of this. Furthermore, adverse events were poorly reported and most trials were not powered to investigate this outcome.

The trials all generally addressed a similar population (similar PICO criteria), so we did not find indirectness to be a problem. We assessed imprecision using confidence intervals and sometimes found it to be a problem, and sometimes downgraded the certainty of the evidence because of this. We did not find publication bias to be a problem when assessed via a funnel plot (Figure 6), however this assessment was only possible for outcomes that included more than 10 studies.

6.

Funnel plot of primary outcome of postoperative pain at rest after 24 hours in the comparison of ESPB vs no block

We assessed risk of bias using the risk of bias tool, and we assessed the influence of studies with some or high risk of bias on the overall results with sensitivity analyses. The sensitivity analyses showed a difference only in two comparisons (versus ICNB and EA). Other sensitivity analyses focusing on fixed‐effect versus random‐effects meta‐analysis, and studies that did not report study flow, did not find differences in outcomes.

In conclusion, we graded the primary outcomes as having a moderate to very low certainty of evidence, with three outcomes graded moderate, seven outcomes graded low and three outcomes graded as very low certainty of evidence.

Potential biases in the review process

We made a decision not to use data with asymmetrical distribution in the meta‐analysis. In addition, we decided to perform subgroup analysis only when more than 10 trials were included for an outcome. We considered these aspects within the GRADE ratings by downgrading if needed.

This review had predetermined outcomes and the included trials did not always report these outcomes. For postoperative time, many trials reported pain scores at times other than those that we were investigating. This may have been a source of bias since we could not include all these results. Equally, this shows the importance of defining a group of core outcomes for trials investigating postoperative pain and wellbeing.

Agreements and disagreements with other studies or reviews

We did not identify any meta‐analysis that has compared ESPB to no block, placebo block and other regional anaesthesia techniques in one review. However, there are many meta‐analyses comparing ESPB to specific blocks, sometimes for specific surgeries. In a literature review of recent years, we found 15 meta‐analyses of randomised controlled trials comparing ESPB to other methods, of which 11 investigated comparable outcomes to this meta‐analysis (Daghmouri 2020; Fanelli 2021; Jiao 2021; Jo 2021; Koo 2021; Leong 2020; Li 2021; Liang 2021; Weng 2021; Xiong 2021; Zhang 2021).

Most of these meta‐analyses investigated ESPB in breast surgery. Several of them found that ESPB reduced pain intensity and morphine consumption compared with no block (Leong 2020; Li 2021; Zhang 2021). They also found that ESPB showed comparable efficacy to PVB but was inferior to PECSB, which is similar to our results. Another review found no difference between PVB and ESPB in breast surgery with a reduction in morphine consumption only when comparing ESPB to a control group receiving opioid analgesia alone (Weng 2021). Interestingly, a meta‐analysis by Xiong et al found that when comparing ESPB and PVB the results depended on the site of surgery (Xiong 2021). For thoracic surgery, PVB had superior effects in reducing pain intensity but in breast surgery, PVB and ESPB showed similar efficacy. Koo et al also showed that ESPB was inferior to PVB but superior to control groups in thoracic surgery (Koo 2021).

One meta‐analysis comparing ESPB to control groups in different surgeries, found that ESPB had an opioid‐sparing effect, which aligns with our results (Jiao 2021). In another comparison over different surgeries, ESPB reduced opioid consumption compared to the control group, but there was no difference compared to other regional anaesthesia groups and ESPB fared worse than PECSB (Fanelli 2021).

Liang et al found that ESPB improved pain outcomes in spinal surgery, but did not assess other outcomes (Liang 2021). Another meta‐analysis looked at trials comparing ESPB to a control group for laparoscopic cholecystectomies and found that the ESPB group reported lower pain scores (Daghmouri 2020). No adverse events were reported in this review.

Authors' conclusions

Implications for practice.

In conclusion, erector spinae plane block (ESPB) in addition to standard care leads to lower postoperative pain intensity 24 hours after surgery compared to no block, but this was only clinically relevant at two hours after surgery. This is possibly related to the fact that only single‐shot nerve blocks using long‐lasting local anaesthetics with a duration of action of six to eight hours were administered (Hussain 2020; Schnabel 2022). It could be speculated that these results would be different if catheter‐based techniques or local anaesthetic adjuvants were used (Chen 2021; El‐Boghdadly 2017; Schnabel 2022). However, the latter result regarding pain intensity may not be that robust because the comparison with placebo blocks showed that ESPB may not have an effect on postoperative pain intensity, even directly after surgery. This has already been shown for the use of pectoralis plane blocks for acute postoperative pain treatment in breast surgery (Meißner 2021). Interestingly a subgroup analysis showed a clinically relevant reduction in postoperative pain intensity at rest 24 hours after thoracic and abdominal surgery. A surgery‐related influence therefore appears to be relevant, but will need to be further explored in the future. Additionally, we observed a decrease in opioid consumption and a lower risk of opioid‐related adverse events (e.g. postoperative nausea and vomiting) in the first 24 hours after surgery, which was still present even in comparison with sham blocks, indicating that this effect is robust, as already seen in other meta‐analyses focusing on specific surgical procedures (Fanelli 2021; Hussain 2020; Jiao 2021). Furthermore, the incidence of block‐related adverse events may be very low in the ESPB group and no different in comparison to other regional anaesthetic techniques (El‐Boghdadly 2017). Therefore, studies investigating ESPB compared to placebo treatment as a control appear to be ethically plausible and are relevant to determine the 'real' role of ESPB in postoperative pain treatment. Finally, when comparing ESPB to other blocks, no clinically relevant differences in postoperative pain outcomes could be determined, but the evidence is limited.

Implications for research.

In the future, the use of perineural adjuvants and catheter techniques should be investigated in order to expand the analgesic effects of ESPB (El‐Boghdadly 2017). Additionally, ESPB may have a robust effect as an opioid‐sparing technique, because it is associated with a reduction in opioid‐related adverse events. However, only a low number of opioid‐related adverse events were reported in the included studies; this should be investigated in the future in order to get a clearer view on the potential benefits of ESPB. Chronic pain was reported by only one trial and this requires further research because this is an important patient‐related outcome, which is also relevant to healthcare providers (Glare 2019; Weinstein 2018). Similarly, we were unable to fully investigate the influence of different types of surgery on the outcomes due to lack of data, although an initial subgroup analysis revealed procedure‐specific differences. Furthermore, our review included only preoperative administration of blocks; it would be interesting to compare this with postoperative administration.

Finally, the certainty of the evidence for most comparisons of ESPB with other regional anaesthetic techniques was low or very low. Further research into which regional block works best for which type of surgery is needed in the future.

History

Protocol first published: Issue 10, 2020

Risk of bias

Risk of bias for analysis 1.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Abdelhamid 2020

Low risk of bias

Patients were randomly assigned to either one of three study groups on the day of surgery. grouping was determined through a computer‐ generated list that was kept in a sealed enve‐ lope. Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. Demographic data, patient characteristics and durations of surgeries were comparable among the three groups.

Low risk of bias

Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. No drop outs. All patients analysed in their allocated group.

Low risk of bias

66 patients. 22 each group. No drop outs. Data available for all patients

Low risk of bias

Prespecified appropriate measures of outcome. Same measurements for the same outcomes. Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged.

Low risk of bias

Based on a pilot study, sample size was cal‐ culated according to the significant differ‐ ence in the mean of vas values over 24 hours (as a primary outcome) between control group (3.21±0.48), taP group (2.23±0.41) and esP group (1.51±0.35) using anova test, with α=0.05, power of 80%, and an effect size of 0.4. therefore, a sample size of 22 patients/group would be required (g‐Power 301). Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas were judged low risk of bias.

Ciftci 2020c

Some concerns

randomising computer program. no information on how blinding was maintained. no difference in baseline demographics.

Low risk of bias

Patients were sedated with midazolam but intervention (or no intervention) happened pre‐operatively. No drop outs. All 90 patients analysed in the groups they were allocated.

Low risk of bias

Data for all patients available.

Low risk of bias

same measurement for same outcomes. outcome assessor was blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged as some concerns of risk of bias.

Ciftci 2020d

Low risk of bias

randomising computer program. Block performed 30 minutes preop. No information on how allocation was kept concealed. No demographic differences at baseline.

Low risk of bias

Block was performed 30mins preop. Patients in both groups were given midazolam but control group did not receive a block. 30 patients in each group, no drop outs.

Low risk of bias

Data were available for all patients.

Low risk of bias

same measurements for same outcomes. outcome assessors were blinded.

Low risk of bias

prospectively registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas low risk of bias.

Elsabeeny 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline Demographics. There was no difference in baseline demographics.

Some concerns

Blocks were performed after induction of general anaesthesia. There were 5 drop outs who were then removed from the analysis.

Low risk of bias

Data was available for all included patients.

Low risk of bias

Outcomes were measured by blinded assesors.

Low risk of bias

Prespecified appropriate measures of outcome. Same measurements for same outcomes.

Low risk of bias

One area was judged at some risk of bias.

Eskin 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

Opening of envelopes and intervention post induction of anaesthesia. No drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

Same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas judged low risk of bias.

Fu 2020

Some concerns

Allocation was made using randomising computer program, no information on concealement of blinding. No difference in the baseline Demographics.

Low risk of bias

Blocks performed preoperatively. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged to be at some concern for risk of bias

Gultekin 2020

Some concerns

Allocation was made using randomising computer program, no information on concealement of allocation. No difference in the baseline Demographics.

Low risk of bias

block performed after induction of anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged at some concern of risk of bias

Park 2021

Some concerns

random computer allocation, no information on concealment of allocation. no baseline difference of demographics.

Some concerns

patients and block performers could not be blinded 1 patient dropped out in each group but they were not analysed later.

Low risk of bias

data available for almost all patients.

Low risk of bias

same measurement for same outcomes.

Low risk of bias

preregistered RCT.

Some concerns

Two areas rated at some risk of bias

Prasad 2020

Some concerns

random allocation using chit and box methid. no information on concealement. no baseline difference of demographics

Some concerns

blocks performedpre‐operatively. no information on patient drop outs.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. assessors were blinded.

Low risk of bias

preregistered RCT.

Some concerns

2 areas judged at some risk of bias.

Singh 2019

Low risk of bias

random computer allocation and revleaed in anaesthetic room by statistician. no baseline difference of demographics.

Low risk of bias

pre‐op blocks while patients were awake. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

preregistered RCT.

Low risk of bias

all areas at low risk of bias.

Singh 2020

Some concerns

random computer allocation. No information onconcealement of allocation. no baseline difference of demographics.

Low risk of bias

preo‐op blocks. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

preregistered RCT.

Some concerns

1 area at some risk of bias.

Sobhy 2020

Low risk of bias

random number generator and sealed enevelope techniques used. no baseline difference of demographics.

Low risk of bias

pre‐op application of block. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. blinded outcome assessors.

Low risk of bias

preregistered RCT.

Low risk of bias

all areas at low risk of bias.

Tulgar 2018

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline demographics.

Low risk of bias

block performed under anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Yaoping 2019

Some concerns

Allocation was randomised but no information on how this was achieved or how it was concealed. No difference in the baseline Demographics

High risk of bias

Block performed pre‐op. No further information.

High risk of bias

Not enough information in the translation .

High risk of bias

Same measurement for same outcomes. No further information from translated text.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

Not enough information in the translation.

Yayik 2018

Some concerns

No information in abstract other than to say that it was a randomised trial. No difference in the baseline Demographics.

High risk of bias

Not enough information in abstract.

High risk of bias

Not enough information in abstract.

High risk of bias

same outcome measurements for same outcomes, no further information.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

Not enough information in abstract.

Zhang 2020

Low risk of bias

Allocation was made randomly by lead investigator and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

block performed pre‐op. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurements for same outcomes.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Zheng 2019

Some concerns

Allocation was made using random number sequence, no information about concealement. No difference in the baseline Demographics.

High risk of bias

blocks performed under anaesthesia. no further information in translation.

High risk of bias

No information on study flow

High risk of bias

same measurement for same outcomes. no information on blinding of assessors.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

High risk of bias

Risk of bias for analysis 1.2 Rates of block‐related adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Abdelhamid 2020

Low risk of bias

Patients were randomly assigned to either one of three study groups on the day of surgery. grouping was determined through a computer‐ generated list that was kept in a sealed enve‐ lope. Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. Demographic data, patient characteristics and durations of surgeries were comparable among the three groups.

Low risk of bias

Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. No drop outs. All patients analysed in their allocated group.

Low risk of bias

66 patients. 22 each group. No drop outs. Data available for all patients

Low risk of bias

Prespecified appropriate measures of outcome. Same measurements for the same outcomes. Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged.

Low risk of bias

Based on a pilot study, sample size was cal‐ culated according to the significant differ‐ ence in the mean of vas values over 24 hours (as a primary outcome) between control group (3.21±0.48), taP group (2.23±0.41) and esP group (1.51±0.35) using anova test, with α=0.05, power of 80%, and an effect size of 0.4. therefore, a sample size of 22 patients/group would be required (g‐Power 301). Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas were judged low risk of bias.

Aksu 2019a

Low risk of bias

Sequentially numbered opaque sealed envelope technique was used for randomization and patients were randomized to one of the two groups as ESPB group and Control group.[14] All patients in ESPB group received bi‐level ESPB and patients in Control group received no intervention.

Low risk of bias

No attrition.

Low risk of bias

Data was available for all included patients.

Low risk of bias

Same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

This study was done after obtaining local Ethic Com‐ mittee’s permission (KIA 2017‐377). Patients’ written informed consents were obtained. The study was registered prior to patient enrollment with clinical‐ trials.gov (NCT03415646) on January 2018.

Low risk of bias

All areas judged low risk of bias.

Aksu 2019b

Low risk of bias

Randomization was performed according to computer‐generated random number tables, and allocation to treatment group was done using the sealed opaque envelope technique. According to randomization, patients were divided into two gro‐ ups to receive either ultrasound‐guided erector spi‐ nae plane block (ESP group) or no intervention (Control group).

Low risk of bias

There was no attrition.

Low risk of bias

Data was available for all included patients.

Low risk of bias

Same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

This prospective randomized controlled study was done after obtaining ethics committee’s permission (KIA 2017‐349/28.11.2017) and written informed consent of the patients between February and April 2018. The study was registered prior to patient enrollment with clinicaltrials.gov (NCT03420703).

Low risk of bias

All areas judged low risk of bias.

Ciftci 2020c

Some concerns

randomising computer program. no information on how blinding was maintained. no difference in baseline demographics.

Low risk of bias

Patients were sedated with midazolam but intervention (or no intervention) happened pre‐operatively. No drop outs. All 90 patients analysed in the groups they were allocated.

Low risk of bias

Data for all patients available.

Low risk of bias

same measurement for same outcomes. outcome assessor was blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged as some concerns of risk of bias.

Elsabeeny 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline Demographics. There was no difference in baseline demographics.

Some concerns

Block was performed after induction of general anaesthesia. 5 patients dropped out. those patients were excluded from analysis.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area was judged at some risk of bias

Gürkan 2018

Low risk of bias

Randomization was achieved using the sequentially numbered opaque sealed envelope technique (SNOSE) [10]. Patients were ran‐ domized to receive either single‐shot ESP block (ESP group) or no in‐ tervention (control group).

Some concerns

It's unclear whether the parctitioners applying the blocks were blinded or not.

Low risk of bias

Data was available for all included patients.

Low risk of bias

Same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

This prospective, randomized controlled study was performed following obtaining of Kocaeli City Clinical Trials Ethical Committee approval (KAEK 2017‐347) and of written informed consent from the patients.

Low risk of bias

One area was judged to be at some risk of bias.

Krishna 2019

Low risk of bias

Allocation was random using a computer randomisation method and allocation was concealed using opaque sealed envelopes. No difference in the baseline Demographics.

Some concerns

4 patients dropped out of study but were excluded from results. patients that dropped out were not analysed. all other patients were analysed inthei allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessor nurses were not involved in the trial.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged at some risk of bias.

Liu 2021

Low risk of bias

Patients were randomised using a random number generator online. The randomized schedule was hidden and generated by the anesthesiologist in charge of the case, who was not part of the study.

Low risk of bias

blocks performed under general anaesthesia. 3 patients dropped out and were excluded from analysis. drop outs were not analysed.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias.

Park 2021

Some concerns

random computer allocation, no information on concealment of allocation. no baseline difference of demographics.

Some concerns

patients and block performers could not be blinded 1 patient dropped out in each group but they were not analysed later.

Low risk of bias

data available for almost all patients.

Low risk of bias

same measurement for same outcomes.

High risk of bias

preregistered RCT.

Some concerns

2 areas judged at some risk of bias.

Prasad 2020

Some concerns

random allocation using chit and box methid. no information on concealement. no baseline difference of demographics

Some concerns

blocks performedpre‐operatively. no information on patient drop outs.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. assessors were blinded.

Low risk of bias

preregistered RCT.

Some concerns

2 areas judged at some risk of bias.

Qiang 2018

Some concerns

random number generator used. no information onconcealement. no baseline difference of demographics.

High risk of bias

blocks performed while patients awake pre‐op. no information on study flow.

Some concerns

not enough information.

High risk of bias

same measurement for same outcomes. no information on blinding of assessors.

Low risk of bias

preregistered RCT.

High risk of bias

high risk of bias.

Seelam 2020

Low risk of bias

random computer allocation and sealed enevelope techniques used. no baseline difference of demographics.

Low risk of bias

single blinded study. no patients dropped out of the groups they were intially assigned to.

Low risk of bias

data available for all patients.

Some concerns

same measurement for same outcomes. no information on blinding of outcome assessors.

Low risk of bias

preregistered RCT.

Some concerns

some risks of bias.

Sharma 2020

Low risk of bias

Patients were allocated to either block group (group B) or control group (group C) using computer‐generated random numbers. Random numbers were concealed using opaque envelopes and were opened just prior to the administration of block. Patients in the block group received ultrasound‐guided erector spinae plane block at T5 level with ropivacaine (0.5%, 0.4 mL/kg) while those in the control group did not receive any intervention.

Some concerns

The anesthetist who performed the block was not blinded to group allocation.

Low risk of bias

No drop outs.

Low risk of bias

Prespecified appropriate measures of outcome. Same measurements for the same outcomes.

Low risk of bias

This study was carried out after getting approval from the institute's ethics committee (PGIMER Institutional ethics committee), reference no. NK/3870/MD/339 dated September 21, 2016, the study was conducted over a period of 1 year from July 2017 to December 2018. This study adheres to the applicable CONSORT guidelines.

Some concerns

One area rated at some risk of bias.

Singh 2019

Low risk of bias

random computer allocation and revleaed in anaesthetic room by statistician. no baseline difference of demographics.

Low risk of bias

pre‐op blocks while patients were awake. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

preregistered RCT.

Low risk of bias

all areas at low risk of bias.

Singh 2020

Some concerns

random computer allocation. No information onconcealement of allocation. no baseline difference of demographics.

Low risk of bias

preo‐op blocks. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

preregistered RCT.

Some concerns

1 area at some risk of bias.

Tulgar 2018

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline demographics.

Low risk of bias

block performed under anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Yaoping 2019

Some concerns

Allocation was randomised but no information on how this was achieved or how it was concealed. No difference in the baseline Demographics

High risk of bias

Block performed pre‐op. No further information.

High risk of bias

Not enough information in the translation.

High risk of bias

Same measurement for same outcomes. No further information from translated text.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

Not enough information in the translation.

Zhang 2020

Low risk of bias

Allocation was made randomly by lead investigator and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

block performed pre‐op. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurements for same outcomes.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Risk of bias for analysis 2.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

El Ghamry 2019b

Low risk of bias

This prospective, double‐blinded, randomised, controlled study was done from January 2018 to January 2019 in Tanta University Hospital after approval from our institutional ethical committee and obtaining the patients’ written informed consent. The observers who collected data and were responsible for postoperative follow‐up were blinded to the study groups.

Low risk of bias

All patients analysed in their allocated group.

Some concerns

Four drop outs in each group.

Low risk of bias

Prespecified appropriate measures of outcome. Same measurements for same outcomes. All parameters and study outcomes were recorded by a research anesthesiologist who was blinded to group assignment.

Low risk of bias

This prospective, double‐blinded, randomised, controlled study was done from January 2018 to January 2019 in Tanta University Hospital after approval from our institutional ethical committee and obtaining the patients’ written informed consent. The trial followed the CONSORT 2010 statement guidelines for conducting a randomised controlled trial.

Some concerns

One are was judged at some risk of bias.

Yang 2019

Some concerns

Allocation was made using random number method. No information as to concealement. No difference in the baseline Demographics.

High risk of bias

block performed under anaesthesia. no information as to study flow.

High risk of bias

not enough information in translation.

High risk of bias

same measurement for same outcomes.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

Most areas judged at high or some concern of bias

Hamed 2019

Low risk of bias

An online randomization program was used to generate a random number list (1:1 ratio). Patient randomization numbers were concealed in opaque envelopes that were opened by the study investigator. No difference in the baseline Demographics.

Low risk of bias

The ESPB group underwent US‐guided ESPB at T9 vertebrae level with 20 ml of bupivacaine 0.5%. The control group underwent the same procedure but had a sham injection (20 ml of saline). no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. All members of the study group involved in obtaining functional data were blinded to randomization for the period of data acquisition and analysis. Patients were randomly allocated to two groups.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas at low risk of bias

Ciftci 2020a

Low risk of bias

Randomising computer program used for allocation. Drugs (real and sham) were drawn up by staff who did not participate in patient care, so blinding was maintained throughout. no baseline difference in demographics

Low risk of bias

Double blinded trial were both groups received block (real and sham). Staff received premade solutions so were unaware which one was real and which was sham.

Low risk of bias

Data was available for all 60 participants.

Low risk of bias

same measurements for same outcomes. outcome assessors were blinded.

Low risk of bias

prospectively registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas judged at low risk of bias

Yao 2020b

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

Sham injections were used. Injections looked the same and were prepared by a trial nurse who was not involved in patient care.

Low risk of bias

Data was available for all included patients.

Low risk of bias

Same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Zhang 2021

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

block performed pre‐operatively. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurements for same outcomes. outcome assessors were blinded.

Low risk of bias

Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas at low risk of bias

Mostafa 2020

Low risk of bias

random computer allocation and sealed enevelope techniques used. no baseline difference of demographics.

Low risk of bias

Sham block in order to ensure double blinding. Blocks were performed by an anaesthesiologist who played no furhter role in the study.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes.

Low risk of bias

preregistered RCT.

Low risk of bias

all areas judged at low risk of bias

Abu 2019

Low risk of bias

Prospective randomized controlled double‐blind study was approved by the Hospital Ethics Committee (31507/04/17). Computer‐generated randomization numbers were used to randomly assign patients into 2 groups using sealed opaque envelopes that were randomly selected by each patient and contained a group number in which the patient was enrolled. Patients in the erector spinae plane block group received general anesthesia with preoperative bilat‐ eral ultrasound‐guided erector spinae plane block. Patients in the control group received general anesthesia with pre‐ operative bilateral sham ultrasound‐guided erector spinae plane block. Sixty patients were enrolled (Figure 2). Table 1 presents demographic data (age, sex, and weight) as well as duration of surgery in the studied groups. No statistically significant difference.

Low risk of bias

Patients in the erector spinae plane block group received general anesthesia with preoperative bilateral ultrasound‐guided erector spinae plane block. Patients in the control group received general anesthesia with pre‐ operative bilateral sham ultrasound‐guided erector spinae plane block. no drop outs and all patients analysed in their allocated groups

Low risk of bias

60 patients. 30 each group. No drop outs.

Low risk of bias

Prespecified appropriate measures of outcome. Same measurements for same outcomes. All parameters and study outcomes were recorded by a research anesthesiologist who was blinded to group assignment.

Low risk of bias

This prospective randomized controlled double‐blind study was approved by the Hospital Ethics Committee (31507/04/17), and informed written consent was obtained from all subjects participating in the trial. The trial was reg‐ istered before patient enrollment at Pan African Clinical Trials Registry (PACTR201706002382366) (principal investi‐ gator: S.F.M.; date of registration: June 27, 2017). Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas were judged low risk of bias.

Risk of bias for analysis 3.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Ciftci 2020c

Some concerns

randomising computer program. no information on how blinding was maintained. no difference in baseline demographics.

Low risk of bias

Patients were sedated with midazolam but intervention (or no intervention) happened pre‐operatively. No drop outs. All 90 patients analysed in the groups they were allocated.

Low risk of bias

Data for all patients available.

Low risk of bias

same measurement for same outcomes. outcome assessor was blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged as some concerns of risk of bias.

El Ghamry 2019a

Low risk of bias

Random allocation via computer randomising program. Allocation concealed using sealed opaque envelopes. No baseline difference in demographic.

Low risk of bias

Intervention performed before induciton of general anaesthesia. No dropouts, all patients analysed in group they were allocated.

Low risk of bias

Data for all patients available.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas judged low risk of bias

Fang 2019

Low risk of bias

Allocation was made using randomising computer program no information on concealement No difference in the baseline Demographics.

Some concerns

block performed ore‐operatively. 3 patients dropped out due to emergency thoracotomy or due to blocks not working. drop outs were excluded from analysis.

Low risk of bias

Data was available for all included patients.

High risk of bias

same measurement for same outcomes. no information regarding blinding of outcome assessors.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

High risk of bias

One area judged some concern and one high risk of bias

Guo 2019

Some concerns

random allocation but not specified how this was achieved, also no information on how allocation was concealed. No difference in the baseline Demographics.

Low risk of bias

block performed after induction of anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged at some concern for bias

Taketa 2019

Low risk of bias

Allocation was made using randomising computer program and performed by an anasesthesiologist who was then no longer involved in the patient's care. No difference in the baseline Demographics.

Low risk of bias

blocks were performed after induction of anaesthesia and the person who performed the block was not involved in the anaesthetic. no drop outs. all patients analysed in their allocated groups. 9 patients dropped out and were subsequently excluded from the analysis.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at risk of bias

Turhan 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline demographics.

Some concerns

block performed pre‐op. no information on study flow.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

Pain scores reported at multiple time points but all time points reported.

Some concerns

one area judged at risk of bias

Zhao 2020

Low risk of bias

Allocation via random number generator. Sealed envelopes were used to conceal allocation. No baseline difference in demographics.

Low risk of bias

Block performed preoperatively. No drop outs, all patients were analysed in their allocated group.

Low risk of bias

Data for all participants available

Low risk of bias

same measurements for same outcomes. outcome assessors blinded.

Low risk of bias

pregegistered RCT

Low risk of bias

all areas at low risk of bias

Risk of bias for analysis 3.2 Rates of block‐related adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Chen 2020

Low risk of bias

random number generator and sealed envelope technique. no statistically significant difference in preop demographics

Some concerns

Intervention happened after general anaesthetic induction. Envelopes were opened after induction. 3 patients dropped out (1 per group) due to surgical technique converison and being lost to follow up. 3 patients dropped out (1 per group) due to surgical technique converison and being lost to follow up. Those 3 patients were excluded. All other patients were analysed in the groups they were allocated. Those 3 patients were excluded. All other patients were analysed in the groups they were allocated.

Low risk of bias

Data available for all included patients. 2.67% patients excluded.

Low risk of bias

same measurements for same outcomes. outcome assessors were blinded

Low risk of bias

prospectively registeres RCT. Pain scores reported at multiple time points but all time points reported.

Some concerns

one area judged some concern of risk of bias.

Ciftci 2020c

Some concerns

randomising computer program. no information on how blinding was maintained. no difference in baseline demographics.

Low risk of bias

Patients were sedated with midazolam but intervention (or no intervention) happened pre‐operatively. No drop outs. All 90 patients analysed in the groups they were allocated.

Low risk of bias

Data for all patients available.

Low risk of bias

same measurement for same outcomes. outcome assessor was blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged as some concerns of risk of bias.

El Ghamry 2019a

Low risk of bias

Random allocation via computer randomising program. Allocation concealed using sealed opaque envelopes. No baseline difference in demographic.

Low risk of bias

Intervention performed before induciton of general anaesthesia. No dropouts, all patients analysed in group they were allocated.

Low risk of bias

Data for all patients available.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas judged low risk of bias

Fang 2019

Low risk of bias

Allocation was made using randomising computer program no information on concealement No difference in the baseline Demographics.

Some concerns

block performed ore‐operatively. 3 patients dropped out due to emergency thoracotomy or due to blocks not working. drop outs were excluded from analysis.

Low risk of bias

Data was available for all included patients.

High risk of bias

same measurement for same outcomes. no information regarding blinding of outcome assessors.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

High risk of bias

One area judged some concern and one high risk of bias

Swisher 2020

Low risk of bias

random computer allocation and sealed enevelope techniques used. no baseline difference of demographics

Low risk of bias

subject blinded trial. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes.

Low risk of bias

preregistered RCT.

Low risk of bias

all areas judged at low risk of bias

Taketa 2019

Low risk of bias

Allocation was made using randomising computer program and performed by an anasesthesiologist who was then no longer involved in the patient's care. No difference in the baseline Demographics.

Low risk of bias

blocks were performed after induction of anaesthesia and the person who performed the block was not involved in the anaesthetic. no drop outs. all patients analysed in their allocated groups. 9 patients dropped out and were subsequently excluded from the analysis.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at risk of bias

Turhan 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline demographics.

Some concerns

block performed pre‐op. no information on study flow.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes.outcome assessors were blinded.

Low risk of bias

Pain scores reported at multiple time points but all time points reported.

Some concerns

one area judged at risk of bias

Risk of bias for analysis 4.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Abdelhamid 2020

Low risk of bias

Patients were randomly assigned to either one of three study groups on the day of surgery. grouping was determined through a computer‐ generated list that was kept in a sealed enve‐ lope. Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. Demographic data, patient characteristics and durations of surgeries were comparable among the three groups.

Low risk of bias

Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. No drop outs. All patients analysed in their allocated group.

Low risk of bias

66 patients. 22 each group. No drop outs. Data available for all patients

Low risk of bias

Prespecified appropriate measures of outcome. Same measurements for the same outcomes. Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged.

Low risk of bias

Based on a pilot study, sample size was cal‐ culated according to the significant differ‐ ence in the mean of vas values over 24 hours (as a primary outcome) between control group (3.21±0.48), taP group (2.23±0.41) and esP group (1.51±0.35) using anova test, with α=0.05, power of 80%, and an effect size of 0.4. therefore, a sample size of 22 patients/group would be required (g‐Power 301). Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas were judged low risk of bias.

Altıparmak 2019a

Low risk of bias

Patients were randomly allocated into two groups based on a computerized randomization table created by a researcher who was not involved in the study. For each randomized patient, the OR anesthe‐ siologist took the corresponding sealed envelope from a folder, which indicated the treatment to be assigned to the patient. Demographics did not vary to a statistically significant amount (p>0.05).

Some concerns

Block performed after anaesthetic induction. Staff opened sealed envelope after induction. 2 patients were exluceded due to change in surgical technique and 2 were excluded due to PCA failure during follow up. These 4 patients were excluded from analysis. All other patients were analysed in the group they were allocated.

Low risk of bias

68 patient. 34 ESPB and 34 OSTAP. 2 Drop outs in each group due to PCA failure and conversion to open OP.

Low risk of bias

The NRS scores were recorded at the postoperative 15th min, 30th min, 60th min, 2nd hour, 12th hour and 24th hour by an anesthesiologist who was blinded to the group allocations. same measurements for same outcomes.

Low risk of bias

This prospective, single‐blinded, randomized, controlled study was conducted after receiving University Ethics Committee approval in ac‐ cordance with the principles outlined in the Declaration of Helsinki. We registered the study with the Australian New Zealand Clinical Trials Registry (Trial ID: ACTRN12618001078246) and used a CONSORT checklist for enrollment and allocation of patients (Fig. 1). Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged some concern.

Kamel 2020

Low risk of bias

Allocation was made using randomising computer program after the end of surgery but before the patient was woken up. No difference in the baseline Demographics.

Low risk of bias

Block performed under anaesthesia. No drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. In the recovery room, the outcome assessor (the anesthesiologist not sharing in the study) assessed the primary and secondary outcomes.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Risk of bias for analysis 4.2 Rates of block‐related adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Abdelhamid 2020

Low risk of bias

Patients were randomly assigned to either one of three study groups on the day of surgery. grouping was determined through a computer‐ generated list that was kept in a sealed enve‐ lope. Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. Demographic data, patient characteristics and durations of surgeries were comparable among the three groups.

Low risk of bias

Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. No drop outs. All patients analysed in their allocated group.

Low risk of bias

66 patients. 22 each group. No drop outs. Data available for all patients

Low risk of bias

Prespecified appropriate measures of outcome. Same measurements for the same outcomes. Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged.

Low risk of bias

Based on a pilot study, sample size was cal‐ culated according to the significant differ‐ ence in the mean of vas values over 24 hours (as a primary outcome) between control group (3.21±0.48), taP group (2.23±0.41) and esP group (1.51±0.35) using anova test, with α=0.05, power of 80%, and an effect size of 0.4. therefore, a sample size of 22 patients/group would be required (g‐Power 301). Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas were judged low risk of bias.

Altıparmak 2019a

Low risk of bias

Patients were randomly allocated into two groups based on a computerized randomization table created by a researcher who was not involved in the study. For each randomized patient, the OR anesthe‐ siologist took the corresponding sealed envelope from a folder, which indicated the treatment to be assigned to the patient. Demographics did not vary to a statistically significant amount (p>0.05).

Some concerns

Block performed after anaesthetic induction. Staff opened sealed envelope after induction. 2 patients were exluceded due to change in surgical technique and 2 were excluded due to PCA failure during follow up. These 4 patients were excluded from analysis. All other patients were analysed in the group they were allocated.

Low risk of bias

68 patient. 34 ESPB and 34 OSTAP. 2 Drop outs in each group due to PCA failure and conversion to open OP.

Low risk of bias

The NRS scores were recorded at the postoperative 15th min, 30th min, 60th min, 2nd hour, 12th hour and 24th hour by an anesthesiologist who was blinded to the group allocations. same measurements for same outcomes.

Low risk of bias

This prospective, single‐blinded, randomized, controlled study was conducted after receiving University Ethics Committee approval in ac‐ cordance with the principles outlined in the Declaration of Helsinki. We registered the study with the Australian New Zealand Clinical Trials Registry (Trial ID: ACTRN12618001078246) and used a CONSORT checklist for enrollment and allocation of patients (Fig. 1). Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged some concern.

Ibrahim 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

sham and real injections prepared by pharmacist no involved in patient care. no drop outs. all patients analysed in their allocated groups. no drop outs, all patients analysed in the groups they were allocated.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Kamel 2020

Low risk of bias

Allocation was made using randomising computer program after the end of surgery but before the patient was woken up. No difference in the baseline Demographics.

Low risk of bias

Block performed under anaesthesia. No drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. In the recovery room, the outcome assessor (the anesthesiologist not sharing in the study) assessed the primary and secondary outcomes.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Risk of bias for analysis 5.1 Rates of block‐related adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Elsabeeny 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline Demographics. There was no difference in baseline demographics.

Some concerns

Block was performed after induction of general anaesthesia. 5 patients dropped out. those patients were excluded from analysis.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area was judged at some risk of bias

Finnerty 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline Demographics.

Low risk of bias

Block were performed after induction of anaesthesia. No drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas judged at low risk of bias.

Risk of bias for analysis 6.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Altıparmak 2019b

Low risk of bias

After the endotracheal intubation, the patients were randomly allocated into two groups based on a computerized randomization table created by a researcher who was not involved in the study. The researcher assigned the random ID to each patient and the blinded anesthesiologist used this ID while collecting the postoperative data in the surgical ward. After the endotracheal intubation, the patients were randomly allocated into two groups based on a computerized randomization table created by a researcher who was not involved in the study. The researcher assigned the random ID to each patient and the blinded anesthesiologist used this ID while collecting the postoperative data in the surgical ward.

Low risk of bias

Patient were not aware because intervention was performed after induction of general anaesthetic. The staff performing the intervention were aware of it only after the patient was asleep. Data collection was performed by a blinded anaesthetist. Blinding not compromised because post operative care was the same for both groups. 2 patients from the PECS block group were lost to follow up due to PCA failure. excluded patients (n=2) were exclude from analysis.

Low risk of bias

38 female patients. 18 PECS, 20 ESPB. 2 Drop outs in PECS group due to PCA failure. Data available for all 38 patient that were included.

Low risk of bias

Prespecified appropriate measures of outcome. Outcome assessors were blinded to allocation.

Low risk of bias

The sample size of the study was calculated using the G*Power programme (v3.1.9) based on a pilot study with 10 patients in each group. NRS scores at 7 different time point (but all reported). Same analysis for all time points.

Low risk of bias

All areas scored low risk

Sinha 2019

Low risk of bias

random computer allocation and sealed enevelope techniques used. no baseline difference of demographics.

Low risk of bias

pre‐op blocks while patients were awake.no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. blinded outcomeassessors.

Low risk of bias

preregistered RCT.

Low risk of bias

all areas at low risk of bias.

Risk of bias for analysis 6.2 Rates of block‐related adverse outcomes.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Gad 2019

Some concerns

Allocation was made using randomising computer program, no information on concealement of allocation. No difference in the baseline Demographics.

Low risk of bias

Blocks performed after induction of anaesthesia. 3 patients were lost to follow up and excluded from analysis. all other patients analysed in their allocated groups. patients that dropped out were not analysed.

Low risk of bias

Data was available for all included patients.

High risk of bias

Same measurement for same outcomes. No information about blinding of staff.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

High risk of bias

One area judged high and one at some concern of risk of bias.

Risk of bias for analysis 7.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Tulgar 2018

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline demographics.

Low risk of bias

block performed under anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Risk of bias for analysis 7.2 Rates of block‐related adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Tulgar 2018

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline demographics.

Low risk of bias

block performed under anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Risk of bias for analysis 8.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Fiorelli 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

Blocks performed before induction of anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas at low risk of bias

Turhan 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline demographics.

Some concerns

block performed pre‐op. no information on study flow.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes.outcome assessors were blinded.

Low risk of bias

Pain scores reported at multiple time points but all time points reported.

Some concerns

one area judged at risk of bias.

Risk of bias for analysis 8.2 Rates of block‐related adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Chen 2020

Low risk of bias

random number generator and sealed envelope technique. no statistically significant difference in preop demographics

Some concerns

Intervention happened after general anaesthetic induction. Envelopes were opened after induction. 3 patients dropped out (1 per group) due to surgical technique converison and being lost to follow up. 3 patients dropped out (1 per group) due to surgical technique converison and being lost to follow up. Those 3 patients were excluded. All other patients were analysed in the groups they were allocated. Those 3 patients were excluded. All other patients were analysed in the groups they were allocated.

Low risk of bias

Data available for all included patients. 2.67% patients excluded.

Low risk of bias

same measurements for same outcomes. outcome assessors were blinded.

Low risk of bias

prospectively registeres RCT. Pain scores reported at multiple time points but all time points reported.

Some concerns

one area judged some concern of risk of bias.

Fiorelli 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

Blocks performed before induction of anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas at low risk of bias

Turhan 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline demographics.

Some concerns

block performed pre‐op. no information on study flow.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes.outcome assessors were blinded.

Low risk of bias

Pain scores reported at multiple time points but all time points reported.

Some concerns

one area judged at risk of bias.

Risk of bias for analysis 9.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Nagaraja 2018

Some concerns

randomised trial but no further information. no baseline difference of demographics.

Some concerns

blocks performed pre‐operatively while patient awake. no information on study flow.

Low risk of bias

data available for all patients.

High risk of bias

same measurement for same outcomes. no information on blinding of outcome assessors.

Low risk of bias

preregistered RCT.

High risk of bias

1 area judged at high risk of bias

Sakae 2020

Low risk of bias

random computer allocation right before surgery. no baseline difference of demographics.

Low risk of bias

single blinded study. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

preregistered RCT.

Low risk of bias

All areas judged at low risk of bias.

Risk of bias for analysis 9.2 Rates of block‐related adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Nagaraja 2018

Some concerns

randomised trial but no further information. no baseline difference of demographics.

Some concerns

blocks performed pre‐operatively while patient awake. no information on study flow.

Low risk of bias

data available for all patients.

High risk of bias

same measurement for same outcomes. no information on blinding of outcome assessors.

Low risk of bias

preregistered RCT.

High risk of bias

1 areas judged at high risk of bias

Risk of bias for analysis 10.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS) ‐ Type of surgery.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.1.1 Abdominal surgery

Abdelhamid 2020

Low risk of bias

Patients were randomly assigned to either one of three study groups on the day of surgery. grouping was determined through a computer‐ generated list that was kept in a sealed enve‐ lope. Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. Demographic data, patient characteristics and durations of surgeries were comparable among the three groups.

Low risk of bias

Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. No drop outs. All patients analysed in their allocated group.

Low risk of bias

66 patients. 22 each group. No drop outs. Data available for all patients

Low risk of bias

Prespecified appropriate measures of outcome. Same measurements for the same outcomes. Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged.

Low risk of bias

Based on a pilot study, sample size was cal‐ culated according to the significant differ‐ ence in the mean of vas values over 24 hours (as a primary outcome) between control group (3.21±0.48), taP group (2.23±0.41) and esP group (1.51±0.35) using anova test, with α=0.05, power of 80%, and an effect size of 0.4. therefore, a sample size of 22 patients/group would be required (g‐Power 301). Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas were judged low risk of bias.

Fu 2020

Some concerns

Allocation was made using randomising computer program, no information on concealement of blinding. No difference in the baseline Demographics.

Low risk of bias

Blocks performed preoperatively. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged to be at some concern for risk of bias

Gultekin 2020

Some concerns

Allocation was made using randomising computer program, no information on concealement of allocation. No difference in the baseline Demographics.

Low risk of bias

block performed after induction of anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged at some concern of risk of bias

Prasad 2020

Some concerns

random allocation using chit and box methid. no information on concealement. no baseline difference of demographics

Some concerns

blocks performedpre‐operatively. no information on patient drop outs.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. assessors were blinded.

Low risk of bias

preregistered RCT.

Some concerns

2 areas judged at some risk of bias.

Subgroup 10.1.2 Breast surgery

Elsabeeny 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline Demographics. There was no difference in baseline demographics.

Some concerns

Block was performed after induction of general anaesthesia. 5 patients dropped out. those patients were excluded from analysis.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area was judged at some risk of bias

Park 2021

Some concerns

random computer allocation, no information on concealment of allocation. no baseline difference of demographics.

Some concerns

patients and block performers could not be blinded 1 patient dropped out in each group but they were not analysed later.

Low risk of bias

data available for almost all patients.

Low risk of bias

same measurement for same outcomes.

Low risk of bias

preregistered RCT.

Some concerns

2 areas judged at some risk of bias.

Singh 2019

Low risk of bias

random computer allocation and revleaed in anaesthetic room by statistician. no baseline difference of demographics.

Low risk of bias

pre‐op blocks while patients were awake. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

preregistered RCT.

Low risk of bias

all areas at low risk of bias

Subgroup 10.1.3 Thoracic surgery

Ciftci 2020c

Some concerns

randomising computer program. no information on how blinding was maintained. no difference in baseline demographics.

Low risk of bias

Patients were sedated with midazolam but intervention (or no intervention) happened pre‐operatively. No drop outs. All 90 patients analysed in the groups they were allocated.

Low risk of bias

Data for all patients available.

Low risk of bias

same measurement for same outcomes. outcome assessor was blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged as some concerns of risk of bias.

Ciftci 2020d

Low risk of bias

randomising computer program. Block performed 30 minutes preop. No information on how allocation was kept concealed. No demographic differences at baseline.

Low risk of bias

Block was performed 30mins preop. Patients in both groups were given midazolam but control group did not receive a block. 30 patients in each group, no drop outs.

Low risk of bias

Data were available for all patients.

Low risk of bias

same measurements for same outcomes. outcome assessors were blinded.

Low risk of bias

prospectively registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas low risk of bias.

Liu 2021

Low risk of bias

Patients were randomised using a random number generator online. The randomized schedule was hidden and generated by the anesthesiologist in charge of the case, who was not part of the study.

Low risk of bias

blocks performed under general anaesthesia. 3 patients dropped out and were excluded from analysis. drop outs were not analysed.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias.

Sobhy 2020

Low risk of bias

random number generator and sealed enevelope techniques used. no baseline difference of demographics.

Low risk of bias

pre‐op application of block. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. blinded outcome assessors.

Low risk of bias

preregistered RCT.

Low risk of bias

all areas at low risk of bias.

Yaoping 2019

Some concerns

Allocation was randomised but no information on how this was achieved or how it was concealed. No difference in the baseline Demographics

High risk of bias

Block performed pre‐op. No further information.

High risk of bias

Not enough information in the translation .

High risk of bias

Same measurement for same outcomes. No further information from translated text.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

Not enough information in the translation .

Zheng 2019

Some concerns

Allocation was made using random number sequence, no information about concealement. No difference in the baseline Demographics.

High risk of bias

blocks performed under anaesthesia. no further information in translation.

High risk of bias

No information on study flow

High risk of bias

same measurement for same outcomes. no information on blinding of assessors.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

High risk of bias

Subgroup 10.1.4 Spine surgery

Eskin 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

Opening of envelopes and intervention post induction of anaesthesia. No drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

Same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas judged low risk of bias.

Singh 2020

Some concerns

random computer allocation. No information onconcealement of allocation. no baseline difference of demographics.

Low risk of bias

preo‐op blocks. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

preregistered RCT.

Some concerns

1 area at some risk of bias.

Yayik 2018

Some concerns

No information in abstract other than to say that it was a randomised trial. No difference in the baseline Demographics.

High risk of bias

Not enough information in abstract.

High risk of bias

Not enough information in abstract.

High risk of bias

same outcome measurements for same outcomes, no further information.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

Not enough information in abstract.

Zhang 2020

Low risk of bias

Allocation was made randomly by lead investigator and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

block performed pre‐op. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurements for same outcomes.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Subgroup 10.1.5 Hip surgery

Tulgar 2018

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline demographics.

Low risk of bias

block performed under anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Risk of bias for analysis 10.2 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS) ‐ Type of local anaesthetic.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.2.1 Bupivacaine 0.25% to 0.5%

Abdelhamid 2020

Low risk of bias

Patients were randomly assigned to either one of three study groups on the day of surgery. grouping was determined through a computer‐ generated list that was kept in a sealed enve‐ lope. Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. Demographic data, patient characteristics and durations of surgeries were comparable among the three groups.

Low risk of bias

Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. No drop outs. All patients analysed in their allocated group.

Low risk of bias

66 patients. 22 each group. No drop outs. Data available for all patients

Low risk of bias

Prespecified appropriate measures of outcome. Same measurements for the same outcomes. Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged.

Low risk of bias

Based on a pilot study, sample size was cal‐ culated according to the significant differ‐ ence in the mean of vas values over 24 hours (as a primary outcome) between control group (3.21±0.48), taP group (2.23±0.41) and esP group (1.51±0.35) using anova test, with α=0.05, power of 80%, and an effect size of 0.4. therefore, a sample size of 22 patients/group would be required (g‐Power 301). Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas were judged low risk of bias.

Ciftci 2020c

Some concerns

randomising computer program. no information on how blinding was maintained. no difference in baseline demographics.

Low risk of bias

Patients were sedated with midazolam but intervention (or no intervention) happened pre‐operatively. No drop outs. All 90 patients analysed in the groups they were allocated.

Low risk of bias

Data for all patients available.

Low risk of bias

same measurement for same outcomes. outcome assessor was blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged as some concerns of risk of bias.

Ciftci 2020d

Low risk of bias

randomising computer program. Block performed 30 minutes preop. No information on how allocation was kept concealed. No demographic differences at baseline.

Low risk of bias

Block was performed 30mins preop. Patients in both groups were given midazolam but control group did not receive a block. 30 patients in each group, no drop outs.

Low risk of bias

Data were available for all patients.

Low risk of bias

same measurements for same outcomes. outcome assessors were blinded.

Low risk of bias

prospectively registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas low risk of bias.

Elsabeeny 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline Demographics. There was no difference in baseline demographics.

Some concerns

Block was performed after induction of general anaesthesia. 5 patients dropped out. those patients were excluded from analysis.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area was judged at some risk of bias

Eskin 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

Opening of envelopes and intervention post induction of anaesthesia. No drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

Same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas judged low risk of bias.

Gultekin 2020

Some concerns

Allocation was made using randomising computer program, no information on concealement of allocation. No difference in the baseline Demographics.

Low risk of bias

block performed after induction of anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged at some concern of risk of bias

Tulgar 2018

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline demographics.

Low risk of bias

block performed under anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Yayik 2018

Some concerns

No information in abstract other than to say that it was a randomised trial. No difference in the baseline Demographics.

High risk of bias

Not enough information in abstract.

High risk of bias

Not enough information in abstract.

High risk of bias

same outcome measurements for same outcomes, no further information.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

Not enough information in abstract.

Subgroup 10.2.2 Ropivacaine 0.3% to 0.5%

Fu 2020

Some concerns

Allocation was made using randomising computer program, no information on concealement of blinding. No difference in the baseline Demographics.

Low risk of bias

Blocks performed preoperatively. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged to be at some concern for risk of bias

Liu 2021

Low risk of bias

Patients were randomised using a random number generator online. The randomized schedule was hidden and generated by the anesthesiologist in charge of the case, who was not part of the study.

Low risk of bias

blocks performed under general anaesthesia. 3 patients dropped out and were excluded from analysis. drop outs were not analysed.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias.

Park 2021

Some concerns

random computer allocation, no information on concealment of allocation. no baseline difference of demographics.

Some concerns

1 patient dropped out in each group but they were not analysed later.

Low risk of bias

data available for almost all patients.

Low risk of bias

same measurement for same outcomes.

Low risk of bias

preregistered RCT.

Some concerns

2 areas judged at some risk of bias.

Prasad 2020

Some concerns

random allocation using chit and box methid. no information on concealement. no baseline difference of demographics

Some concerns

blocks performedpre‐operatively. no information on patient drop outs.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. assessors were blinded.

Low risk of bias

preregistered RCT.

Some concerns

2 areas judged at some risk of bias.

Singh 2019

Low risk of bias

random computer allocation and revleaed in anaesthetic room by statistician. no baseline difference of demographics.

Low risk of bias

pre‐op blocks while patients were awake. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

preregistered RCT.

Low risk of bias

all areas at low risk of bias.

Singh 2020

Some concerns

random computer allocation. No information onconcealement of allocation. no baseline difference of demographics.

Low risk of bias

preo‐op blocks. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

preregistered RCT.

Some concerns

1 area at some risk of bias.

Sobhy 2020

Low risk of bias

random number generator and sealed enevelope techniques used. no baseline difference of demographics.

Low risk of bias

pre‐op application of block. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. blinded outcome assessors.

Low risk of bias

preregistered RCT.

Low risk of bias

all areas at low risk of bias.

Yaoping 2019

Some concerns

Allocation was randomised but no information on how this was achieved or how it was concealed. No difference in the baseline Demographics

High risk of bias

Block performed pre‐op. No further information.

High risk of bias

Not enough information in the translation .

High risk of bias

Same measurement for same outcomes. No further information from translated text.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

Not enough information in the translation .

Zhang 2020

Low risk of bias

Allocation was made randomly by lead investigator and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

Allocation was made randomly by lead investigator and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurements for same outcomes.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Zheng 2019

Some concerns

Allocation was made using random number sequence, no information about concealement. No difference in the baseline Demographics.

High risk of bias

blocks performed under anaesthesia. no further information in translation.

High risk of bias

No information on study flow

High risk of bias

same measurement for same outcomes. no information on blinding of assessors.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

High risk of bias

Risk of bias for analysis 10.3 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS) ‐ Volume of local anaesthetic.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.3.1 15 mL to 19 mL

Abdelhamid 2020

Low risk of bias

Patients were randomly assigned to either one of three study groups on the day of surgery. grouping was determined through a computer‐ generated list that was kept in a sealed enve‐ lope. Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. Demographic data, patient characteristics and durations of surgeries were comparable among the three groups.

Low risk of bias

Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. No drop outs. All patients analysed in their allocated group.

Low risk of bias

66 patients. 22 each group. No drop outs. Data available for all patients

Low risk of bias

Prespecified appropriate measures of outcome. Same measurements for the same outcomes. Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged.

Low risk of bias

Based on a pilot study, sample size was cal‐ culated according to the significant differ‐ ence in the mean of vas values over 24 hours (as a primary outcome) between control group (3.21±0.48), taP group (2.23±0.41) and esP group (1.51±0.35) using anova test, with α=0.05, power of 80%, and an effect size of 0.4. therefore, a sample size of 22 patients/group would be required (g‐Power 301). Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas were judged low risk of bias.

Subgroup 10.3.2 20 mL to 24 mL

Ciftci 2020c

Some concerns

randomising computer program. no information on how blinding was maintained. no difference in baseline demographics.

Low risk of bias

Patients were sedated with midazolam but intervention (or no intervention) happened pre‐operatively. No drop outs. All 90 patients analysed in the groups they were allocated.

Low risk of bias

Data for all patients available.

Low risk of bias

same measurement for same outcomes. outcome assessor was blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged as some concerns of risk of bias.

Ciftci 2020d

Low risk of bias

randomising computer program. Block performed 30 minutes preop. No information on how allocation was kept concealed. No demographic differences at baseline.

Low risk of bias

Block was performed 30mins preop. Patients in both groups were given midazolam but control group did not receive a block. 30 patients in each group, no drop outs.

Low risk of bias

Data were available for all patients.

Low risk of bias

same measurements for same outcomes. outcome assessors were blinded.

Low risk of bias

prospectively registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas low risk of bias.

Eskin 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

Opening of envelopes and intervention post induction of anaesthesia. No drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

Same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas judged low risk of bias.

Fu 2020

Some concerns

Allocation was made using randomising computer program, no information on concealement of blinding. No difference in the baseline Demographics.

Low risk of bias

Blocks performed preoperatively. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged to be at some concern for risk of bias

Gultekin 2020

Some concerns

Allocation was made using randomising computer program, no information on concealement of allocation. No difference in the baseline Demographics.

Low risk of bias

block performed after induction of anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged at some concern of risk of bias

Prasad 2020

Some concerns

random allocation using chit and box methid. no information on concealement. no baseline difference of demographics

Some concerns

blocks performedpre‐operatively. no information on patient drop outs.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. assessors were blinded.

Low risk of bias

preregistered RCT.

Some concerns

2 areas judged at some risk of bias.

Singh 2019

Low risk of bias

random computer allocation and revleaed in anaesthetic room by statistician. no baseline difference of demographics.

Low risk of bias

pre‐op blocks while patients were awake. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

preregistered RCT.

Low risk of bias

all areas at low risk of bias.

Singh 2020

Some concerns

random computer allocation. No information onconcealement of allocation. no baseline difference of demographics.

Low risk of bias

preo‐op blocks. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

preregistered RCT.

Some concerns

1 area at some risk of bias.

Sobhy 2020

Low risk of bias

random number generator and sealed enevelope techniques used. no baseline difference of demographics.

Low risk of bias

pre‐op application of block. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. blinded outcome assessors.

Low risk of bias

preregistered RCT.

Low risk of bias

all areas at low risk of bias.

Tulgar 2018

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline demographics.

Low risk of bias

block performed under anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Yayik 2018

Some concerns

No information in abstract other than to say that it was a randomised trial. No difference in the baseline Demographics.

High risk of bias

Not enough information in abstract.

High risk of bias

Not enough information in abstract.

High risk of bias

same outcome measurements for same outcomes, no further information.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

Not enough information in abstract.

Subgroup 10.3.3 25 mL to 30 mL

Elsabeeny 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline Demographics. There was no difference in baseline demographics.

Some concerns

Block was performed after induction of general anaesthesia. 5 patients dropped out. those patients were excluded from analysis.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area was judged at some risk of bias

Liu 2021

Low risk of bias

Patients were randomised using a random number generator online. The randomized schedule was hidden and generated by the anesthesiologist in charge of the case, who was not part of the study.

Low risk of bias

blocks performed under general anaesthesia. 3 patients dropped out and were excluded from analysis. drop outs were not analysed.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias.

Park 2021

Some concerns

random computer allocation, no information on concealment of allocation. no baseline difference of demographics.

Some concerns

patients and block performers could not be blinded 1 patient dropped out in each group but they were not analysed later.

Low risk of bias

data available for almost all patients.

Low risk of bias

same measurement for same outcomes.

Low risk of bias

preregistered RCT.

Some concerns

2 areas judged at some risk of bias.

Yaoping 2019

Some concerns

Allocation was randomised but no information on how this was achieved or how it was concealed. No difference in the baseline Demographics

High risk of bias

Block performed pre‐op. No further information.

High risk of bias

Not enough information in the translation .

High risk of bias

Same measurement for same outcomes. No further information from translated text.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

Not enough information in the translation.

Zhang 2020

Low risk of bias

Allocation was made randomly by lead investigator and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

block performed pre‐op. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurements for same outcomes.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Zheng 2019

Some concerns

Allocation was made using random number sequence, no information about concealement. No difference in the baseline Demographics.

High risk of bias

blocks performed under anaesthesia. no further information in translation.

High risk of bias

No information on study flow

High risk of bias

same measurement for same outcomes. no information on blinding of assessors.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

high risk of bias

Risk of bias for analysis 10.4 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS) ‐ Fixed multimodal analgesia.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.4.1 Non‐opioid + PCA

Ciftci 2020d

Low risk of bias

randomising computer program. Block performed 30 minutes preop. No information on how allocation was kept concealed. No demographic differences at baseline.

Low risk of bias

Block was performed 30mins preop. Patients in both groups were given midazolam but control group did not receive a block. 30 patients in each group, no drop outs.

Low risk of bias

Data were available for all patients.

Low risk of bias

same measurements for same outcomes. outcome assessors were blinded.

Low risk of bias

prospectively registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas low risk of bias.

Eskin 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

Opening of envelopes and intervention post induction of anaesthesia. No drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

Same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas judged low risk of bias.

Liu 2021

Low risk of bias

Patients were randomised using a random number generator online. The randomized schedule was hidden and generated by the anesthesiologist in charge of the case, who was not part of the study.

Low risk of bias

blocks performed under general anaesthesia. 3 patients dropped out and were excluded from analysis. drop outs were not analysed.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias.

Park 2021

Some concerns

random computer allocation, no information on concealment of allocation. no baseline difference of demographics.

Some concerns

patients and block performers could not be blinded 1 patient dropped out in each group but they were not analysed later.

Low risk of bias

data available for almost all patients.

Low risk of bias

same measurement for same outcomes.

Low risk of bias

preregistered RCT.

Some concerns

2 areas judged at some risk of bias.

Prasad 2020

Some concerns

random allocation using chit and box methid. no information on concealement. no baseline difference of demographics

Some concerns

blocks performedpre‐operatively. no information on patient drop outs.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. assessors were blinded.

Low risk of bias

preregistered RCT.

Some concerns

2 areas judged at some risk of bias.

Sobhy 2020

Low risk of bias

random number generator and sealed enevelope techniques used. no baseline difference of demographics.

Low risk of bias

pre‐op application of block. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. blinded outcome assessors.

Low risk of bias

pre‐registered RCT

Low risk of bias

all areas at low risk of bias

Tulgar 2018

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline demographics.

Low risk of bias

block performed under anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Zhang 2020

Low risk of bias

Allocation was made randomly by lead investigator and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

block performed pre‐op. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurements for same outcomes.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Subgroup 10.4.2 Non‐opioid + rescue opioids

Abdelhamid 2020

Low risk of bias

Patients were randomly assigned to either one of three study groups on the day of surgery. grouping was determined through a computer‐ generated list that was kept in a sealed enve‐ lope. Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. Demographic data, patient characteristics and durations of surgeries were comparable among the three groups.

Low risk of bias

Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. No drop outs. All patients analysed in their allocated group.

Low risk of bias

66 patients. 22 each group. No drop outs. Data available for all patients

Low risk of bias

Prespecified appropriate measures of outcome. Same measurements for the same outcomes. Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged.

Low risk of bias

Based on a pilot study, sample size was cal‐ culated according to the significant differ‐ ence in the mean of vas values over 24 hours (as a primary outcome) between control group (3.21±0.48), taP group (2.23±0.41) and esP group (1.51±0.35) using anova test, with α=0.05, power of 80%, and an effect size of 0.4. therefore, a sample size of 22 patients/group would be required (g‐Power 301). Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas were judged low risk of bias.

Ciftci 2020c

Some concerns

randomising computer program. no information on how blinding was maintained. no difference in baseline demographics.

Low risk of bias

Patients were sedated with midazolam but intervention (or no intervention) happened pre‐operatively. No drop outs. All 90 patients analysed in the groups they were allocated.

Low risk of bias

Data for all patients available.

Low risk of bias

same measurement for same outcomes. outcome assessor was blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged as some concerns of risk of bias.

Elsabeeny 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline Demographics. There was no difference in baseline demographics.

Some concerns

Block was performed after induction of general anaesthesia. 5 patients dropped out. those patients were excluded from analysis.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area was judged at some risk of bias

Fu 2020

Some concerns

Allocation was made using randomising computer program, no information on concealement of blinding. No difference in the baseline Demographics.

Low risk of bias

Blocks performed preoperatively. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged to be at some concern for risk of bias

Gultekin 2020

Some concerns

Allocation was made using randomising computer program, no information on concealement of allocation. No difference in the baseline Demographics.

Low risk of bias

block performed after induction of anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged at some concern of risk of bias

Singh 2019

Low risk of bias

random computer allocation and revleaed in anaesthetic room by statistician. no baseline difference of demographics.

Low risk of bias

pre‐op blocks while patients were awake. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

preregistered RCT.

Low risk of bias

all areas at low risk of bias.

Singh 2020

Some concerns

random computer allocation. No information onconcealement of allocation. no baseline difference of demographics.

Low risk of bias

preo‐op blocks. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

preregistered RCT.

Some concerns

1 area at some risk of bias.

Subgroup 10.4.3 No specified multimodal analgesia

Yaoping 2019

Some concerns

Allocation was randomised but no information on how this was achieved or how it was concealed. No difference in the baseline Demographics

High risk of bias

Block performed pre‐op. No further information.

High risk of bias

Not enough information in the translation .

High risk of bias

Same measurement for same outcomes. No further information from translated text.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

Not enough information in the translation .

Yayik 2018

Some concerns

No information in abstract other than to say that it was a randomised trial. No difference in the baseline Demographics.

High risk of bias

Not enough information in abstract.

High risk of bias

Not enough information in abstract.

High risk of bias

same outcome measurements for same outcomes, no further information.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

Not enough information in abstract.

Zheng 2019

Some concerns

Allocation was made using random number sequence, no information about concealement. No difference in the baseline Demographics.

High risk of bias

blocks performed under anaesthesia. no further information in translation.

High risk of bias

No information on study flow

High risk of bias

same measurement for same outcomes. no information on blinding of assessors.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

High risk of bias

Risk of bias for analysis 10.5 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS) ‐ Baseline non‐opioid analgesia.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.5.1 Paracetamol

Abdelhamid 2020

Low risk of bias

Patients were randomly assigned to either one of three study groups on the day of surgery. grouping was determined through a computer‐ generated list that was kept in a sealed enve‐ lope. Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. Demographic data, patient characteristics and durations of surgeries were comparable among the three groups.

Low risk of bias

Both patients and the anesthetists involved in postoperative data collection were blinded as to the treatment group to which each patient belonged. all groups initially received general anesthesia and one of three analgesic techniques was then conducted. No drop outs. All patients analysed in their allocated group.

Low risk of bias

66 patients. 22 each group. No drop outs. Data available for all patients

Low risk of bias

same measurements for same outcome

Low risk of bias

Based on a pilot study, sample size was cal‐ culated according to the significant differ‐ ence in the mean of vas values over 24 hours (as a primary outcome) between control group (3.21±0.48), taP group (2.23±0.41) and esP group (1.51±0.35) using anova test, with α=0.05, power of 80%, and an effect size of 0.4. therefore, a sample size of 22 patients/group would be required (g‐Power 301). Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas were judged low risk of bias.

Gultekin 2020

Some concerns

Allocation was made using randomising computer program, no information on concealement of allocation. No difference in the baseline Demographics.

Low risk of bias

block performed after induction of anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged at some concern of risk of bias

Park 2021

Some concerns

random computer allocation, no information on concealment of allocation. no baseline difference of demographics.

Some concerns

1 patient dropped out in each group but they were not analysed later.

Low risk of bias

data available for almost all patients.

Low risk of bias

same measurement for same outcomes.

Low risk of bias

preregistered RCT.

Some concerns

2 areas judged at some risk of bias.

Singh 2020

Some concerns

random computer allocation. No information onconcealement of allocation. no baseline difference of demographics.

Low risk of bias

preo‐op blocks. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

preregistered RCT.

Some concerns

1 area at some risk of bias.

Sobhy 2020

Low risk of bias

random number generator and sealed enevelope techniques used. no baseline difference of demographics.

Low risk of bias

pre‐op application of block. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. blinded outcome assessors.

Low risk of bias

preregistered RCT.

Low risk of bias

preregistered RCT.

Subgroup 10.5.2 NSAID

Ciftci 2020d

Low risk of bias

randomising computer program. Block performed 30 minutes preop. No information on how allocation was kept concealed. No demographic differences at baseline.

Low risk of bias

Block was performed 30mins preop. Patients in both groups were given midazolam but control group did not receive a block. 30 patients in each group, no drop outs.

Low risk of bias

Data were available for all patients.

Low risk of bias

same measurements for same outcomes. outcome assessors were blinded.

Low risk of bias

prospectively registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas low risk of bias.

Liu 2021

Low risk of bias

Patients were randomised using a random number generator online. The randomized schedule was hidden and generated by the anesthesiologist in charge of the case, who was not part of the study.

Low risk of bias

blocks performed under general anaesthesia. 3 patients dropped out and were excluded from analysis. drop outs were not analysed.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias.

Tulgar 2018

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline demographics.

Low risk of bias

block performed under anaesthesia. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Yaoping 2019

Some concerns

Allocation was randomised but no information on how this was achieved or how it was concealed. No difference in the baseline Demographics

High risk of bias

Block performed pre‐op. No further information.

High risk of bias

Not enough information in the translation .

High risk of bias

Same measurement for same outcomes. No further information from translated text.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

Not enough information in the translation .

Zheng 2019

Some concerns

Allocation was made using random number sequence, no information about concealement. No difference in the baseline Demographics.

High risk of bias

blocks performed under anaesthesia. no further information in translation.

High risk of bias

No information on study flow

High risk of bias

same measurement for same outcomes. no information on blinding of assessors.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

high risk of bias in several areas

Subgroup 10.5.3 Not specified

Ciftci 2020c

Some concerns

randomising computer program. no information on how blinding was maintained. no difference in baseline demographics.

Low risk of bias

Patients were sedated with midazolam but intervention (or no intervention) happened pre‐operatively. No drop outs. All 90 patients analysed in the groups they were allocated.

Low risk of bias

Data for all patients available.

Low risk of bias

same measurement for same outcomes. outcome assessor was blinded.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

One area judged as some concerns of risk of bias.

Fu 2020

Some concerns

Allocation was made using randomising computer program, no information on concealement of blinding. No difference in the baseline Demographics.

Low risk of bias

Blocks performed preoperatively. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area judged to be at some concern for risk of bias

Yayik 2018

Some concerns

No information in abstract other than to say that it was a randomised trial. No difference in the baseline Demographics.

High risk of bias

Not enough information in abstract.

High risk of bias

Not enough information in abstract.

High risk of bias

same outcome measurements for same outcomes, no further information.

Some concerns

Pain scores reported at multiple time points but all time points reported.

High risk of bias

Not enough information in abstract.

Zhang 2020

Low risk of bias

Allocation was made randomly by lead investigator and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

block performed pre‐op. no drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurements for same outcomes.

Low risk of bias

pre‐registered RCT. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

all areas at low risk of bias

Subgroup 10.5.4 Both paracetamol + NSAID

Elsabeeny 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes. No difference in the baseline Demographics. There was no difference in baseline demographics.

Some concerns

Block was performed after induction of general anaesthesia. 5 patients dropped out. those patients were excluded from analysis.

Low risk of bias

Data was available for all included patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Some concerns

One area was judged at some risk of bias

Eskin 2020

Low risk of bias

Allocation was made using randomising computer program and concealed using sealed envelopes No difference in the baseline Demographics.

Low risk of bias

Opening of envelopes and intervention post induction of anaesthesia. No drop outs. all patients analysed in their allocated groups.

Low risk of bias

Data was available for all included patients.

Low risk of bias

Same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

pre‐registered trial. Pain scores reported at multiple time points but all time points reported.

Low risk of bias

All areas judged low risk of bias.

Prasad 2020

Some concerns

random allocation using chit and box methid. no information on concealement. no baseline difference of demographics

Some concerns

blocks performedpre‐operatively. no information on patient drop outs.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. assessors were blinded.

Low risk of bias

preregistered RCT.

Some concerns

2 areas judged at some risk of bias.

Singh 2019

Low risk of bias

random computer allocation and revleaed in anaesthetic room by statistician. no baseline difference of demographics.

Low risk of bias

pre‐op blocks while patients were awake. no drop outs, all patients analysed in allocated groups.

Low risk of bias

data available for all patients.

Low risk of bias

same measurement for same outcomes. outcome assessors were blinded.

Low risk of bias

preregistered RCT.

Low risk of bias

all areas at low risk of bias.

Appendices

Appendix 1. MEDLINE search strategy from initial search on 4 January 2021 and repeat search on 3 January 2022

1 (erector spinae plane or ESP block*).mp. (438)
2 ((erector or paraspinal or thoracic neuropathic pain) and block*).mp. (608)
3 (erector spinae and (an?esth* or analg*)).mp. (357)
4 (paraspinal and ((regional or local) adj2 (an?esth* or analg*))).mp. (115)
5 Paraspinal Muscles/ or Cardiac Surgical Procedures/ (53124)
6 exp Nerve Block/ or Anesthesia, Local/ (37546)
7 5 and 6 (192)
8 1 or 2 or 3 or 4 or 7 (719)
9 ((randomized controlled trial or controlled clinical trial).pt. or randomi?ed.ab. or placebo.ab. or drug therapy.fs. or randomly.ab. or trial.ab. or groups.ab.) not (exp animals/ not humans.sh.) (4036322)
10 8 and 9 (156)
11 "Systematic Review"/ or Meta‐analysis/ or exp meta‐analysis as topic/ or (meta analy* or metaanaly*).ti,ab. or ((review* or search*) adj10 (systemat* or literature* or international database* or medical database* or medline or pubmed or embase or cochrane or cinahl or biosis)).ti,ab. (655500)
12 8 and 11 (12)
13 10 or 12 (163)

Data and analyses

Comparison 1. Erector spinae plane block vs no block.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS)	17	958	Mean Difference (IV, Random, 95% CI)	‐0.77 [‐1.08, ‐0.46]
1.2 Rates of block‐related adverse events	18	1045	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
1.3 Mean difference in postoperative pain intensity during activity (24 hours postoperatively) (VAS)	9	518	Mean Difference (IV, Random, 95% CI)	‐1.31 [‐1.69, ‐0.93]
1.4 Mean difference in postoperative pain intensity at rest ‐ 2 hours (VAS)	10	611	Mean Difference (IV, Random, 95% CI)	‐2.08 [‐2.59, ‐1.58]
1.5 Mean difference in postoperative pain intensity at rest ‐ 48 hours (VAS)	6	360	Mean Difference (IV, Random, 95% CI)	‐0.47 [‐0.79, ‐0.14]
1.6 Mean difference in postoperative pain intensity during activity ‐ 2 hours (VAS)	4	240	Mean Difference (IV, Random, 95% CI)	‐3.03 [‐4.29, ‐1.77]
1.7 Mean difference in postoperative pain intensity during activity ‐ 48 hours (VAS)	5	280	Mean Difference (IV, Random, 95% CI)	‐0.48 [‐0.81, ‐0.16]
1.8 Rate of chronic post‐surgical pain after 3 months	1	58	Risk Ratio (M‐H, Random, 95% CI)	0.80 [0.46, 1.40]
1.9 Rate of chronic post‐surgical pain after 6 months	1	58	Risk Ratio (M‐H, Random, 95% CI)	0.53 [0.27, 1.06]
1.10 Cumulative mean oral morphine requirement in mg ‐ 2 hours	1	60	Mean Difference (IV, Random, 95% CI)	‐13.47 [‐15.49, ‐11.44]
1.11 Cumulative mean oral morphine requirement in mg ‐ 24 hours	23	1341	Mean Difference (IV, Random, 95% CI)	‐33.59 [‐56.37, ‐10.81]
1.12 Cumulative mean oral morphine requirement in mg ‐ 48 hours	4	260	Mean Difference (IV, Random, 95% CI)	‐54.14 [‐69.09, ‐39.20]
1.13 Rates of opioid‐related adverse events (PONV)	21	1111	Risk Ratio (M‐H, Random, 95% CI)	0.46 [0.36, 0.60]
1.14 Rates of opioid‐related adverse events (pruritus)	7	440	Risk Ratio (M‐H, Random, 95% CI)	0.41 [0.26, 0.63]

Comparison 2. Erector spinae plane block vs placebo treatment.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS)	8	499	Mean Difference (IV, Random, 95% CI)	‐0.14 [‐0.29, 0.00]
2.2 Rates of block‐related adverse events	10	592	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
2.3 Mean difference in postoperative pain intensity during activity (24 hours postoperatively) (VAS)	3	199	Mean Difference (IV, Random, 95% CI)	‐0.10 [‐0.19, ‐0.02]
2.4 Mean difference in postoperative pain intensity at rest ‐ 2 hours (VAS)	7	439	Mean Difference (IV, Random, 95% CI)	‐1.46 [‐2.44, ‐0.48]
2.5 Mean difference in postoperative pain intensity at rest ‐ 48 hours (VAS)	2	120	Mean Difference (IV, Random, 95% CI)	‐0.49 [‐0.84, ‐0.14]
2.6 Cumulative mean oral morphine requirement in mg ‐ 24 hours	11	677	Mean Difference (IV, Random, 95% CI)	‐21.05 [‐33.17, ‐8.93]
2.7 Cumulative mean oral morphine requirement in mg ‐ 48 hours	3	195	Mean Difference (IV, Random, 95% CI)	‐26.62 [‐46.41, ‐6.84]
2.8 Rates of opioid‐related adverse events (PONV)	11	695	Risk Ratio (M‐H, Random, 95% CI)	0.39 [0.28, 0.56]
2.9 Rates of opioid‐related adverse events (pruritus)	2	110	Risk Ratio (M‐H, Random, 95% CI)	0.51 [0.21, 1.28]

Comparison 3. Erector spinae plane block vs paravertebral block.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
3.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS)	7	478	Mean Difference (IV, Random, 95% CI)	0.23 [‐0.06, 0.52]
3.2 Rates of block‐related adverse events	7	522	Risk Ratio (M‐H, Random, 95% CI)	0.10 [0.02, 0.55]
3.3 Mean difference in postoperative pain intensity during activity (24 hours postoperatively) (VAS)	6	408	Mean Difference (IV, Random, 95% CI)	0.24 [‐0.06, 0.53]
3.4 Mean difference in postoperative pain intensity at rest ‐ 2 hours (VAS)	3	211	Mean Difference (IV, Random, 95% CI)	0.26 [‐0.24, 0.76]
3.5 Mean difference in postoperative pain intensity at rest ‐ 48 hours	4	298	Mean Difference (IV, Random, 95% CI)	‐0.01 [‐0.14, 0.13]
3.6 Mean difference in postoperative pain intensity during activity ‐ 2 hours	2	141	Mean Difference (IV, Random, 95% CI)	0.23 [‐1.14, 1.60]
3.7 Mean difference in postoperative pain intensity during activity ‐ 48 hours (VAS)	4	298	Mean Difference (IV, Random, 95% CI)	‐0.02 [‐0.31, 0.27]
3.8 Cumulative mean oral morphine requirement in mg ‐ 2 hours	1	90	Mean Difference (IV, Random, 95% CI)	0.30 [‐0.51, 1.11]
3.9 Cumulative mean oral morphine requirement in mg ‐ 24 hours	8	496	Mean Difference (IV, Random, 95% CI)	6.94 [‐2.65, 16.53]
3.10 Cumulative mean oral morphine requirement in mg ‐ 48 hours	5	336	Mean Difference (IV, Random, 95% CI)	9.57 [‐4.33, 23.47]
3.11 Rates of opioid‐related adverse events (PONV)	10	668	Risk Ratio (M‐H, Random, 95% CI)	1.13 [0.81, 1.58]

3.8. Analysis.

Comparison 3: Erector spinae plane block vs paravertebral block, Outcome 8: Cumulative mean oral morphine requirement in mg ‐ 2 hours

Comparison 4. Erector spinae plane block vs transversus abdominis plane block.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
4.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS)	3	160	Mean Difference (IV, Random, 95% CI)	‐0.16 [‐0.46, 0.14]
4.2 Rates of block‐related adverse events	4	202	Risk Ratio (M‐H, Random, 95% CI)	1.00 [0.15, 6.84]
4.3 Mean difference in postoperative pain intensity at rest ‐ 2 hours (VAS)	3	160	Mean Difference (IV, Random, 95% CI)	‐0.33 [‐1.74, 1.08]
4.4 Cumulative mean oral morphine requirement in mg ‐ 24 hours	3	158	Mean Difference (IV, Random, 95% CI)	‐17.31 [‐41.42, 6.79]
4.5 Rates of opioid‐related adverse events (PONV)	3	160	Risk Ratio (M‐H, Random, 95% CI)	0.71 [0.44, 1.16]

Comparison 5. Erector spinae plane block vs serratus anterior plane block.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
5.1 Rates of block‐related adverse events	2	110	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
5.2 Mean difference in postoperative pain intensity at rest ‐ 2 hours (VAS)	1	60	Mean Difference (IV, Random, 95% CI)	‐0.10 [‐0.40, 0.20]
5.3 Cumulative mean oral morphine requirement in mg ‐ 24 hours	2	110	Mean Difference (IV, Random, 95% CI)	‐3.36 [‐17.50, 10.78]
5.4 Rates of opioid‐related adverse events (PONV)	3	170	Risk Ratio (M‐H, Random, 95% CI)	0.70 [0.29, 1.71]
5.5 Rates of opioid‐related adverse events (pruritus)	2	120	Risk Ratio (M‐H, Random, 95% CI)	0.57 [0.19, 1.75]

5.5. Analysis.

Comparison 5: Erector spinae plane block vs serratus anterior plane block, Outcome 5: Rates of opioid‐related adverse events (pruritus)

Comparison 6. Erector spinae plane block vs pectoralis plane block.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
6.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS)	2	98	Mean Difference (IV, Random, 95% CI)	0.24 [‐0.11, 0.58]
6.2 Rates of block‐related adverse outcomes	1	47	Risk Ratio (M‐H, Random, 95% CI)	2.88 [0.12, 67.29]
6.3 Mean difference in postoperative pain intensity at rest ‐ 2 hours (VAS)	3	162	Mean Difference (IV, Random, 95% CI)	0.21 [‐0.16, 0.58]
6.4 Cumulative mean oral morphine requirement in mg ‐ 24 hours	3	145	Mean Difference (IV, Random, 95% CI)	14.08 [4.21, 23.94]

Comparison 7. Erector spinae plane block vs quadratus lumborum block.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
7.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS)	1	40	Mean Difference (IV, Random, 95% CI)	0.05 [‐0.21, 0.31]
7.2 Rates of block‐related adverse events	1	40	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
7.3 Mean difference in postoperative pain intensity during activity (24 hours postoperatively)	1	80	Mean Difference (IV, Random, 95% CI)	0.00 [‐0.65, 0.65]
7.4 Cumulative mean oral morphine requirement in mg ‐ 24 hours	2	120	Mean Difference (IV, Random, 95% CI)	‐0.31 [‐2.22, 1.61]
7.5 Rates of opioid‐related adverse events (PONV)	2	120	Risk Ratio (M‐H, Random, 95% CI)	1.00 [0.35, 2.84]

Comparison 8. Erector spinae plane block vs intercostal nerve block.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
8.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS)	2	131	Mean Difference (IV, Random, 95% CI)	‐0.33 [‐3.02, 2.35]
8.2 Rates of block‐related adverse events	3	181	Risk Ratio (M‐H, Random, 95% CI)	0.09 [0.01, 1.56]
8.3 Mean difference in postoperative pain intensity during activity (24 hours postoperatively) (VAS)	2	131	Mean Difference (IV, Random, 95% CI)	‐1.52 [‐4.61, 1.56]
8.4 Mean difference in postoperative pain intensity at rest ‐ 48 hours (VAS)	1	60	Mean Difference (IV, Random, 95% CI)	‐1.37 [‐2.47, ‐0.27]
8.5 Mean difference in postoperative pain intensity during activity ‐ 48 hours (VAS)	1	60	Mean Difference (IV, Random, 95% CI)	‐1.73 [‐3.02, ‐0.44]
8.6 Cumulative mean oral morphine requirement in mg ‐ 24 hours	2	121	Mean Difference (IV, Random, 95% CI)	9.02 [3.05, 14.99]
8.7 Cumulative mean oral morphine requirement in mg ‐ 48 hours	2	121	Mean Difference (IV, Random, 95% CI)	7.98 [2.60, 13.36]
8.8 Rates of opioid‐related adverse events (PONV)	2	121	Risk Ratio (M‐H, Random, 95% CI)	1.26 [0.40, 3.98]

8.4. Analysis.

Comparison 8: Erector spinae plane block vs intercostal nerve block, Outcome 4: Mean difference in postoperative pain intensity at rest ‐ 48 hours (VAS)

8.5. Analysis.

Comparison 8: Erector spinae plane block vs intercostal nerve block, Outcome 5: Mean difference in postoperative pain intensity during activity ‐ 48 hours (VAS)

Comparison 9. Erector spinae plane block vs epidural analgesia.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
9.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS)	2	81	Mean Difference (IV, Random, 95% CI)	1.20 [‐2.52, 4.93]
9.2 Rates of block‐related adverse events	1	50	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
9.3 Mean difference in postoperative pain intensity during activity (24 hours postoperatively) (VAS)	1	50	Mean Difference (IV, Random, 95% CI)	‐0.72 [‐1.14, ‐0.30]
9.4 Mean difference in postoperative pain intensity at rest ‐ 2 hours (VAS)	1	31	Mean Difference (IV, Random, 95% CI)	3.00 [1.52, 4.48]
9.5 Mean difference in postoperative pain intensity at rest ‐ 48 hours (VAS)	1	50	Mean Difference (IV, Random, 95% CI)	‐1.20 [‐1.78, ‐0.62]
9.6 Mean difference in postoperative pain intensity during activity ‐ 48 hours (VAS)	1	50	Mean Difference (IV, Random, 95% CI)	‐1.36 [‐1.96, ‐0.76]
9.7 Cumulative mean oral morphine requirement in mg ‐ 24 hours	1	31	Mean Difference (IV, Random, 95% CI)	1.75 [‐1.44, 4.94]

9.2. Analysis.

Comparison 9: Erector spinae plane block vs epidural analgesia, Outcome 2: Rates of block‐related adverse events

9.3. Analysis.

Comparison 9: Erector spinae plane block vs epidural analgesia, Outcome 3: Mean difference in postoperative pain intensity during activity (24 hours postoperatively) (VAS)

9.4. Analysis.

Comparison 9: Erector spinae plane block vs epidural analgesia, Outcome 4: Mean difference in postoperative pain intensity at rest ‐ 2 hours (VAS)

9.5. Analysis.

Comparison 9: Erector spinae plane block vs epidural analgesia, Outcome 5: Mean difference in postoperative pain intensity at rest ‐ 48 hours (VAS)

9.6. Analysis.

Comparison 9: Erector spinae plane block vs epidural analgesia, Outcome 6: Mean difference in postoperative pain intensity during activity ‐ 48 hours (VAS)

9.7. Analysis.

Comparison 9: Erector spinae plane block vs epidural analgesia, Outcome 7: Cumulative mean oral morphine requirement in mg ‐ 24 hours

Comparison 10. Subgroup analysis: erector spinae plane block vs no block.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
10.1 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS) ‐ Type of surgery	18	1038	Mean Difference (IV, Random, 95% CI)	‐0.79 [‐1.07, ‐0.50]
10.1.1 Abdominal surgery	4	225	Mean Difference (IV, Random, 95% CI)	‐1.09 [‐1.26, ‐0.92]
10.1.2 Breast surgery	3	148	Mean Difference (IV, Random, 95% CI)	0.00 [‐0.53, 0.53]
10.1.3 Thoracic surgery	6	385	Mean Difference (IV, Random, 95% CI)	‐1.20 [‐1.62, ‐0.79]
10.1.4 Spine surgery	4	240	Mean Difference (IV, Random, 95% CI)	‐0.66 [‐0.88, ‐0.45]
10.1.5 Hip surgery	1	40	Mean Difference (IV, Random, 95% CI)	0.20 [‐0.04, 0.44]
10.2 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS) ‐ Type of local anaesthetic	18	1038	Mean Difference (IV, Random, 95% CI)	‐0.79 [‐1.07, ‐0.50]
10.2.1 Bupivacaine 0.25% to 0.5%	8	479	Mean Difference (IV, Random, 95% CI)	‐0.67 [‐1.13, ‐0.21]
10.2.2 Ropivacaine 0.3% to 0.5%	10	559	Mean Difference (IV, Random, 95% CI)	‐0.93 [‐1.22, ‐0.64]
10.3 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS) ‐ Volume of local anaesthetic	18	1038	Mean Difference (IV, Random, 95% CI)	‐0.79 [‐1.07, ‐0.50]
10.3.1 15 mL to 19 mL	1	44	Mean Difference (IV, Random, 95% CI)	‐1.00 [‐1.98, ‐0.02]
10.3.2 20 mL to 24 mL	11	646	Mean Difference (IV, Random, 95% CI)	‐0.70 [‐1.06, ‐0.33]
10.3.3 25 mL to 30 mL	6	348	Mean Difference (IV, Random, 95% CI)	‐0.94 [‐1.46, ‐0.42]
10.4 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS) ‐ Fixed multimodal analgesia	18	1038	Mean Difference (IV, Random, 95% CI)	‐0.79 [‐1.07, ‐0.50]
10.4.1 Non‐opioid + PCA	8	499	Mean Difference (IV, Random, 95% CI)	‐0.70 [‐1.18, ‐0.23]
10.4.2 Non‐opioid + rescue opioids	7	379	Mean Difference (IV, Random, 95% CI)	‐0.62 [‐1.01, ‐0.23]
10.4.3 No specified multimodal analgesia	3	160	Mean Difference (IV, Random, 95% CI)	‐1.47 [‐1.99, ‐0.94]
10.5 Mean difference in postoperative pain intensity at rest (24 hours postoperatively) (VAS) ‐ Baseline non‐opioid analgesia	18	1038	Mean Difference (IV, Random, 95% CI)	‐0.79 [‐1.07, ‐0.50]
10.5.1 Paracetamol	5	262	Mean Difference (IV, Random, 95% CI)	‐0.65 [‐1.12, ‐0.18]
10.5.2 NSAID	5	280	Mean Difference (IV, Random, 95% CI)	‐1.12 [‐1.94, ‐0.31]
10.5.3 Not specified	4	265	Mean Difference (IV, Random, 95% CI)	‐0.77 [‐1.18, ‐0.36]
10.5.4 Both paracetamol + NSAID	4	231	Mean Difference (IV, Random, 95% CI)	‐0.58 [‐0.93, ‐0.24]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Abdelhamid 2020.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained Grouping was determined through a computer‐generated list that was kept in a sealed envelope. Both the patients and the anaesthetists involved in postoperative data collection were blinded to the treatment group to which each patient belonged.
Participants	Setting: Cairo University hospitals Sample size: 66 adult patients Participant (baseline) characteristics: Age 18 to 59 years ASA II‐III BMI > 40 kg/m2 Inclusion criteria: adult patients scheduled for elective sleeve gastrectomy Exclusion criteria: presence of coagulopathy or thrombocytopenia with a count less than 100,000/mL, hepatic or renal disease, opioid addiction, sepsis (due to the increased risk of meningitis, infection at the puncture site, pre‐existing neurologic disease, history of allergy to local anaesthetic)
Interventions	Intervention group: ESPB Comparator groups: TAPB and no block
Outcomes	Primaryoutcome: pain assessment by VAS at 30 minutes, as well as at 2, 4, 6, 8, 12, 18 and 24 hours postoperatively Secondary outcomes: total intraoperative fentanyl requirements, intraoperative heart rate and mean arterial blood pressure, duration of anaesthesia (time from induction of GA to extubation), incidence of complications, including nerve injury, haematoma formation, LA toxicity and intravascular injections, cumulative consumption of pethidine during the first 24 hours postoperatively, duration of time before the first request for rescue analgesia post‐surgically
Notes	—

Abu 2019.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained Computer‐generated randomised numbers were used to randomly assign patients into 2 groups using sealed opaque envelopes that were randomly selected by each patient and contained a group number in which the patient was enrolled.
Participants	Setting: Faculty of Medicine, Tanta University, Tanta, Egypt Sample size: 60 adult patients Participant (baseline) characteristics: Age 18 to 65 years ASA I‐II Inclusion criteria: patients scheduled for open epigastric hernia repair Exclusion criteria: contraindication to regional anaesthesia (e.g. patient refusal, coagulation abnormality or local infection), allergy to local anaesthetics, uncooperative or psychiatric patients, and hepatic or renal dysfunction
Interventions	Intervention group: ESPB Comparator group: placebo block
Outcomes	Primaryoutcome: pain severity measured by VAS at 2 hours postoperatively Secondary outcomes: pain at rest up to 24 hours postoperatively (VAS), consumption of intraoperative fentanyl, first 24 hours postoperative rescue analgesia
Notes	—

Aksu 2019a.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained A sequentially numbered, opaque, sealed envelope technique was used for randomisation and patients were randomised to either ESPB or control groups (no block).
Participants	Setting: Kocaeli University, Kocaeli, Turkey Sample size: 50 female patients Participant (baseline) characteristics: Age 25 to 70 years ASA I‐II Inclusion criteria: patients scheduled to undergo elective breast cancer surgery Exclusion criteria: allergies to any of the study drugs, infection of the skin at the site of needle puncture area, coagulopathy, obesity (BMI > 35 kg/m2) and chronic use of opioids were excluded from the study
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primary outcome: postoperative morphine consumption recorded at 1, 6, 12 and 24 hours postoperatively Secondary outcomes: opioid‐related side effects, postoperative nausea and vomiting (PONV) and NRS pain scores
Notes	—

Aksu 2019b.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained Randomisation was performed according to computer‐generated random number tables, and allocation to treatment group was done using the sealed, opaque envelope technique.
Participants	Setting: Kocaeli University, Kocaeli, Turkey Sample size: 46 adult patients Participant (baseline) characteristics: Age 20 to 75 years ASA I‐II Inclusion criteria: patients scheduled for elective laparoscopic cholecystectomy Exclusion criteria: ASA physical status III‐IV, obesity (BMI > 35 kg/m2), infection of the skin at the site of needle puncture, allergies to any of the study drugs, coagulopathy and recent opioid use
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primary outcome: cumulative morphine consumption within the first 24 hours after surgery Secondary outcomes: postoperative NRS scores, opioid‐related side effects, nausea and vomiting
Notes	—

Altıparmak 2019a.

Study characteristics
Methods	Prospective, single‐blinded, randomised controlled trial Ethics approval obtained Patients were randomly allocated into 2 groups based on a computerised randomisation table created by a researcher who was not involved in the study. For each randomised patient, the OR anaesthesiologist took the corresponding sealed envelope from a folder, which indicated the treatment to be assigned to the patient.
Participants	Setting: Muğla Sıtkı Koçman University, Department of Anesthesiology and Reanimation, Muğla, Turkey Sample size: 76 adult patients Participant (baseline) characteristics: Age 18 to 70 years ASA I‐II Inclusion criteria: patients scheduled for an elective laparoscopic cholecystectomy surgery Exclusion criteria: coagulation disorder, infection at the injection site, allergy to local anaesthetics, advanced hepatic or renal failure, chronic opioid consumption and BMI ≥ 35 kg/m2
Interventions	Intervention group: ESPB Comparator group: TAPB
Outcomes	Primary outcome: tramadol consumption 24 hours postoperative Secondary outcomes: intraoperative fentanyl need, NRS scores at different time point, total rescue analgesic consumption in the first 24 postoperative hours and the incidence of complications.
Notes	—

Altıparmak 2019b.

Study characteristics
Methods	Prospective, single‐blinded, randomised controlled trial Ethics approval obtained After endotracheal intubation, patients were randomly allocated into 2 groups based on a computerised randomisation table created by a researcher who was not involved in the study. The researcher assigned the random ID to each patient and the blinded anaesthesiologist (MKT) used this ID while collecting the postoperative data in the surgical ward.
Participants	Setting: Muğla Sıtkı Koçman University, Department of Anesthesiology and Reanimation, Muğla, Turkey Sample size: 40 adult patients Participant (baseline) characteristics: Age 18 to 70 years ASA I‐II Inclusion criteria: adult patients scheduled for an elective unilateral modified radical mastectomy surgery with axillary lymph node dissection Exclusion criteria: coagulation disorders, allergy to study drugs, obesity (BMI > 35 kg/m2), infection at the injection site, chronic opioid consumption and an inability to use a patient‐controlled analgesia (PCA) device
Interventions	Intervention group: ESPB Comparator group: PECS block
Outcomes	Primaryoutcome: total postoperative tramadol consumption in the first 24 hours Secondary outcomes: NRS scores at 7 different time points (1, 2, 6, 12 and 24 hours), intraoperative fentanyl need
Notes	—

Anushree 2020.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval not documented
Participants	Setting: Manipal Hospital, Bangalore Sample size: 42 adult patients Participant (baseline) characteristics: Age over 18 years ASA I‐II Inclusion criteria: adults scheduled for laparoscopic living donor nephrectomy Exclusion criteria: —
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: mean fentanyl consumption was recorded for 24 hours postoperatively Secondary outcomes: NRS pain scores, sedation, incidence of nausea and vomiting, requirement for additional analgesia and patient satisfaction were recorded
Notes	Only conference abstract available. No email address available to request further information.

Arora 2019.

Study characteristics
Methods	Randomised, double‐blinded, controlled trial Trial registered retrospectively Ethics approval obtained Computer‐generated randomisation, sequentially numbered, sealed, opaque envelopes
Participants	Setting: Nehru Hospital, Chandigarh, India Sample size: 60 women Participant (baseline) characteristics: Age 18 to 60 years ASA I‐II Inclusion criteria: women undergoing elective breast cancer surgery under general anaesthesia Exclusion criteria: infection at the block site, coagulopathy, morbid obesity (BMI > 40), allergy to local anaesthetics, uncontrolled hypertension or ischaemic heart disease, renal dysfunction, pre‐existing neurological deficits, psychiatric illness
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: 24‐hour morphine consumption Secondary outcomes: time to first rescue analgesia, pain scores, adverse effects
Notes	Only abstract and study protocol available. No email address available to request further information.

Aygun 2020.

Study characteristics
Methods	Prospective, randomised, comparative, double‐blinded, controlled trial Ethics approval obtained Patients were randomised in the operating room using sequentially numbered, opaque, sealed envelopes to ESPB and QLB‐II groups. Randomisation was overseen by a single author (ASP) with each patient assigned a study number.
Participants	Setting: Cigli Regional Training Hospital, Izmir, Turkey Sample size: 80 patients Participant (baseline) characteristics: Age 18 to 70 years ASA I‐II Inclusion criteria: patients scheduled to undergo laparoscopic cholecystectomy Exclusion criteria: patients not consenting, allergies to local anaesthetics, bleeding diathesis or history of anticoagulant use, history of liver or renal pathology effecting drug elimination, psychiatric disease, use of medicines such as gabapentin/pregabalin that could affect pain perception and those unable to operate PCA
Interventions	Intervention group: ESPB Comparator group: QLB
Outcomes	Primaryoutcome: first 24 hours opioid consumption Secondary outcomes: NRS scores during rest and coughing/movement, ondansetron use for nausea/vomiting within the first 24 hours
Notes	—

Calia 2019.

Study characteristics
Methods	Randomised controlled trial. The study protocol was approved by the local review board. Ethics approval was not reported
Participants	Setting: inpatient hospital and operating rooms Sample size: 29 patients Participant (baseline) characteristics: Age over 18 years Inclusion criteria: patients undergoing open lumbar decompression surgery Exclusion criteria: —
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: postoperative NRS score at various time periods during the first 24 hours following surgery Secondary outcomes: opioid consumption, rescue analgesia, opioid‐related side effects
Notes	Only conference abstract available ‐ requested full text from authors (primary results)

Chen 2020.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained Patients were allocated randomly to PVB group, ICNB group and ESPB group according to a computer‐generated random number table. Allocation to treatment group was performed using the sealed, opaque envelope technique. All patients and an investigator who was responsible for follow‐up during the 48 postoperative hours were blinded to the randomisation groups.
Participants	Setting: Tongji Hospital, Huazhong University of Science and Technology, Wuhan, China Sample size: 75 adults patients Participant (baseline) characteristics: Age 18 to 75 years ASA I–II Inclusion criteria: patients undergoing elective thoracoscopic partial pulmonary resection surgery Exclusion criteria: allergy to local anaesthetics, coagulation disorders, severe heart diseases, hepatic or renal insufficiency, a history of chronic pain or chronic opioid use and psychiatric disease
Interventions	Intervention group: ESPB Comparator groups: PVB or ICNB
Outcomes	Primaryoutcome: cumulative morphine consumption at 24 hours postoperatively Secondary outcomes: VAS pain scores at rest and while coughing at 0, 2, 4, 8, 24 and 48 hours postoperatively, cumulative morphine consumption at other observed times and rescue analgesia requirement
Notes	—

Ciftci 2020a.

Study characteristics
Methods	Prospective, randomised, controlled, double‐blind pilot trial Ethics approval obtained Participants were randomly assigned to the groups following simple randomisation procedures (computerised random numbers). The study drugs were prepared by an anaesthesiologist who did not participate in the study. The anaesthesiologist who performed the ESPB or sham block did not know whether the drug was a local anaesthetic or serum saline. The postoperative pain scores were evaluated by a nurse anaesthetist who was blinded to the study.
Participants	Setting: Medipol University Hospital Sample size: 60 patients Participant (baseline) characteristics: Age 18 to 65 years ASA I or II Inclusion criteria: patients undergoing unilateral arthroscopic shoulder surgery under general anaesthesia Exclusion criteria: bleeding diathesis or anticoagulant treatment, pregnancy or lactation, allergy to the study drugs (local anaesthetic or opioids), local tissue infections in the block procedure area, patient refusal
Interventions	Intervention group: ESPB Comparator group: placebo block
Outcomes	Primaryoutcome: comparison of perioperative and postoperative (48 hours) opioid consumption Secondary outcomes: postoperative pain scores, rescue analgesia needed, adverse events related to opioid consumption (itching, nausea, vomiting etc.)
Notes	—

Ciftci 2020b.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained A computer program was used to randomise the participants into 3 groups. Each group (mTLIP block, ESPB and control) was composed of 30 patients.
Participants	Setting: Medipol University Hospital Sample size: 90 patients Participant (baseline) characteristics: Age 18 to 65 years ASA I‐II Inclusion criteria: patients undergoing single‐level lumbar discectomy and hemilaminectomy surgery under general anaesthesia Exclusion criteria: bleeding diathesis, receiving anticoagulant treatment, allergies to medications used in the study (local anaesthetics, nonsteroidal anti‐inflammatory drugs and opioids), skin infections at the site of block area, pregnancy or lactation, and refusal to participate in the study
Interventions	Intervention group: ESPB Comparator groups: no block and mTLIP block
Outcomes	Primaryoutcome: fentanyl consumption Secondary outcomes: VAS pain scores, rescue analgesia, block procedure time and side effects
Notes	—

Ciftci 2020c.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained The patients were divided into groups using a randomising computer program. A pain nurse anaesthetist, who was blinded to the study, evaluated and recorded the postoperative pain and opioid consumption.
Participants	Setting: Medipol University Hospital Sample size: 90 adult patients Participant (baseline) characteristics: Age 18 to 65 ASA I‐II Inclusion criteria: patients scheduled for VATS lobectomies/wedge resections under general anaesthesia Exclusion criteria: bleeding diathesis, pregnancy or breastfeeding, receiving anticoagulant treatment, a known history of allergy to the study drugs (local anaesthetic or opioids), infections at the region of the blocks and refusal to take part in the procedure
Interventions	Intervention group: ESPB Comparator groups: no block and PVB
Outcomes	Primaryoutcome: perioperative and postoperative (48 h) opioid consumption Secondary outcomes: postoperative pain scores using VAS, block performance time, success of one‐time puncture, rescue analgesic usage and adverse events related to opioid consumption (itching, nausea and vomiting etc.)
Notes	—

Ciftci 2020d.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained The patients were divided into 2 groups of 30 patients each using a randomising computer program
Participants	Setting: Medipol University Hospital Sample size: 60 patients Participant (baseline) characteristics: Age 18 to 65 years ASA I‐II Inclusion criteria: patients diagnosed with primary lung carcinomas who were undergoing VATS lobectomies under general anaesthesia Exclusion criteria: bleeding diathesis histories, patients receiving anticoagulant treatment, patients with known local anaesthetic or opioid allergies, patients with skin infections at the needle puncture site, pregnant or lactating patients, and those who did not agree to the procedure
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: postoperative opioid consumption Secondary outcomes: postoperative VAS pain scores, adverse events related to the opioid treatments (allergic reactions, nausea and vomiting)
Notes	—

Ekinci 2020.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained Sealed envelopes were prepared for this randomisation technique before the beginning of the trial. On arrival in the preoperative regional anaesthesia room, the investigators randomly selected a sealed envelope. Then the procedure was explained to the patient (whether ESPB or SAPB was performed).
Participants	Setting: Medipol Mega Hospital Complex Sample size: 60 patients were included in this study (30 in each group) Participant (baseline) characteristics: Age 18 to 65 years ASA I‐II Inclusion criteria: patients undergoing unilateral thoracoscopic lobectomies/wedge resection under general anaesthesia Exclusion criteria: bleeding diathesis, anticoagulant treatment, pregnancy, lactation, known allergies to the drugs used in the study (i.e. local anaesthetics or opioids), or local tissue infections at the block procedure area, patient refusal
Interventions	Intervention group: ESPB Comparator group: SAPB
Outcomes	Primaryoutcome: comparison of perioperative and postoperative (48 hours) opioid consumption Secondary outcomes: evaluation of postoperative pain scores as measured by VAS, block performance time, one‐time puncture success, rescue analgesic usage, adverse events related to opioid consumption (e.g. itching, nausea, vomiting, etc.)
Notes	—

El Ghamry 2019a.

Study characteristics
Methods	Prospective, double‐blinded randomised controlled trial Ethics approval obtained Patients were randomly assigned to 2 groups (35 each) with a 1:1 allocation ratio using computer‐generated random numbers. The numbers were concealed in sealed, opaque envelopes. A blinded nurse (not participating in the study or data collection) used the random numbers to assign patients to their groups. Group I included those to receive TPVB and group II were to receive ESPB.
Participants	Setting: Tanta University, El Geish Street, Tanta, Egypt Sample size: 70 adult female patients Participant (baseline) characteristics: Age 20 to 60 years ASA I‐II Inclusion criteria: adult female patients scheduled for unilateral modified radical mastectomy under general anaesthesia Exclusion criteria: severe respiratory or cardiac disorders, hepatic or renal insufficiency, coagulopathy, local infection at the injection site, spine or chest wall deformity, allergy to any of the study drugs, opioid addiction or severe obesity (BMI > 35 kg/m2). Uncooperative patients and those who could not express pain intensity by VAS were also excluded.
Interventions	Intervention group: ESPB Comparator group: PVB
Outcomes	Primaryoutcome: postoperative 24‐hour morphine consumption Secondary outcomes: intraoperative fentanyl consumption, time of the first request for analgesia, postoperative VAS at 2, 4, 6, 8, 12, 18 and 24 h, heart rate (HR) and mean blood pressure (MBP)
Notes	Authors replied with some further data

El Ghamry 2019b.

Study characteristics
Methods	Prospective, double‐blinded, randomised, controlled study Ethics approval obtained Patients were randomised into 2 equal groups (30 patients in each) using computer‐generated random numbers concealed in sealed envelopes indicating the group of the assignment
Participants	Setting: Tanta University Hospital Sample size: 60 adult patients Participant (baseline) characteristics: Age 18 to 60 years ASA I or II Inclusion criteria: patients scheduled for L3‐L5 posterior lumbar interbody fusion (PLIF) under general anaesthesia (GA) Exclusion criteria: patient refusal, BMI > 35 kg/m2, infection at the site of the needle puncture, allergy to any of the study drugs, coagulopathy, and uncooperative patients or who could not express pain via a VAS
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primary outcome: postoperative morphine consumption Secondary outcomes: intraoperative fentanyl consumption, time to first analgesic request, static and dynamic VAS score, haemodynamic changes, length of PACU stay and complications
Notes	—

Elsabeeny 2020.

Study characteristics
Methods	Prospective, randomised, parallel‐group, double‐blinded, controlled trial Ethics approval obtained A computer system‐generated randomisation was used and patients were randomly allocated by the anaesthesia resident using the closed envelope technique into one of 3 study groups
Participants	Setting: National Cancer Institute, Cairo University, Cairo, Egypt Sample size: 75 adult female patients Participant (baseline) characteristics: Age 18 to 65 years ASA I‐II Inclusion criteria: adult female patients with breast cancer and scheduled for modified radical mastectomy (MRM) Exclusion criteria: patient refusal, local infection at site of injection, coagulation defects, thrombocytopenia, allergy to any of the used drugs, impaired hepatic or renal function and patients with any chronic pain syndrome receiving chronic pain medications
Interventions	Intervention group: ESPB Comparator groups: SAPB and no block
Outcomes	Primary outcomes: postoperative VAS scores, total postoperative morphine consumption and first time to receive morphine Secondary outcomes: intraoperative and postoperative haemodynamics
Notes	—

Eskin 2020.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained A sealed, opaque envelope containing allocated randomisation was opened in the operating room after induction of GA. The patients were randomly allocated in a 1:1:1 ratio to one of 3 groups: erector spinae plane block (ESPB, n = 40), mid‐transverse to pleura block (MTPB, n = 40) and control n = 40).
Participants	Setting: University of Health Sciences, Gulhane Training and Research Hospital Sample size: 120 patients Participant (baseline) characteristics: Age 18 to 80 years ASA I‐III Inclusion criteria: patients scheduled for elective lumbar decompression surgery for one or two vertebral levels under general anaesthesia Exclusion criteria: patient refusal, pregnancy, history of allergy to study drugs, neurological and cognitive disorders, coagulopathy, chronic pain disorders and infections at the injection site
Interventions	Intervention group: ESPB Comparator groups: MTPB and no block
Outcomes	Primaryoutcome: mean pain scores Secondary outcomes: consumption of opioid rescue analgesia and amount of analgesic delivered by patient‐controlled analgesia (PCA) in the postoperative period
Notes	—

Fang 2019.

Study characteristics
Methods	Single‐centre, randomised, double‐blind, parallel, 2‐arm controlled study Ethics approval obtained Eligible patients were informed about the study and their written informed consents were received
Participants	Setting: Shanghai General Hospital Sample size: 94 adult patients Participant (baseline) characteristics: Age 18 to 81 years ASA I‐II Inclusion criteria: patients scheduled for the wedge or segmental resection of the lung, or pulmonary lobectomy by posterolateral thoracotomy Exclusion criteria: BMI ≥ 30 or ≤ 18 kg/m2, infection of the skin at the site of the needle puncture, allergies to any of the study drugs, pre‐existing pain syndromes, pregnancy, severe hepatic disorders, renal disorders or cardiovascular disorders (New York Heart Association functional class greater than II)
Interventions	Intervention group: ESPB Comparator group: PVB
Outcomes	Primaryoutcome: VAS pain scores at rest and when coughing at 1, 6, 12 and 24 hours postoperatively Secondary outcomes: total press times of PCA, adverse effects, puncture time and success rate of one puncture
Notes	—

Finnerty 2020.

Study characteristics
Methods	Preregistered randomised controlled trial Ethics approval obtained Patients were assigned to one of the trial groups using a computer‐generated random number table. Numbers ending in an even number integer were designated ESPB patients, and those ending in odd numbers as SAPB patients. The patient study code number and group allocation were typed on separate pages, folded and concealed in sequentially numbered, sealed envelopes. Block randomisation in groups of 6 individuals was applied to ensure an even number in each group as the study progressed. The groups were named ‘ESP’ (erector spinae plane) and ‘SAP’ (serratus anterior plane). A randomisation key was held by an independent third party.
Participants	Setting: Mater Misericordiae University Hospital, Dublin, Ireland Sample size: 60 adult patients Participant (baseline) characteristics: Age 18 to 80 years ASA I‐III Inclusion criteria: adult patients undergoing minimally invasive thoracic surgery under general anaesthesia Exclusion criteria: pre‐existing infection at block site, contraindication to regional anaesthesia, history of opiate abuse, and pre‐existing chronic pain or cognitive dysfunction, which would impede accurate engagement with postoperative quality of recovery and analgesia assessment
Interventions	Intervention group: ESPB Comparator group: SAPB
Outcomes	Primaryoutcome: quality of patient recovery at 24 hours, using the Quality of Recovery‐15 (QoR‐15) scale Secondary outcomes: area under the curve (AUC) of pain verbal rating scale (VRS) over time, time to first opioid analgesia, postoperative 24‐hour opioid consumption, in‐hospital comprehensive complication index (CCI) score, hospital stay
Notes	—

Fiorelli 2020.

Study characteristics
Methods	Prospective, randomised, single‐centre, single‐blinded, controlled study Ethics approval obtained From July to October 2019, 60 consecutive patients scheduled for elective lobectomy or wedge resection via mini‐thoracotomy were included in the study
Participants	Setting: Sant' Andrea Hospital, Rome, Italy Sample size: 60 consecutive adult patients Participant (baseline) characteristics: Age over 18 years ASA I‐III Inclusion criteria: adult patients scheduled to undergo mini‐thoracotomy for lung resection in general anaesthesia Exclusion criteria: emergency surgery, pregnancy, patient refusal or inability to give consent, allergy or contraindication to the use of any of the study drugs, age ≤ 18 years, ASA ≥ IV, intraoperative rib fractures, and preoperative chronic or neuropathic pain
Interventions	Intervention group: ESPB Comparator group: ICNB
Outcomes	Primaryoutcome: intensity of postoperative pain at rest, assessed with a NRS Secondary outcomes: dynamic NRS values (during cough), perioperative analgesic requirements, patient satisfaction, on the basis of a verbal scale (Likert scale), respiratory muscle strength, considering the maximum inspiratory pressure (MIP) and maximum expiratory pressure (MEP) variation from baseline
Notes	—

Fu 2020.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained All patients were randomised to receive ultrasound‐guided ESPB (Group A, n = 30) or no intervention (Group B, n = 30)
Participants	Setting: Linyi Central Hospital Sample size: 60 adult patients Participant (baseline) characteristics: Age 20 to 65 years ASA I‐II Inclusion criteria: adult patients undergoing partial hepatectomy Exclusion criteria: infection of the skin at the site of the needle puncture, coagulopathy, morbid obesity, hypersensitivity to local anaesthetic drugs, serious heart and lung diseases, liver and kidney dysfunction, or mental diseases
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: postoperative pain determined by VAS scores on movement and at rest Secondary outcomes: time to initial anal exhaust, analgesic consumption, side effects, complications associated with postoperative pain management
Notes	—

Gaballah 2019.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained Patients were randomised using a computerised randomisation schedule with allocations sealed in opaque envelopes
Participants	Setting: Menoufia university hospitals Sample size: 60 adult patients Participant (baseline) characteristics: Age over 18 years ASA I‐II Inclusion criteria: adult scheduled to undergo unilateral VATS Exclusion criteria: severe respiratory disease, pre‐existing neurologic deficit‚ local infection, history of opioid dependence, or allergy/contraindication to the use of any of the study drugs. Patients who required thoracotomy or postoperative mechanical ventilation and patients who had previous VATS and other thoracic surgeries were also excluded.
Interventions	Intervention group: ESPB Comparator group: SPB
Outcomes	Primaryoutcome: pain severity, time of first postoperative analgesia requirement, intraoperative and postoperative analgesic requirements Secondary outcomes: perioperative mean arterial pressure (MAP), heart rate (HR), respiratory rate, arterial blood gases (ABGs)
Notes	—

Gad 2019.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained Patients were assigned randomly by a computer‐generated assignment into 2 groups: ESPB group (n = 24) and PECS block group (n = 23)
Participants	Setting: Faculty of Medicine, Mansoura University, Mansoura, Egypt Sample size: 50 patients Participant (baseline) characteristics: Age 18 to 65 years ASA I and II Inclusion criteria: Patients undergoing unilateral modified radical mastectomy surgeries including axillary lymph node dissection Exclusion criteria: known allergy to the study drugs, coagulation disorders, obesity (BMI > 35 kg/m2), injection site infection or opioid abuse
Interventions	Intervention group: ESPB Comparator group: PECS block
Outcomes	Primaryoutcome: total morphine consumption in the first 24 hours postoperatively, levels of stress hormones Secondary outcomes: VAS score, number of patients who requested analgesic, intraoperative fentanyl needed, postoperative complications
Notes	—

Gultekin 2020.

Study characteristics
Methods	Single‐centre randomised controlled trial Ethics approval obtained Randomisation was performed using a computer‐based list before the operation. Patients were randomly assigned to the ESPB (n = 30) and CA (n = 30) groups
Participants	Setting: Mengucek Gazi Education and Research Hospital, Erzincan Binali Yıldırım University, Erzincan, Turkey Sample size: 60 patients Participant (baseline) characteristics: Age older than 18 years > 2 cm kidney stones Inclusion criteria: patients undergoing percutaneous nephrolithotomy for the treatment of renal stones Exclusion criteria: BMI of > 30, contraindication for ESPB, substance abuse, allergy to local anaesthetics, cardiovascular, respiratory, neurological or metabolic diseases, severe coagulopathy, complex stones requiring more than 2 renal access punctures, musculoskeletal deformities, low cognitive function that would make VAS score difficult to assess, missing data and patient withdrawal from the study
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: pain score Secondary outcomes: time to first rescue analgesic, rates of tramadol and paracetamol use, length of hospital stay, postoperative stone‐free rates
Notes	—

Guo 2019.

Study characteristics
Methods	Randomised controlled trial Ethics approval obtained
Participants	Setting: Hospital of Southern Medical University, Shenzhen, China Sample size: 40 patients Participant (baseline) characteristics: Age 60 to 75 years ASA I‐II Inclusion criteria: patients undergoing elective thoracoscopic surgery Exclusion criteria: obesity (BMI > 35 kg/m2), skin infection at the puncture site, any known allergies to study drugs and recent opioid use
Interventions	Intervention group: ESPB Comparator group: PVB
Outcomes	Primaryoutcome: intraoperative remifentanil and postoperative sufentanil doses Secondary outcomes: numerical rating scale (NRS) score at 1, 6, 12 and 24 postoperative hours, intraoperative hypotension, postoperative nausea and vomiting, operation time
Notes	Translated from Chinese using Google translator

Gürkan 2018.

Study characteristics
Methods	Prospective, randomised, controlled, single‐blinded study Ethics approval obtained Randomisation was achieved using the sequentially numbered, opaque, sealed envelope technique. A pain nurse blinded to the study groups was responsible for postoperative follow‐up.
Participants	Setting: School of Medicine, Kocaeli University, Kocaeli, Turkey Sample size: 50 female patients Participant (baseline) characteristics: Age 20 to 65 years ASA I‐II Inclusion criteria: adult female patients scheduled for elective surgery for breast cancer. Only unilateral surgical procedures were included. Exclusion criteria: obesity (BMI > 35 kg/m2), infection of the skin at the site of the needle puncture, known allergies to any of the study drugs, coagulopathy and recent use of opioid drugs
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: mean morphine consumption Secondary outcomes: numeric rating scale (NRS) pain scores recorded at 1, 6, 12 and 24 hours postoperatively
Notes	—

Gürkan 2020.

Study characteristics
Methods	Double‐blinded, prospective, randomised controlled study Ethics approval obtained Randomisation was performed according to computer‐generated random number tables, and allocation to treatment group was done using the sealed, opaque envelope technique
Participants	Setting: Kocaeli University Hospital, Kocaeli, Turkey Sample size: 75 female patients Participant (baseline) characteristics: Age 18 to 65 years ASA I–II Inclusion criteria: adult female patients scheduled for elective breast cancer surgery Exclusion criteria: obesity (BMI > 35 kg/m2), infection of the skin at the site of needle puncture area, patients with known allergies to any of the study drugs, coagulopathy, recent use of opioid drugs, inability to comprehend or use the numeric rating pain scoring system or patient‐controlled analgesia (PCA) pump
Interventions	Intervention group: ESPB Comparator groups: no block and PVB
Outcomes	Primaryoutcome: morphine consumption Secondary outcomes: numeric rating scale (NRS) scores for pain were recorded at 1, 6, 12 and 24 hours postoperatively
Notes	—

Hamed 2019.

Study characteristics
Methods	Prospective, randomised, single‐blind, controlled study Ethics approval obtained An online randomisation program was used to generate a random number list (1:1 ratio). Patient randomisation numbers were concealed in opaque envelopes that were opened by the study investigator.
Participants	Setting: Faculty of Medicine, Fayoum University, Fayoum, Egypt Sample size: 60 female patients Participant (baseline) characteristics: Age 40 to 70 years 50 kg to 90 kg ASA Ι to ΙΙΙ Inclusion criteria: participants were scheduled for elective total abdominal hysterectomy via Pfannenstiel incision under general anaesthesia Exclusion criteria: local infection at the incision site; a history of haematological disorders including coagulation abnormality, previous abdominal surgeries, severe hepatic impairment or chronic pain, or a known allergy to the study drugs
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: total fentanyl consumption in the first 24 hours Secondary outcomes: VAS pain score (ranging from 0 indicating no pain to 10 indicating extreme pain) at several postoperative time points, hospital stay duration, any complications
Notes	—

Ibrahim 2019.

Study characteristics
Methods	Single‐blinded, randomised and controlled trial Ethics approval obtained Patients were randomised with a 1:1 allocation ratio. The allocated intervention was written on a slip of paper then placed in sealed, serially numbered and opaque envelopes. The envelopes were serially opened to determine in which group the patient would be included. Blindness to the treatment group involved the surgical and anaesthesia teams, operating theatre staff, post‐anaesthesia care unit (PACU) and surgical ward nurses.
Participants	Setting: Faculty of Medicine, Department of Anaesthesiology, Zagazig University, Zagazig, Egypt Sample size: 56 patients Participant (baseline) characteristics: Age 18 to 65 years ASA I or II Inclusion criteria: patients scheduled for first‐stage unilateral PCNL to extract stones, which were large, multiple or resistant to shock wave lithotripsy from the upper ureter and the pelvicalyceal system Exclusion criteria: patients with psychological disorders, morbid obesity, hepatic or renal dysfunction, coagulation disorders, sepsis, history of chronic analgesic use and history of allergy to local anaesthetic agents
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: postoperative 24‐hour morphine consumption Secondary outcomes: intraoperative fentanyl consumption, time to the first use of PCA (started from tracheal extubation), PACU stay, NRS on awakening in PACU (time of PACU admission = 0 time) and at 2, 4, 6, 12 and 24 hours after surgery
Notes	—

Ibrahim 2020.

Study characteristics
Methods	Double‐blinded, randomised and controlled trial Ethics approval was obtained Blindness to the treatment group involved the patients, surgical and anaesthesia teams, operating theatre staff, and PACU and surgical ward nurses
Participants	Setting: Al Jedaani Group of Hospitals and Ibn Sina Medical College Hospital, Jeddah Sample size: 70 adult patients Participant (baseline) characteristics: Age 20 to 60 years ASA I and II BMI 20 to 35 kg/m2 Inclusion criteria: adults planned to undergo elective laparoscopic cholecystectomy Exclusion criteria: allergy to LAs, presence of bleeding disorders, infection at injection sites, prior addiction or analgesic abuse (opioid and nonsteroidal antiinflammatory drug medications), liver or renal insufficiency, history of psychiatric or neurological disease, deafness and previous open surgery, patients who need to be converted to open surgery with more tissue trauma, and ASA physical status class above II patients
Interventions	Intervention group: ESPB Comparator group: TAPB
Outcomes	Primaryoutcome: morphine consumption in the 24 hours postoperatively Secondary outcomes: VAS at 0 (PACU), 2, 4, 6, 12 and 24 hours postoperatively, intraoperative fentanyl (μg) consumed during surgery, equivalent morphine dose at PACU, first analgesic dose, extubation time, PONV at PACU and 24 hours postoperative were recorded, ESP block and OSTAP block complications (including LA systemic toxicity, vascular injury and intravascular injection of LA), surgical time (defined as the time between the incision and the completion of the dressing)
Notes	—

Kamel 2020.

Study characteristics
Methods	Prospective, double‐blind, randomised controlled clinical trial The study was approved by the university's institutional review board
Participants	Setting: Faculty of Medicine, Zagazig University, Alsharkia, Egypt Sample size: 48 women Participant (baseline) characteristics: Age 40 to 60 years BMI 25 to 35 kg/m2 ASA I‐II Inclusion criteria: women scheduled to undergo open total abdominal hysterectomy Exclusion criteria: local infection at site of puncture, altered mental status and history of allergy to study drugs (bupivacaine or morphine), chronic pain, severe hepatic or kidney impairment, and haematologic disorders, including coagulation abnormality or on anticoagulants
Interventions	Intervention group: ESPB Comparator group: TAPB
Outcomes	Primaryoutcome: pain intensity using VAS score at 30 minutes, 2, 4, 6, 8, 12, 16, 20 and 24 hours postoperative, time for requirement of first morphine dose (hours), total amount of morphine consumption during the first 24 hours of postoperative period Secondary outcomes: overall patient satisfaction at the end of 24 hours postoperatively (1 to 3 verbal scale (1 = unsatisfactory analgesia, 2 = satisfactory analgesia, 3 = excellent analgesia)), morphine‐related side effects (incidence of nausea and vomiting, respiratory depression, bradycardia, pruritus and urine retention), bupivacaine‐related side effects (lightheadedness, circumoral numbness, tongue paraesthesia, drowsiness, irritability, muscle twitches, convulsions, bradycardia, hypotension, hypoventilation and cardiac arrest), adverse effects of the block techniques (local site infection, haematoma formation, bowel perforation and pneumothorax)
Notes	—

Kang 2019.

Study characteristics
Methods	Prospective randomised controlled trial A hospital staff member who was not otherwise involved in the study performed computer‐generated block randomisation in a 1:1 ratio to allocate enrolled patients to an intrathecal morphine (ITM) group (n = 27) or an ESPB group (n = 27). The allocation of each patient was concealed in an opaque envelope, which was only opened on the day of surgery by the anaesthesiologist performing the blocks. These investigators were not involved in outcome assessment or further patient management. All other personnel involved in intraoperative and postoperative care of the patients (anaesthesiologists, surgeons and nursing staff), as well as assessors of study outcomes, were blinded to group allocation.
Participants	Setting: Samsung Medical Center, Seoul, Korea Sample size: 54 patients Participant (baseline) characteristics: Age over 18 years ASA I‐II Inclusion criteria: patients scheduled for elective laparoscopic right hepatectomy Exclusion criteria: patients who refused to participate in the study or had contraindications to peripheral nerve block or allergy to local anaesthetics
Interventions	Intervention group: ESPB Comparator group: ITM
Outcomes	Primaryoutcome: resting pain score at 24 hours postoperatively Secondary outcomes: highest, lowest and average pain scores in PACU, resting pain scores at 8, 48 and 72 hours postoperatively, cumulative opioid consumption in PACU, and at 8, 24, 48 and 72 hours postoperatively, incidence of postoperative nausea, vomiting or pruritus within 24 hours after surgery, donor satisfaction with pain relief at 24 hours postoperatively, sleep disturbance on the first night
Notes	—

Khorasanizadeh 2020.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained Patients in the study were divided into the 2 groups in a random manner before inducing general anaesthesia
Participants	Setting: Shohadaye Tajrish Hospital, Tehran, Iran Sample size: 64 women Participant (baseline) characteristics: Age 18 to 50 years ASA I and II Inclusion criteria: women scheduled for breast lump removal Exclusion criteria: allergy to local anaesthetics, psychotic disorders, drug addiction, coagulation disorder, local skin infection, and chronic use of analgesics and nonsteroidal anti‐inflammatory drugs
Interventions	Intervention group: ESPB Comparator group: PECS block
Outcomes	Primaryoutcome: pain score based on VAS (at recovery time and 1, 2, 4 and 6 hours after surgery) Secondary outcomes: nausea and vomiting, opioids (meperidine, at the doses of 0 mg, 25 mg and 50 mg), intraoperative systolic and diastolic blood pressure and pulse rate
Notes	—

Krishna 2019.

Study characteristics
Methods	Prospective, randomised, controlled, single‐blinded study Ethics approval obtained Randomisation using computer‐generated random number table and then sealed in opaque envelope
Participants	Setting: All India Institute of Medical Sciences, New Delhi, India Sample size: 106 adult patients Participant (baseline) characteristics: Age 20 to 70 years Inclusion criteria: patients undergoing elective cardiac surgery with cardiopulmonary bypass Exclusion criteria: emergency surgery, left main stem disease, moderate to severe left ventricular dysfunction, preoperative inotropic support or IABP, mitral stenosis with left atrial thrombus, re‐operation, low cardiac output syndrome, abnormal coagulation, renal or hepatic disorders, uncontrolled diabetes, COPD, allergies to local anaesthetics, atrial fibrillation on anticoagulants
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: pain levels using NRS score Secondary outcomes: postoperative opioid use was also measured
Notes	—

Liu 2021.

Study characteristics
Methods	Prospective, randomised, single‐centre, double‐blinded study The study was approved by the hospital ethics committee The patients were randomised (1:1) and divided into a control group (n = 40) and an ultrasound‐guided ESPB group (n = 40) by the random number table method (www.randomization.com). The randomised schedule was hidden and generated by the anaesthesiologist in charge of the case, who was not part of the study.
Participants	Setting: Hospital of Soochow University in China Sample size: 80 patients Participant (baseline) characteristics: Age 18 to 75 years ASA I–II Inclusion criteria: patients scheduled for uniportal VATS (including lobectomy, wedge resection, segmentectomy) under general anaesthesia Exclusion criteria: refusal to participate in the trial; severe cardiovascular diseases, EF < 40%; severe pulmonary diseases (FEV1/FVC < 40%); liver and kidney dysfunction; chronic pain; puncture site infection; conversion to thoracotomy during the operation; requirement for mechanical ventilation after surgery and transfer to the intensive care unit (ICU); and reoperation due to adverse events such as severe bleeding
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: sufentanil dose within 24 hours after surgery Secondary outcomes: postoperative pain scores at 2, 4, 8 and 24 hours evaluated using a NRS, intraoperative opioid dosage, levels of inflammatory cytokines, including tumour necrosis factor‐alpha (TNF‐α), interleukin‐6 (IL‐6) and interleukin‐10 (IL‐10) in the plasma, block side effect profile, length of postoperative hospital stay
Notes	—

Mostafa 2020.

Study characteristics
Methods	Prospective, double‐blinded randomised controlled trial Approval of the Hospital Ethics Committee obtained Patients randomly assigned into 2 groups using computer‐generated randomisation numbers. The randomisation sequence was concealed in sealed, opaque envelopes.
Participants	Setting: Faculty of Medicine, Tanta University, Tanta, Egypt Sample size: 60 adults Participant (baseline) characteristics: Age 18 to 65 years BMI ≥ 40 kg/m2 ASA III Inclusion criteria: morbidly obese adult patients undergoing laparoscopic bariatric surgery Exclusion criteria: patient refusal as well as patients with hepatic or renal dysfunction, coagulation disorders, uncontrolled cardiovascular or respiratory conditions, psychiatric or uncooperative patients, and allergy to the study drug
Interventions	Intervention group: ESPB Comparator group: placebo block
Outcomes	Primaryoutcome: postoperative pain scores Secondary outcomes: intraoperative fentanyl consumption, cumulative 24‐hour postoperative morphine consumption and postoperative pulmonary function
Notes	—

Nagaraja 2018.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained Randomisation was performed to 2 groups of 25 each, Group A (thoracic epidural analgesia, TEA) and Group B (ESPB) using the closed envelope method
Participants	Setting: Sri Jayadeva Institute of Cardiovascular Sciences and Research, Bengaluru, Karnataka, India Sample size: 50 patients Participant (baseline) characteristics: Age over 18 years Inclusion criteria: adult elective cardiac surgical patients undergoing median sternotomy Exclusion criteria: emergency surgery, left main coronary artery disease, left ventricular ejection fraction < 40%, anomalies of vertebral column, blood or CSF tap during the procedure, failed blocks, patient on anticoagulants, bleeding diathesis and patients who expired before extubation
Interventions	Intervention group: ESPB Comparator group: TEA
Outcomes	Primaryoutcome: intraoperative fentanyl consumption Secondary outcomes: postoperative pain assessment using a VAS at rest and during cough at 0 (extubation), 3, 6, 12, 24, 36 and 48 hours. Peak inspiratory flow spirometry (incentive spirometry) was performed at 3, 6, 12, 24, 36 and 48 hours postextubation, ventilator duration, ICU stay and any complications.
Notes	—

Oksuz 2019.

Study characteristics
Methods	Prospective, randomised, controlled, double‐blind trial Ethics approval obtained Patients were randomly separated into 2 groups using the closed envelope method to receive tumescent anaesthesia or erector spinae block
Participants	Setting: Sutcu Imam University School of Medicine, Kahramanmaras, Turkey Sample size: 44 female patients Participant (baseline) characteristics: Age 20 to 65 years Inclusion criteria: adult female patients undergoing breast reduction surgery, without adjunctive liposuction on the breast, under general anaesthesia Exclusion criteria: local anaesthesia allergy, coagulation impairment, any renal, cardiac, neurological or hepatic disease, or if they did not wish to participate in the study
Interventions	Intervention group: ESPB Comparator group: tumescent analgesia
Outcomes	Primaryoutcome: 24‐hour tramadol consumption with PCA Secondary outcomes: NRS scores were compared at 30 minutes postoperatively and then at 1, 2, 4, 6, 12 and 24 hours, additional analgesia requirements, patient satisfaction
Notes	—

Park 2021.

Study characteristics
Methods	Prospective, randomised clinical trial Approved by the Institutional Review Board of Samsung Medical Center An Internet program (www.randomizer.org) was used for randomisation in a 1:1 ratio, to allocate enrolled patients to group P (n = 30) or group E (n = 30). Study authors did not perform the sham block and both patients and physicians could not be blinded to the type of pain control. Other investigators for outcome assessment were blinded to group enrolment, as they were not involved in performing the procedure.
Participants	Setting: Samsung Medical Center, Seoul, Korea Sample size: 60 patients Participant (baseline) characteristics: Age 20 to 70 years ASA I‐II Inclusion criteria: patients scheduled for an elective unilateral IRB with a tissue expander after mastectomy Exclusion criteria: patient refusal to participate in the study, chronic pain, psychological disorders, allergy to the study drugs (especially local anaesthetics), a history of drug abuse, coagulation disorders, infection at the injection site and renal dysfunction (preoperative serum creatinine level > 2.0 mg/dL)
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: total opioid consumption during 24 hours postoperatively Secondary outcomes: patient satisfaction, pain score at rest and on shoulder movement using an NRS, incidence of postoperative nausea and vomiting (PONV), short form of the Brief Pain Inventory (BPI‐SF) at 3 and 6 months after surgery
Notes	—

Prasad 2020.

Study characteristics
Methods	Prospective, double‐blind study Approval from the institutional ethical committee was obtained
Participants	Setting: Department of Anaesthesia in a tertiary care centre in northern part of India Sample size: 61 adult patients Participant (baseline) characteristics: Age 18 to 65 years ASA I and II Serum creatinine value of less than 1.5 mg/dL Inclusion criteria: patients undergoing elective unilateral percutaneous nephrolithotomy Exclusion criteria: history of allergy, creatinine level more than 1.5 and any drug abuse
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: pain intensity using a VAS Secondary outcomes: haemodynamic variables (heart rate, systolic, diastolic and mean blood pressure)
Notes	—

Qiang 2018.

Study characteristics
Methods	Randomised controlled study Study was approved by the ethics committee Patients were divided into 2 groups, using a random number table method
Participants	Setting: Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College Sample size: 40 patients Participant (baseline) characteristics: Age 30 to 64 years BMI 18 to 24 kg/m2 ASA I or II Inclusion criteria: patients scheduled for elective thoracoscopic radical resection of lung cancer Exclusion criteria: puncture site infection, abnormal coagulation function, history of ropivacaine allergy, severe heart, lung, liver, kidney insufficiency, mental illness incompatibility, history of radiotherapy and chemotherapy before surgery, sedative drugs and narcotic analgesics used recently
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: intraoperative sufentanil and remifentanil dosage Secondary outcomes: anaesthesia recovery time, tracheal extubation time and postoperative hospital stay. Comparison of salvage analgesia rates between the 2 groups. Adverse reactions, PONV and satisfaction scores.
Notes	Translated from Chinese using Google translator

Sakae 2020.

Study characteristics
Methods	Single‐blind, non‐inferiority, randomised clinical trial Approval by the Research Ethics Committee of Universidade do Sul de Santa Catarina obtained Patients were chosen randomly by random number draw of the Epiinfo 6 software Epitable function, for group samples: 1) "Positive control" group with epidural anaesthesia; and 2) Intervention group with ESPB. Postoperative assessors were blind to the group to which patients belonged and to intraoperative management and complications (hypotension, bradycardia, use of vasoactive drugs).
Participants	Setting: Servico de Anestesiologia SIANEST, Florianópolis, SC, Brazil Sample size: 31 adult patients Participant (baseline) characteristics: Age over 18 years ASA I and II BMI ranged between 18 and 38 kg/m2 Inclusion criteria: patients undergoing open cholecystectomy surgery and elective Kocher incision Exclusion criteria: urgent/emergency indication cholecystectomy, intraoperative cholangiography or posterior biliary tract handling, ASA III or higher, history of allergy to any study medication, use of anticoagulants, block failure, use of anticonvulsants or postoperative cognitive deficit that prevented the patient from understanding/answering the questionnaire
Interventions	Intervention group: ESPB Comparator group: epidural analgesia
Outcomes	Primaryoutcome: opioid required in the PACU Secondary outcomes: intraoperative complications (bradycardia, hypotension, vasoactive drug required), rescue opioid use in postoperative 24 hours, postoperative nausea and vomiting and adverse reactions (pruritus, urinary retention)
Notes	Authors replied with further data

Seelam 2020.

Study characteristics
Methods	Prospective, randomised, single‐blinded study Approved by local Institutional Ethics Committee Computer‐generated randomisation was performed for both groups (www.random.org)
Participants	Setting: Basavatarakam Indo‐American Cancer Hospital and Research Institute, Hyderabad, Telangana, India Sample size: 100 female patients Participant (baseline) characteristics: Age 20 to 65 years ASA I–II Inclusion criteria: adult female patients scheduled for unilateral, elective modified radical mastectomy for breast cancer Exclusion criteria: patient not willing to participate, BMI more than 35 kg/m2, known allergy to LA, ASA‐PS III and beyond, underlying coagulopathies, renal and hepatic dysfunction
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: total morphine consumption at the end of 24 hours Secondary outcomes: rates of PONV, VAS scores
Notes	—

Sharma 2020.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained Patients were allocated to either the block group (group B) or control group (group C) using computer‐generated random numbers. Random numbers were concealed using opaque envelopes and were opened just prior to the administration of block. Pain score was recorded by an investigator blinded to the study
Participants	Setting: Nehru Hospital, Chandigarh, India Sample size: 60 female patients Participant (baseline) characteristics: Age 18 to 60 years ASA I or II Inclusion criteria: patients scheduled to undergo total mastectomy and axillary clearance under general anaesthesia Exclusion criteria: infection at the local site, coagulopathy, morbid obesity (BMI > 40 kg/m2), uncontrolled hypertension or ischaemic heart disease, renal dysfunction, allergy to local anaesthetics, pre‐existing neurological deficits and psychiatric illness
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: pain score on a numeric rating scale (NRS) recorded at 0, 1⁄2, 1, 2, 4, 6, 8, 12 and 24 h Secondary outcomes: 24‐hour morphine consumption, rescue analgesia within 1st hour of surgery, rates of PONV and block related complications
Notes	—

Shim 2020.

Study characteristics
Methods	Prospective, randomised, single‐centre, double‐blind trial Approved by the institutional ethics committee Randomisation was conducted at a 1:1 ratio using a web‐based response system (http:// www.randomization.com). Assignments were sealed in serially numbered envelopes. Randomisation and blinding procedures were performed by an independent researcher who was not involved in the trial. The patients and physician assessing the outcomes were kept blinded to the grouping process (double‐blind study).
Participants	Setting: Samsung Hospital, Seoul, Republic of Korea Sample size: 46 patients Participant (baseline) characteristics: Age 19 to 85 years ASA I‐III Inclusion criteria: adult patients scheduled for unilateral lobectomy under complete VATS with 3 trochar ports for lung cancer Exclusion criteria: patients with abnormal clotting haemostatic test results, patients on anticoagulant treatment, patients with a history of allergy to LA agents, patients with skin problems at the needle puncture site, pregnant patients and highly obese patients (BMI > 30 kg/m2)
Interventions	Intervention group: ESPB Comparator group: placebo block
Outcomes	Primaryoutcome: NRS score 12 hours postoperatively Secondary outcomes: Riker Sedation‐Agitation Scale (SAS) score for emergence agitation, postoperative cumulative opioid consumption, length of post‐anaesthesia care unit (PACU) stay, incidence of postoperative nausea and vomiting (PONV) and dizziness, ESPB‐related adverse events
Notes	—

Singh 2019.

Study characteristics
Methods	Prospective randomised controlled trial Approved by Institutional Research Committee Patients were randomly allocated into groups by using a computer‐generated random table. Randomisation was done by a statistician and the group of the patient was revealed only when the included patient was shifted to the preanaesthetic room.
Participants	Setting: Indira Gandhi Institute of Medical Sciences, Patna, Bihar, India Sample size: 40 female patients Participant (baseline) characteristics: Age 20 to 55 years ASA I‐II Inclusion criteria: patients scheduled to undergo modified radical mastectomy under general anaesthesia. Exclusion criteria: —
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: cumulative morphine consumption for 24 hours after surgery Secondary outcomes: NRS measured at 0, 2, 4, 6, 8, 10, 12 and 24 hours, PONV rates, patient satisfaction score
Notes	—

Singh 2020.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained Participants were randomly allocated to 1 of 2 groups using a computer‐generated random table
Participants	Setting: Institute of Medical Education and Research, Chandigarh, Punjab Sample size: 40 patients Participant (baseline) characteristics: Age 18 to 65 years ASA I to III Inclusion criteria: patients scheduled to undergo elective lumbar spine surgery (prolapsed lumbar intervertebral disk, lumbar stenosis or laminectomy) Exclusion criteria: chronic analgesic therapy, history of opioid dependence, anticoagulation or experiencing any bleeding disorder, unable to communicate with the investigators
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: cumulative morphine consumption during the first 24 hours after surgery Secondary outcomes: pain score at rest, patient satisfaction score
Notes	—

Sinha 2019.

Study characteristics
Methods	Prospective, randomised, single‐blinded study Approval from the Institutional Ethics Committee obtained The statistician using computer‐generated random numbers randomly allocated them into 2 groups of 32 each. The group allocation numbers were concealed in sealed, opaque envelopes that were opened only after shifting the patient to preoperative holding area.
Participants	Setting: AIIMS, Patna, Bihar, India Sample size: 64 female patients Participant (baseline) characteristics: Age 18 to 60 years ASA I and II Inclusion criteria: adult female patients scheduled for unilateral modified radical mastectomy (MRM) under general anaesthesia Exclusion criteria: patients scheduled for bilateral surgery, suffering from psychiatric disorder, chronic neurological disease, not willing to give consent, having BMI > 40
Interventions	Intervention group: ESPB Comparator group: PECS block
Outcomes	Primaryoutcome: total analgesic consumption in 24 hours Secondary outcomes: duration of analgesia (time to first rescue analgesia after administration of block), level of sensory blockade as assessed preoperatively and postoperative pain scores, adverse effects such as hypotension and respiratory depression
Notes	—

Sobhy 2020.

Study characteristics
Methods	Prospective, randomised, observer‐blind, controlled clinical trial Institutional Ethical Committee approval Patients were randomly chosen to receive either ESPB plus conventional opioid analgesics (group ESPB, n = 30) or receive only the conventional opioid analgesics (group C, n = 30) by a random sequence number generated by the computer kept in sealed envelopes. The sealed envelopes were opened on the day of surgery when the patient in the operation room and participants received either ESPB or not as per the envelope. The observer anaesthesiologist postoperative will be blinded to which group the patient belongs.
Participants	Setting: Benha University Hospital Sample size: 60 patients Participant (baseline) characteristics: Age over 18 years ASA I–IV Inclusion criteria: patients scheduled for an elective thoracotomy Exclusion criteria: refusal of the patient to provide written consent, age less than 18, coagulation disorders, known allergy to study drugs, obesity (BMI > 40 kg/m2), infection at the injection site and pregnant females
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: total postoperative morphine consumption in the first 24 hours Secondary outcomes: VAS scores at 5 different time points (0 "on arrival", 6th hour, 12th hour, 18th hour and 24th hour) at rest and during coughing, incidence of complications related to the technique, morphine‐related side effects (nausea, vomiting, pruritus and excessive sedation), patient satisfaction with postoperative analgesia after 72 hours postoperatively according to a satisfaction score (poor = 0, fair = 1, good = 2, excellent = 3), total postoperative hospital stay
Notes	—

Sotome 2021.

Study characteristics
Methods	Single‐blinded, prospective randomised controlled trial Approved by the Sapporo Medical University Research Ethics Board 50 consecutive patients were randomly allocated to receive ESPB or RLB under ultrasound guidance. The randomisation protocol was prepared outside the study centre. Randomisation was performed using a computer‐generated randomisation sequence (https://www.randomizerd.org). Group allocation was concealed using consecutively numbered, opaque, sealed envelopes prepared by a research fellow not involved in the study. The patients were blinded to group assignment. The operator performing ESPB or RLB confirmed group allocation just before the patient entered the operating room.
Participants	Setting: Sapporo Medical University Hospital (Sapporo, Japan) Sample size: 50 adult female patients Participant (baseline) characteristics: Age over 18 years ASA physical status I or II Inclusion criteria: adult female patients scheduled for unilateral mastectomy with sentinel lymph node biopsy under general anaesthesia Exclusion criteria: history of daily consumption of analgesics, allergy to local anaesthetics, liver dysfunction, and infection or previous surgery in the thoracic vertebral region
Interventions	Intervention group: ESPB Comparator group: RLB
Outcomes	Primaryoutcome: analgesic efficacy in terms of time to first postoperative rescue analgesic after the block procedure Secondary outcomes: consumption of remifentanil during anaesthesia, pain intensity at rest for 24 hours postoperatively, occurrence of postoperative nausea and vomiting (PONV)
Notes	—

Swisher 2020.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained Participants undergoing unilateral or bilateral non‐mastectomy breast surgery were randomised to a single‐injection ESPB or PVB in a participant‐blinded fashion (ropivacaine 0.5% with epinephrine; 20 mL unilateral or 16 mL/side for bilateral)
Participants	Setting: University of California San Diego, San Diego, California, USA Sample size: 100 patients Participant (baseline) characteristics: Age ≥ 18 years Inclusion criteria: patients presenting for unilateral or bilateral breast surgery with at least moderate postoperative pain anticipated Exclusion criteria: planned regional analgesic with perineural catheter placement, morbid obesity as defined as a BMI > 40 kg/m2, renal insufficiency (preoperative creatinine > 1.5 mg/dL), current chronic opioid use (daily equivalent of > 20 mg of oxycodone within the 2 weeks prior to surgery and duration of use > 4 weeks), history of opioid abuse, any comorbidity that results in moderate or severe functional limitation, inability to communicate with the investigators or hospital staff, pregnancy, incarceration and allergy to study medications (ropivacaine)
Interventions	Intervention group: ESPB Comparator group: PVB
Outcomes	Primaryoutcome: pain scores within the recovery room and opioid requirements in the operating and recovery rooms combined Secondary outcomes: pain scores were recorded using a NRS. Within the PACU, pain scores, opioid requirements and antiemetic administration were recorded by nursing staff blinded to treatment. The morning following surgery, all participants were contacted by telephone or in person (if hospitalised) to record lowest, average, highest and current pain scores, number of sleep disturbances, and nausea using a 0 to 10 Likert scale (0 = no nausea; 10 = vomiting). For ambulatory participants, opioid requirements were recorded by patients while inpatient participants had opioid requirements extracted from the electronic medical record. In addition, for inpatient participants, antiemetic use and nursing‐recorded pain scores were extracted from the electronic medical record. The times at which participants identified block resolution and the time at which they consumed their first oral opioid analgesic following PACU discharge were recorded. In addition, possible block‐related side effects and adverse events were recorded.
Notes	—

Taketa 2019.

Study characteristics
Methods	Prospective randomised controlled trial Ethics approval obtained Allocation was automated, using a computer‐generated identifier. The anaesthesiologist oversaw patients’ enrolment and allocation. All procedures were performed by anaesthesiologists other than those in charge of the operation, who were skilled in ultrasound‐guided nerve blockade.
Participants	Setting: Ehime Prefectural Central Hospital, Matsuyama, Japan Sample size: 88 patients Participant (baseline) characteristics: Age over 20 years ASA I‐III Inclusion criteria: patients scheduled for radical lobectomy under complete video‐assisted thoracic surgery (VATS), with 3 trochal ports, for lung cancer Exclusion criteria: age < 20 or > 80 years, allergy to local anaesthetics, BMI > 30 kg/m2 or the presence of coagulopathy, history of ipsilateral thoracotomy, daily use of opioid analgesics, mental or neurological disorders, allergy to analgesics (acetaminophen, non‐steroidal anti‐inflammatory drugs or opioids), disorder in communication and scheduled thoracotomy with rib cutting or pleurectomy
Interventions	Intervention group: ESPB Comparator group: PVB
Outcomes	Primaryoutcome: median differences in NRS scores, at rest, 24 hours postoperatively Secondary outcomes: NRS scores at rest, dynamic pain scores, amount of rescue fentanyl use per weight, number of anaesthetised dermatomes (6 and 20 hours after the initial bolus injection), total plasma levobupivacaine concentration, number of additional analgesics used by each patient, PONV scores
Notes	—

Tulgar 2018.

Study characteristics
Methods	Randomised, double‐blinded, prospective efficiency study The Institutional Review Board approved the study Patients were assigned a random identification number before surgery and all data were collected using this identification number. Group assignments were determined using simple randomisation with the sealed envelope technique. All data were collected blindly. Blocks were performed by the authors who did not play any role in data collection or analysis.
Participants	Setting: Maltepe State Hospital, Istanbul, Turkey Sample size: 72 patients Participant (baseline) characteristics: Age 18 to 65 years ASA I‐III Inclusion criteria: patients undergoing hip or proximal femoral surgery under general anaesthesia Exclusion criteria: refusal at enrolment, request for removal from the study, inability to give informed consent, the presence of contraindications to LA agents used in this study, known solid organ diseases, pathologies of medication use leading to bleeding diathesis, chronic use of opioids or corticosteroids, condition leading to inability to operate patient‐controlled analgesia (PCA) system and psychiatric disorders. For better standardisation of data, surgical procedures lasting < 60 min or > 180 min were also excluded from the study.
Interventions	Intervention group: ESPB Comparator groups: no block and QLB
Outcomes	Primaryoutcome: tramadol consumption and additional rescue analgesic requirement were measured Secondary outcomes: pain intensity using numeric rating scores
Notes	—

Turhan 2020.

Study characteristics
Methods	Randomised, prospective, single‐blind study Approval from Istanbul University Istanbul Faculty of Medicine Clinical Researches Ethics Committee (2019/51) Randomisation was achieved by computer‐generated random numbers and concealed using opaque, closed envelopes
Participants	Setting: Istanbul University, Istanbul Faculty of Medicine, Turkey Sample size: 111 adult patients Participant (baseline) characteristics: Age over 18 years ASA I‐III Inclusion criteria: patients undergoing thoracoscopic lung surgery under general anaesthesia Exclusion criteria: known chronic pain and opioid use, local anaesthetic allergy, spinal deformity and psychiatric disorder. Technical failure in the blocks and the conversion to an open surgical procedure were also excluded.
Interventions	Intervention group: ESPB Comparator groups: PVB and ICNB
Outcomes	Primaryoutcome: dynamic pain score at first hour postoperatively Secondary outcomes: static pain scores, additional analgesic requirement and side effects
Notes	—

Wang 2019b.

Study characteristics
Methods	Single‐centre, randomised, controlled clinical study Approved by the Institutional Ethics Committee of National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College All patients were randomly allocated into the 2 groups using a computer‐generated random number table. The group allocation numbers were concealed in sealed, opaque envelopes that were opened after enrolment of the patients.
Participants	Setting: Chinese Academy of Medical Sciences and Peking Union Medical College Sample size: 72 patients Participant (baseline) characteristics: Age 30 to 65 years ASA I and II Inclusion criteria: patients undergoing open thoracic surgery for lower thoracic oesophageal cancer Exclusion criteria: infections at the site of injection for ultrasound‐guided ESPB or wound infiltration; neuropathy; coagulation disorders; morbid obesity (BMI > 40 kg/m2); allergy to ropivacaine; greater than first‐degree heart block; bradycardia (heart rate < 60 bpm); pregnancy; clinically significant cardiovascular, pulmonary, hepatic or renal diseases; psychiatric illnesses that would interfere with assessment of pain score and those given a painkiller within 1 week before surgery
Interventions	Intervention group: ESPB Comparator group: no block (wound infiltration)
Outcomes	Primaryoutcome: intraoperative sufentanil consumption Secondary outcomes: perioperative opioid consumption, pain scores at rest and during coughing immediately after surgery, at postoperative day (POD) 1 and POD 2, consumption of rescue analgesic tramadol and postoperative opioid‐related adverse events
Notes	—

Yang 2019.

Study characteristics
Methods	Randomised controlled trial Ethics approval was obtained Patients were randomly allocated
Participants	Setting: Shengjing Hospital Affiliated of China Medical University Sample size: 60 patients Participant (baseline) characteristics: Age 18 to 75 years BMI 18 to 30 kg/m2 Inclusion criteria: participants who received video‐assisted thoracoscopic surgery under general anaesthesia Exclusion criteria: local anaesthesia allergy, local infection at the puncture point, coagulopathies, history of chronic pain, recent opioids use, mental disorders
Interventions	Intervention group: ESPB Comparator group: placebo block
Outcomes	Primaryoutcome: VAS scores at 1, 2, 4, 8, 12 and 24 hours postoperative Secondary outcomes: opiate doses, adverse reactions and degree of satisfaction
Notes	Translated from Chinese using Google translator

Yao 2020a.

Study characteristics
Methods	Randomised controlled study Ethical approval for this study was provided by the Ethical Committees of Fujian Provincial Hospital, Fujian, China Randomisation was carried out in a 1:1 ratio using a computer‐generated randomisation sequence by an assistant not involved in the study. Allocations were concealed in numbered, opaque envelopes. A research nurse who was not involved in the study opened the envelopes and prepared the study drug on the day of surgery.
Participants	Setting: Fujian Provincial Hospital, Fujian, China Sample size: 76 patients Participant (baseline) characteristics: Age 18 to 65 years ASA I or II Inclusion criteria: patients scheduled for elective VATS Exclusion criteria: patients with a history of allergy to local anaesthetics or non‐steroidal anti‐inflammatory drugs, known coagulation disorders, infection near the puncture site, a history of chronic pain or regular analgesic use, inability to communicate, or other reasons that were not appropriate for the study
Interventions	Intervention group: ESPB Comparator group: placebo block
Outcomes	Primaryoutcome: postoperative QoR after VATS, assessed by using the Chinese version of the global QoR‐40 score at postoperative day (POD) 1 Secondary outcomes: post‐anaesthesia care unit (PACU) discharge time, acute postoperative pain, cumulative opioid consumption, incidence of postoperative nausea or vomiting (PONV), patient satisfaction
Notes	—

Yao 2020b.

Study characteristics
Methods	Randomised controlled trial Randomisation was performed with a 1:1 ratio using a computerised random number generator (http://www.randomization.com). Assignments were sealed in sequentially numbered envelopes. Randomisation, blinding procedures and study drug preparations were handled by a research nurse who was not involved in the trial. The study drugs, 0.5% ropivacaine and saline, were drawn up into syringes that looked identical. All participants, the surgeon and the anaesthesiologist, postanaesthesia care unit (PACU) personnel and research staff who collected the study data were not informed of the group assignments.
Participants	Setting: Clinical Medical College, Fujian Medical University, Fuzhou, China Sample size: 82 female participants Participant (baseline) characteristics: Age 18 to 65 years ASA I or II Inclusion criteria: patients scheduled for elective unilateral modified radical mastectomy Exclusion criteria: local anaesthetics or non‐steroidal anti‐inflammatory drugs allergies, infection near the puncture site, known coagulation disorders, chronic pain, use of pain medications and inability to provide consent
Interventions	Intervention group: ESPB Comparator group: placebo block
Outcomes	Primaryoutcome: QoR, assessed 24 hours postoperatively using the 15‐item QoR questionnaire (QoR‐15) Secondary outcomes: postoperative pain scores, postoperative cumulative opioid consumption, postanaesthesia care unit (PACU) discharge time, postoperative nausea or vomiting and dizziness
Notes	—

Yaoping 2019.

Study characteristics
Methods	Randomised controlled study This study was approved by the hospital ethics committee (201703‐05) The patients were randomly divided into 2 groups: ESPB recovery under ultrasound guidance and general anaesthesia group (E group) and simple general anaesthesia group (G group)
Participants	Setting:? Sample size: 60 patients Participant (baseline) characteristics: Age 30 to 70 years Weight 50 kg to 90 kg ASA I or II Inclusion criteria: patients with chronic empyema scheduled for elective pleural decortication with video‐assisted thoracoscopic surgery Exclusion criteria: severe hypertension and abnormal heart function, liver and kidney function is not total, nervous system disease, local anaesthetic allergy, puncture site feeling dye
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: remifentanil dose Secondary outcomes: dermatomes of sensory block on midclavicular line were recorded at 20 min after ESPB, duration of stay in postanaesthesia care unit, frequency of PCA pressing, pain VAS scores during rest and movement at 1, 4, 12, 24 and 48 hours after operation
Notes	Translated from Chinese using Google translator

Yayik 2018.

Study characteristics
Methods	Randomised clinical study Patients were randomly assigned into 2 groups
Participants	Setting: Baskent University School of Medicine, Department of Anesthesiology and Reanimation, Konya, Turkey Sample size: 60 patients Participant (baseline) characteristics: Age 18 to 65 years ASA I‐III Inclusion criteria: patients scheduled for 1 or 2‐level lumbar disc herniation surgery Exclusion criteria: —
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: VAS score in PACU, at 2, 4, 8, 12 and 24 hours Secondary outcomes: tramadol consumption 24 hours postoperative
Notes	Only abstract and study protocol available. No email address available to request further information.

Yuzhong 2018a.

Study characteristics
Methods	Randomised controlled study This study was approved by the medical ethics committee of the hospital, and the patients signed an informed consent form Patients were divided into 2 groups using a random number table
Participants	Setting: Affiliate Hospital of Zhengzhou University, China Sample size: 85 patients Participant (baseline) characteristics: Age 18 to 64 years ASA grade II or III BMI 18 to 24 kg/m2 Inclusion criteria: patients undergoing elective video‐assisted thoracoscopic pulmonary lobectomy Exclusion criteria: local anaesthetic allergy, language communication difficulties, history of chronic pain and mental illness, long‐term history of opioid use, abnormality in coagulation function
Interventions	Intervention group: ESPB Comparator group: placebo block
Outcomes	Primaryoutcome: QoR‐40 (quality of recovery) questionnaire to assess the early postoperative quality of recovery at 1 day before surgery and 1 and 2 days after surgery Secondary outcomes: consumption of intraoperative remifentanil and postoperative sufentanil, requirement for rescue analgesics and occurrence of postoperative adverse reactions were recorded
Notes	Translated from Chinese using Google translator

Yuzhong 2018b.

Study characteristics
Methods	Randomised controlled study Study approved by the Medical Ethics Committee of the hospital Using a random number table method, participants were divided into 2 groups (n = 45 each): ultrasound‐guided paravertebral nerve block group (P group) and ultrasound‐guided ESPB group (E group)
Participants	Setting: Affiliate Hospital of Zhengzhou University, China Sample size: 90 patients Participant (baseline) characteristics: Age 18 to 64 years BMI 20 to 27 kg/m2 ASA I‐II Inclusion criteria: patients scheduled for elective VATS Exclusion criteria: history of local anaesthetic allergy, history of chronic pain and mental illness, long‐term history of opioid use, abnormal blood coagulation function
Interventions	Intervention group: ESPB Comparator group: PVB
Outcomes	Primaryoutcome: operation time, anaesthesia time and total consumption of intraoperative remifentanil Secondary outcomes: cumulative consumption of morphine was recorded at 2, 4, 6, 24 and 48 hours after surgery, development of postoperative nausea and vomiting, and respiratory depression was recorded
Notes	Translated from Chinese using Google translator

Zhang 2020.

Study characteristics
Methods	Prospective randomised controlled study This study was approved by the Ethics Committee of Beijing Hospital Enrolment was performed by a physician investigator. Allocation of T12 ESPB group or control group was performed by the physician investigator with treatment divulged by sequentially numbered, sealed envelopes before injection. T12 group received bilateral ultrasound‐guided single‐injection ESPB at the T12 level with ropivacaine. The control group just received an ultrasound scan but did not receive a block. The patient, anaesthesiologist responsible for the operation and evaluating physician (i.e. outcome adjudicator) were blinded to the assignment.
Participants	Setting: Beijing Hospital, China Sample size: — Participant (baseline) characteristics: Age 18 to 80 years ASA I‐II Inclusion criteria: patients with prolapsed lumbar intervertebral disk, lumbar stenosis, undergoing open posterior lumbar decompression surgery Exclusion criteria: hepatic or renal insufficiency, preoperative cognitive dysfunction or communication disorder, allergy to amide‐type local anaesthetics, back puncture site infection, neuromuscular disease, emergency surgery, chronic opioid use and participation in another clinical study
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: Modified Observer’s Assessment of Alertness/Sedation (MOAA/S) score at 10 minutes after extubation Secondary outcomes: intraoperative sufentanil consumption postoperative morphine consumption, first time to ambulation after surgery, hospital length of stay after surgery
Notes	—

Zhang 2021.

Study characteristics
Methods	The study protocol was approved by the Ethics Committee of Peking University People's Hospital Randomised controlled study The randomisation numbers were computer‐generated and stored in sealed, opaque envelopes. Participants were randomised in a 1:1 ratio to receive either ultrasound‐guided bilateral ESPB (ESPB group) or sham subcutaneous infiltration (control group). The interventions were performed by an anaesthesiologist otherwise not involved in the study. The surgeons, anaesthesiologists providing intraoperative care, nursing staff, investigators and patients were blinded to patient group allocation throughout the study.
Participants	Setting: Peking University People's Hospital, Beijing, China Sample size: 60 patients Participant (baseline) characteristics: Age 20 to 75 years ASA I‐II Inclusion criteria: adult patients scheduled to undergo primary open posterior lumbar spinal fusion surgery under general anaesthesia. Patients with involvement of ≤ 3 spinal levels were eligible. Exclusion criteria: ASA classification III or higher; BMI > 31 kg/m2; chronic opioid use; coagulation abnormality (defined as international normalised ratio > 91.4 or prothrombin time prolonged > 4 s), or platelet count < 100*10 /L; injection site infection; hepatic insufficiency or more than twice the upper normal limit of hepatic enzymes; renal insufficiency, defined as an estimated glomerular filtration rate < 90 mL/min/1.73 m2; allergy or contraindication to drugs in the study protocol
Interventions	Intervention group: ESPB Comparator group: placebo block
Outcomes	Primaryoutcome: pain intensity at rest within 12 hours postoperatively using a NRS Secondary outcomes: NRS pain scores at rest and on movement, postoperative opioid consumption, proportions of patients requiring opioids during the first 48 hours after surgery
Notes	—

Zhao 2020.

Study characteristics
Methods	Randomised controlled study The study protocol was approved by the Institutional Review Board of Peking University People's Hospital Patients were randomly allocated to receive single shot of ESPB (Group ESPB) or paravertebral block (Group PVB) at T4 and T6 levels according to a random number list generated by a computer that was stored in sealed envelopes. An independent experienced anaesthesiologist performed the blocks according to the random number. The study co‐ordinator, attending anaesthesiologist, data collection resident and the patients were all blinded to the treatment group assignment. Both blocks were applied on the patient's back in a lateral position; thereafter the patient could not tell the difference.
Participants	Setting: Peking University People's Hospital in Beijing, China Sample size: 66 patients Participant (baseline) characteristics: Age 18 and 75 years Inclusion criteria: patients diagnosed as having lung nodules and scheduled for VATS under general anaesthesia Exclusion criteria: allergy to NSAIDs, history of asthma, peptic ulcer disease, inflammatory bowel disease or renal deficiency
Interventions	Intervention group: ESPB Comparator group: PVB
Outcomes	Primaryoutcome: oxycodone rescue consumption at 48 hours postoperative Secondary outcomes: episodes of intraoperative hypotension, pain scores rated with a NRS ranging from 0 to 10, QoR15 at postoperative 24 and 48 hours
Notes	—

Zheng 2019.

Study characteristics
Methods	Randomised controlled study Study approved by the Ethics Committee of Wenzhou People’s Hospital in Zhejiang Province The patients were divided into 3 groups according to a random number table: combination 1 group, combination 2 group and control group, with 20 participants in each group
Participants	Setting: Wenzhou People′s Hospital Sample size: 60 patients Participant (baseline) characteristics: Age 40 to 75 years ASA I‐II Body weight 50 kg to 80 kg BMI 20 to 30 kg/m2 Inclusion criteria: patients who underwent thoracoscopic lobectomy under general anaesthesia Exclusion criteria: puncture site infection, severe cardiopulmonary disease, liver and kidney dysfunction, mental disease, severe coagulation abnormality, incompatibility
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Primaryoutcome: ? Secondary outcomes: Mean arterial pressure (MAP), heart rate, plethysmography index (SPI), state entropy (SE) and reaction entropy (RE) before anaesthesia induction, immediately after intubation, at the time of incision, and thoracoscopic cannulation, 30 min after surgery, and at the time of thoracoscopic cannula withdrawal were recorded. The VAS score at rest and cough extubation immediately and 1, 6, 12, 24, 48 hours after extubation were recorded. The compression number of the analgesia pump, number of rescue sufentanil and incidence of adverse events 48 hours after extubation were recorded. The levels of venous blood norepinephrine (NE), epinephrine (E) and cortisol immediately after surgery and 24 hours after extubation were measured.
Notes	Translated from Chinese using Google translator

Authors contacted via email whenever necessary but for most, no reply was received.

Abbreviations:

ASA: American Society of Anesthesiology
BMI: body mass index
CONSORT: Consolidated Standards of Reporting Trials
CSF: cerebrospinal fluid
ESPB: erector spinae plane block
EV: ejection fraction
FEV1: forced expiratory volume in one second
FVC: forced vital capacity
GA: general anaesthesia
ICNB: intercostal nerve block
ITM: intrathecal morphine
IV: intravenous
LA: local anaesthetic
MITS: minimally invasive thoracic surgery
mTLIP: modified thoracolumbar interfascial plane block
MTPB: mid‐transverse to pleura block
NPS: numerical pain scale
NRS: numerical rating scale
OR: operating room
OSTAP: oblique subcostal transversus abdominis plane block
PACU: post‐anaesthesia care unit 
PCA: patient‐controlled analgesia 
PECS: pectoralis plane block
PLIF: posterior lumbar interbody fusion
PONV: postoperative nausea and vomiting
PVB: paravertebral block
QLB‐t: transmuscular quadratus lumborum block
QLB: quadratus lumborum block
QoR: quality of recovery 
RLB: retrolaminar block
SAPB: serratus anterior plane block
SPB: serratus plane block
TAPB: transversus abdominis plane block
TEA: thoracic epidural analgesia
TPVB: thoracic paravertebral block
US: ultrasound
VAS: visual analogue scale
VATS: video‐assisted thoracic surgery

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Qiulan 2019	This study investigated the wrong outcome for this meta‐analysis.
Sajna 2020	In this trial, surgery was not performed under general anaesthetic, which is an exclusion criterion for this meta‐analysis.
Tao 2019	This study investigated the wrong outcome for this meta‐analysis.
Tulgar 2019	The paper was retracted.
Wang 2019a	No translation of the manuscript was possible.

Characteristics of studies awaiting classification [ordered by study ID]

Abd 2021.

Methods	Randomised, controlled, double‐blinded trial Patients were randomly allocated into 3 equal groups to receive either ESPB, QLB or standard care without block
Participants	Setting: Zagazig University Hospitals Sample size: 75 adult patients Participant (baseline) characteristics: Age 21 to 60 years ASA II‐III BMI ≤ 30 kg/m2 Inclusion criteria: adult patients aged 21 to 60 years old, of either sex, BMI ≤ 30 kg/m2, ASA II or ASA III scheduled for elective open nephrectomy with flank incision under general anaesthesia Exclusion criteria: patients with a history of allergy to the local anaesthetic agents used in this study, skin lesion at needle insertion site, those with bleeding disorders, sepsis, liver disease and psychiatric disorders were excluded from this study.
Interventions	Intervention group: ESPB Comparator groups: QLB and standard care without block
Outcomes	24‐hour postoperative cumulative morphine consumption First rescue analgesic time Postoperative pain score Time to perform technique Sensory block coverage
Notes	—

Abraham 2021.

Methods	Preregistered randomised controlled trial with ethics approval Patients were randomly allocated into 2 groups to receive PECS block or ESPB after induction of anaesthesia. Fentanyl was administered for postoperative pain relief via PCA pump.
Participants	Setting: specialist hospital, Saket, India Sample size: 80 adult female patients Participant (baseline) characteristics: Age 18 to 80 years ASA grades I‐II Inclusion criteria: adult female patients (ASA grades I‐II), aged 18 to 80 years old, weighing over 50 kg and undergoing modified radical mastectomy under general anaesthesia Exclusion criteria: pregnancy, any bleeding disorders, infection at injection site, severe/morbid obesity, allergy to local anaesthetics
Interventions	Intervention group: ESPB Comparator group: PECS block
Outcomes	NRS scores Time to first rescue analgesia Intraoperative and postoperative fentanyl requirement Incidence of PONV Patient satisfaction
Notes	—

Agarwal 2021.

Methods	Prospective, randomised, double‐blinded clinical trial. Ethics approval was received. Randomisation and blinding were done using computer‐generated random numbers, and a sealed, opaque envelope concealing the group allocation number was opened after enrolment of the patients.
Participants	Setting: AIIMS, India Sample size: 80 female patients Participant (baseline) characteristics: Age 18 to 70 years ASA I and II Inclusion criteria: female patients aged 18 to 70 years, belonging to physical status ASA I and II, undergoing modified radical mastectomy were included in the study. Exclusion criteria: body mass index (BMI) > 35 kg/m2, infection at the site of injection, coagulopathy, spine deformity, history of opioid dependence, history of allergy to opioids or local anaesthetics
Interventions	Intervention group: ESPB Comparator group: PVB
Outcomes	Time to first rescue analgesia Total doses of rescue analgesics NRS scores at 0 minutes, 30 minutes, 1, 2, 6, 12 and 24 hours Patient satisfaction at 24 hours
Notes	—

Asar 2022.

Methods	Randomised controlled trial in which 78 patients undergoing elective open lumbar spine surgery randomised into 2 groups, ESPB or no block
Participants	Setting: Aydın Adnan Menderes University Hospital Sample size: 78 adult patients Participant (baseline) characteristics: Age over 18 years Inclusion criteria: adults undergoing elective open lumbar spine surgery Exclusion criteria: emergency surgery and children
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Postoperative opioid consumption as morphine equivalent dose Numerical rating scale Mobilisation time Discharge time Side effects Bolus deliveries Rescue analgesia doses
Notes	—

Athar 2021.

Methods	Prospective, randomised, double‐blinded clinical trial
Participants	Setting: single‐centre, tertiary care hospital with university affiliation Sample size: 30 adult patients Participant (baseline) characteristics: BMI 19 to 30 kg/m2 Age 18 to 60 years Inclusion criteria: adult patients undergoing elective on‐pump single‐vessel coronary artery bypass grafting or valve replacement under general anaesthesia Exclusion criteria: emergency surgery
Interventions	Intervention group: ESPB Comparator group: sham block
Outcomes	24‐hour postoperative analgesic consumption Sedation Duration of mechanical ventilation
Notes	—

Dost 2021.

Methods	Prospective, randomised, placebo‐controlled, double‐blinded trial
Participants	Setting: Ondokuz Mayis University Hospital Sample size: 50 adult patients Participant (baseline) characteristics: Age 18 to 65 years ASA I‐III Inclusion criteria: adult patients undergoing elective open radical prostatectomies Exclusion criteria: emergency surgery
Interventions	Intervention group: ESPB Comparator group: sham block
Outcomes	Morphine consumption in the first 24 hours after surgery NRS pain scores at rest and coughing Intraoperative remifentanil consumption Need for rescue analgesic during the first 24 hours after surgery
Notes	—

Elsabeeny 2021.

Methods	Single‐blinded randomised controlled trial
Participants	Setting: National Cancer Institute in Cairo, Egypt Sample size: 51 adults patients Participant (baseline) characteristics: Age over 18 years Inclusion criteria: adult patients with cancer undergoing elective thoracotomy Exclusion criteria: emergency surgery
Interventions	Intervention group: ESPB Comparator groups: TEA and SAPB
Outcomes	24 hours postoperative visual analogue scale Intraoperative rescue fentanyl consumption Perioperative heart rate Mean blood pressure Total postoperative morphine consumption
Notes	—

Finnerty 2021.

Methods	Randomised controlled trial, approved by Ethics Review Board of the Mater Misericordiae University Hospital Randomisation was done by computer‐generated number sequences and blinding was achieved using folded and sealed sequentially numbered, opaque envelopes, which were opened after induction of general anaesthesia.
Participants	Setting: Mater Private Hospital, Dublin, Ireland Sample size: 60 adult patients Participant (baseline) characteristics: Age over 18 years ASA physical status I‐IV Inclusion criteria: adults undergoing open thoracolumbar decompressive spinal surgery Exclusion criteria: patients for whom peripheral regional anaesthesia was contraindicated, those with a history of opioid abuse or chronic pain, and patients whose cognition was sufficiently impaired to preclude written informed consent or postoperative assessment
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Summed QoR‐15 score at 24 hours Pain and opioid consumption for the first 24 hours Time to first intravenous opioid
Notes	—

Genc 2021.

Methods	Randomised controlled trial with ethical approval
Participants	Setting: University Faculty of Medicine, Department of Anesthesiology and Reanimation, Samsun, Turkey Sample size: 90 adult patients Participant (baseline) characteristics: Age over 18 years Inclusion criteria: adult patients undergoing elective breast surgery Exclusion criteria: non‐elective breast surgery
Interventions	Intervention group: ESPB Comparator groups: PECS block and no block
Outcomes	Time to onset of block Intraoperative remifentanil consumption Time to first request for analgesia via patient‐controlled analgesia (PCA) Morphine consumption in the first postoperative 24 hours Number of patients who requested rescue analgesia VAS pain scores of the arm at rest and in abduction at hours 1, 3, 6, 9, 12 and 24 hours VAS pain scores in the third month following discharge
Notes	—

Iqbal 2021.

Methods	Prospective, randomised, double‐blinded study
Participants	Setting: inpatient hospital Sample size: 47 adult patients Participant (baseline) characteristics: Age over 18 years Inclusion criteria: adults undergoing elective percutaneous nephrolithotomy under general anaesthesia Exclusion criteria: emergency surgery
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Pain level was measured using a VAS Opioid consumption at 1, 2, 6, 12, 18, 24, 48 and 72 hours postoperatively
Notes	—

Jin 2021.

Methods	Prospective, double‐blind, randomised, controlled, single‐centre trial
Participants	Setting: Hospital of Shandong University, Shandong University, Jinan, Shandong, People's Republic of China Sample size: 62 adult patients Participant (baseline) characteristics: Age over 18 years Inclusion criteria: patients undergoing elective lumbar laminoplasty under general anaesthesia Exclusion criteria: emergency surgery
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Postoperative pain scores for the first 48 hours after surgery Intraoperative anaesthetic usage Perioperative analgesic consumption Return of bowel function Satisfaction for acute pain management indicated by overall benefit of analgesia score (OBAS)
Notes	—

Khan 2021.

Methods	Randomised controlled trial using the sealed envelope method
Participants	Setting: Shifa International Hospital Sample size: 82 adult patients Participant (baseline) characteristics: Age 18 to 45 years ASA I‐II Inclusion criteria: adult patients undergoing elective living related donor hepatectomies for liver transplant surgery Exclusion criteria: neurological impairment, allergy, hypersensitivity or any other contraindications to local anaesthetic, anatomical variation for block landmarks
Interventions	Intervention group: continuous ESPB Comparator group: continuous TEA
Outcomes	Postoperative pain scores using VAS for pain at rest in the postoperative period at 0, 1, 6, 12, 24 and 48 hours Dosage of adjunct opioid used Postoperative nausea and vomiting (PONV) Lung volume on incentive spirometry Haemodynamic parameters in the postoperative period
Notes	—

Kim 2021.

Methods	Prospective, randomised, controlled study
Participants	Setting: Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea Sample size: 70 adult patients Participant (baseline) characteristics: Age over 18 years Inclusion criteria: adult patients undergo laparoscopic liver resection. Exclusion criteria: —
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Cumulative postoperative opioid consumption at 24 hours (morphine equivalent) Rescue opioid (fentanyl) dose in the postanaesthesia care unit (PACU) and pain severity at 1, 6, 12, 24, 48 and 72 hours, assessed using a NRS score
Notes	—

Ozdemir 2021.

Methods	Randomised controlled, prospective trial.
Participants	Setting: Baskent University Research and Training Hospital, Konya, Turkey Sample size: 64 adult patients Participant (baseline) characteristics: Age over 18 years Inclusion criteria: adult patients undergoing elective laparoscopic cholecystectomy surgery Exclusion criteria: emergency surgery
Interventions	Intervention group: ESPB Comparator group: STAPB
Outcomes	Pain scores at rest and during movement Fentanyl requirement Postoperative walking time Duration of hospital stay
Notes	—

Park 2021a.

Methods	Prospective, randomised clinical trial approved by the Institutional Review Board of Samsung Medical Center
Participants	Setting: Samsung Medical Centre, Seoul, Korea Sample size: 58 adult patients Participant (baseline) characteristics: Age 20 to 70 years ASA status I‐II Inclusion criteria: adult women undergoing immediate breast reconstruction with a tissue expander after mastectomy Exclusion criteria: patients who refused to participate in the study and had chronic pain, psychological disorders, a known allergy to the study drugs (especially local anaesthetics), a history of drug abuse, coagulation disorders, infection at the injection site, and renal dysfunction (preoperative serum creatinine level > 2.0 mg/dL)
Interventions	Intervention group: ESPB with PCA Comparator group: no block with PCA
Outcomes	Total amount of opioid consumption during 24 hours postoperatively Patient satisfaction Pain score at rest and on shoulder movement using a numerical rating scale Incidence of postoperative nausea and vomiting (PONV) Brief pain inventory (BPI‐SF) at 3 and 6 months after surgery
Notes	—

Piskin 2021.

Methods	Prospective, randomised, controlled study
Participants	Setting: Zonguldak Bulent Ecevit University, Zonguldak, Turkey Sample size: 80 adult patients Participant (baseline) characteristics: Age 18 to 75 years ASA physical status I‐III Inclusion criteria: adult patients undergoing elective video‐assisted thoracoscopic surgery Exclusion criteria: emergency surgery
Interventions	Intervention group: continuous ESPB Comparator group: no block
Outcomes	Tramadol and meperidine consumption VAS pain scores at 0, 1, 4, 8, 12, 24, 36 and 48 hours postoperatively Opioid‐related side effects at 0, 1, 4, 8, 12, 24, 36 and 48 hours postoperatively
Notes	—

Ramachandran 2021.

Methods	Prospective, double‐blinded, randomised, parallel‐group study
Participants	Setting: in‐patient hospital Sample size: 66 adult patients Participant (baseline) characteristics: Age over 18 years Inclusion criteria: patients undergoing PCNL Exclusion criteria: patients under the age of 18 years
Interventions	Intervention group: ESPB Comparator group: subcutaneous infiltration
Outcomes	NRS scores assessed at intervals of 30 minutes, 60 minutes, then hourly for 6 hours, followed by 4‐hourly up to 24 hours Analgesic requirement Incidence of complications
Notes	—

Rao 2021.

Methods	Prospective, randomised, controlled study registered with the Australian and New Zealand Clinical Trials Registry. Institutional Human Ethics and Research Committee (HREC/18/CALHN/456) approval was obtained. Groups were randomised to the intervention allocation based on a computer‐generated sequence.
Participants	Setting: The Queen Elizabeth Hospital (TQEH), part of Central Adelaide Local Health Network (CALHN) Sample size: 72 adult patients Participant (baseline) characteristics: Age 18 to 85 years ASA I‐III Inclusion criteria: adult patients undergoing elective laparoscopic assisted colonic surgery Exclusion criteria: communication barriers, sensitivity or allergy to local anaesthetics, were pregnant, had a pre‐operative daily use of opioids equivalent to 10 mg/day of morphine or above or if the procedure could not be performed laparoscopically
Interventions	Intervention group: ESPB Comparator group: wound infiltration
Outcomes	Postoperative pain score utilising a verbal numerical rating score (NRS, 0 to 10) on rest and coughing in the postanaesthetic care unit (PACU) and in the first 24 hours Opioid usage Length of stay Clinical adverse events
Notes	—

Sarkar 2021.

Methods	Randomised controlled trial
Participants	Setting: inpatient hospital Sample size: 34 adult patients Participant (baseline) characteristics: Age 18 to 60 years ASA I‐II Inclusion criteria: adult patients undergoing unilateral percutaneous nephrolithotomy Exclusion criteria: —
Interventions	Intervention group: ESPB Comparator group: sham block
Outcomes	Total 24‐hour rescue opioid analgesic requirement Time to first demand for rescue analgesic and pain scores at 2, 12 and 24 hours after surgery
Notes	—

Shanthanna 2021.

Methods	Randomised, controlled, double‐blinded trial
Participants	Setting: inpatient hospital Sample size: 62 adult patients Participant (baseline) characteristics: Age over 18 years Inclusion criteria: adult patients undergoing arthroscopic shoulder repair under general anaesthesia Exclusion criteria: —
Interventions	Intervention group: ESPB Comparator group: sham block
Outcomes	Resting pain score in recovery Pain scores with movement Opioid use Patient satisfaction Adverse effects in hospital Outcomes at 24 hours and 1 month
Notes	—

Shen 2021.

Methods	Randomised controlled trial
Participants	Setting: inpatient hospital Sample size: 62 adult patients Participant (baseline) characteristics: Age ≥ 65 years Inclusion criteria: patients undergoing elective laparoscopic colorectal surgery with general anaesthesia Exclusion criteria: —
Interventions	Intervention group: ESPB Comparator group: TAPB
Outcomes	Visual analogue scale (VAS) pain score during the first 24 postoperative hours in resting and active states Postoperative consumption of sufentanil Satisfaction score Number of patients who required antiemetics Incidence of block‐related complications and other side effects
Notes	—

Sifaki 2021.

Methods	Double‐blinded, randomised, controlled, prospective study
Participants	Setting: inpatient hospital Sample size: 60 adult patients Participant (baseline) characteristics: Age over 18 years Inclusion criteria: adult patients undergoing elective laparoscopic cholecystectomy under general anaesthesia Exclusion criteria: emergency surgery
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Median 24‐hour morphine consumption Median NRS scores at all time points Mobilisation time
Notes	—

Taskaldiran 2021.

Methods	Randomised controlled trial
Participants	Setting: inpatient hospital Sample size: 60 adult patients Participant (baseline) characteristics: Age over 18 years Inclusion criteria: patients scheduled for elective lumbar herniated disc surgery Exclusion criteria: —
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Fentanyl consumption Visual analogue scale (VAS) scores were recorded at 15 minutes, 1, 6, 12 and 24 hours postoperatively
Notes	—

Theodoraki 2021.

Methods	Double‐blind, randomised, controlled, prospective study
Participants	Setting: inpatient hospital Sample size: 60 adult patients Participant (baseline) characteristics: Age over 18 years Inclusion criteria: patients undergoing laparoscopic cholecystectomy Exclusion criteria: —
Interventions	Intervention group: ESPB Comparator group: sham block
Outcomes	Difference in total postoperative morphine consumption Difference in total intraoperative remifentanil consumption Time spent in PACU NRS pain scores at arrival and at discharge from PACU and at 3, 6, 12 and 24 hours after surgery Postoperative nausea and vomiting Time until first mobilisation of the patient Hospitalisation days Satisfaction score of the patients (using a scale from 0 to 6) among groups
Notes	—

Verma 2020.

Methods	Double‐blinded, prospective randomised controlled trial
Participants	Setting: Department of Anaesthesiology, SGPGIMS, Lucknow, Uttar Pradesh, India Sample size: 85 adult patients Participant (baseline) characteristics: Age 18 to 60 ASA I‐II Inclusion criteria: adult patients undergoing elective laparoscopic cholecystectomy Exclusion criteria: patient refusal, morbid obesity, any lesion obscuring sonoanatomy, allergies to local anaesthetics, bleeding diathesis, use of anticoagulants and psychiatric disorders
Interventions	Intervention group: ESPB Comparator group: sham block
Outcomes	Pain score as per the VAS Intraoperative requirement of fentanyl Postoperative use of diclofenac Time to ambulation after surgery Presence of any pain after surgery
Notes	—

Wahdan 2021.

Methods	Double‐blinded, prospective randomised controlled trial Research Ethical Committee of Faculty of Medicine Cairo University (MD‐80‐2019)
Participants	Setting: Faculty of Medicine, Cairo University, Cairo, Egypt Sample size: 140 adult patients Participant (baseline) characteristics: Age 18 to 65 years ASA I‐II Inclusion criteria: patients scheduled for elective lumbar spine surgery Exclusion criteria: refusal to participate, inability to evaluate level of pain, allergy to the study drugs, morbid obesity or a procedure on more than two intervertebral spaces
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Total amount of morphine consumed intraoperatively and in the first 24 postoperative hours Time to first request for rescue analgesia and incidence of adverse effects
Notes	—

Wang 2021.

Methods	Randomised controlled trial The study was approved by the Ethics Committee of First Affiliated Hospital of Anhui Medical University (No. PJ2018‐11‐01).
Participants	Setting: University of Science and Technology of China, Hefei, China Sample size: 304 adult patients Participant (baseline) characteristics: Age over 18 years ASA I‐III Inclusion criteria: adult patients undergoing lumbar spine fusion Exclusion criteria: history of severe psychiatric illness (major depression or generalised anxiety disorder), preplanned overnight hospitalisation, pre‐existing chronic pain (lasting at least 3 months) or being opioid‐dependent
Interventions	Intervention group: ESPB Comparator group: TLIP
Outcomes	NRS pain scores in movement state and static state and the total consumption of opioids in the perioperative period during the 48 hours Side effects of opioids (nausea and vomiting, pruritus, respiratory depression) Frequency of PCA compressions Remedial analgesic administration Quality of life score during the 6 months after surgery
Notes	—

Wang 2021a.

Methods	Randomised controlled trial approved by the Ethics Committee of Zhongshan Hospital, Fudan University (No. B2020‐011R) and prospectively registered
Participants	Setting: Zhongshan Hospital, Fudan University, Shanghai, People’s Republic of China Sample size: 108 adult patients Participant (baseline) characteristics: Age over 18 years ASA I‐II Inclusion criteria: patients scheduled for elective open hepatectomy Exclusion criteria: allergy to local anaesthetics, chronic use of opioids or pregnancy
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Average cough‐elicited pain numeric rating scale (NRS) scores recorded at the 7 follow‐up time points of 20:00 on the day of surgery and 9:00 and 15:00 on postoperative day 1 to day 3 (POD1 to POD3) Opioid use and time to start oral intake postoperatively
Notes	—

Yildiz 2021.

Methods	Prospective randomised controlled trial Ethical approval from the Ethics Committee of Necmettin Erbakan University Medical Faculty (numbered 2019/185 and dated 9 January 2019)
Participants	Setting: University of Health Science, Konya City Hospital, Department of Anesthesiology and Reanimation, Konya, Turkey Sample size: 68 adult patients Participant (baseline) characteristics: Age over 18 years Inclusion criteria: adult patients undergoing laparoscopic cholecystectomy Exclusion criteria: —
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Postoperative pain intensity associated with a lower opioid requirement and significant respiratory function improvement
Notes	—

Yörükoğlu 2021.

Methods	Randomised controlled trial
Participants	Setting: Anesthesiology and Reanimation Clinic, Tatvan State Hospital, Bitlis, Turkey Sample size: 54 adult patients Participant (baseline) characteristics: Age 18 to 65 years ASA I‐II Inclusion criteria: patients scheduled for elective discectomy surgery Exclusion criteria: —
Interventions	Intervention group: ESPB Comparator group: sham block
Outcomes	Morphine consumption Numeric rating scale (NRS) scores for pain were recorded 1, 6, 12 and 24 hours after surgery
Notes	—

Zengin 2021.

Methods	Randomised controlled trial
Participants	Setting: Marmara University Medical Faculty, Istanbul, Turkey Sample size: 63 adult patients Participant (baseline) characteristics: Age over 18 years Inclusion criteria: adult patients with morbid obesity undergoing laparoscopic bariatric surgery Exclusion criteria: —
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Total opioid dose Pain score using VAS scores for 24 hours following the operation
Notes	—

Zhang 2021a.

Methods	Blinded, randomised, controlled study The study protocol was approved by the Ethics Committee of Peking University People's Hospital (2019PHB123‐01)
Participants	Setting: Peking University People's Hospital, Beijing, China Sample size: 60 adult patients Participant (baseline) characteristics: Age 18 to 75 years ASA grade I or II Inclusion criteria: patients scheduled for lumbar spinal fusion surgery Exclusion criteria: —
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Pain intensity at rest within 12 hours postoperatively using a NRS NRS pain scores at rest and on movement, postoperative opioid consumption and proportions of patients requiring opioids during the first 48 hours after surgery
Notes	—

Zhang 2021b.

Methods	Randomised controlled trial, using random numbers sealed in envelopes
Participants	Setting: Beijing Hospital, National Center of Gerontology, Beijing, China Sample size: 60 adult patients Participant (baseline) characteristics: Age 18 to 80 years ASA I‐III Inclusion criteria: adult patients scheduled to undergo lumbar spine surgery with general anaesthesia Exclusion criteria: refusal to participate in the study, ASA physical classification of IV or higher, preoperative cognitive dysfunction or communication disorder, infection at the puncture site, allergy to amide‐type local anaesthetics, neuromuscular disease, hepatic or renal insufficiency, emergency surgery, chronic opioid use or participation in another clinical study
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Time to first PCA bolus Morphine consumption at 24 and 48 hours post operation Numerical rating scale (NRS) scores at 24 and 48 hours post operation Intraoperative sufentanil and cisatracurium consumption Duration of stay in the PACU and time to first ambulation after surgery
Notes	—

Zhang 2021c.

Methods	Randomised controlled study approved by the Institutional Review Board of Zhongshan Hospital, Fudan University (B2019‐ 074R)
Participants	Setting: Zhongshan Hospital, Fudan University, Shanghai, China Sample size: 120 adult patients Participant (baseline) characteristics: Age 20 to 70 years Inclusion criteria: adult patients undergoing elective VATS Exclusion criteria: pre‐existing infection at the block site, history of chronic pain, significant coagulopathy, contraindication to techniques or drugs used in the protocol and conversion to open thoracotomy
Interventions	Intervention group: ESPB Comparator groups: TEA and PCA
Outcomes	Pain severity Time to first postoperative analgesia requirement Intraoperative and postoperative analgesic requirements
Notes	—

Zhao 2021.

Methods	Randomised controlled trial, using computer‐generated random numbers table
Participants	Setting: Beijing Jishuitan Hospital, Beijing, China Sample size: 80 adult patients Participant (baseline) characteristics: Aged 18 to 65 years ASA I‐III Inclusion criteria: patients with multiple rib fractures undergoing surgery under general anaesthesia Exclusion criteria: bilateral chest trauma, rib fracture < 3 or > 6 sites, local anaesthetic drug allergy, local infection at the puncture site, cardiac insufficiency, renal insufficiency, nervous system abnormality or coagulation dysfunction
Interventions	Intervention group: ESPB Comparator group: RLB
Outcomes	Intraoperative remifentanil Morphine dosage Pain assessment Blood gas analysis Adverse reactions before and after surgery
Notes	—

Zhu 2021.

Methods	Prospective, randomised controlled study, single‐centre Approved by the ethics committee of Taizhou Hospital of Zhejiang Province (approval number K20190801)
Participants	Setting: Taizhou Hospital of Zhejiang Province Affiliated to Wenzhou Medical University, Zhejiang Province, China Sample size: 40 adult patients Participant (baseline) characteristics: Age 45 to 70 years ASA I‐II Inclusion criteria: adult patients undergoing posterior lumbar fusion surgery Exclusion criteria: BMI ≥ 30 kg/m2, known allergies to local anaesthetic drug, infection near the puncture site, abnormal coagulation function, use of painkillers or communication difficulties with the medical staff
Interventions	Intervention group: ESPB Comparator group: sham block
Outcomes	Total dosage of oxycodone Remifentanil consumption Postoperative pain scores Postoperative adverse events Safety and range of cold hypoesthesia
Notes	—

Zimmerer 2021.

Methods	Randomised controlled trial
Participants	Setting: University Medicine Greifswald, Greifswald, Germany Sample size: 68 adult patients Participant (baseline) characteristics: Age over 18 years Inclusion criteria: adult patients undergoing arthroscopic therapy for femoroacetabular impingement syndrome Exclusion criteria: —
Interventions	Intervention group: ESPB Comparator group: sham
Outcomes	Pain scores (numeric pain score out of 10) during the first 24 hours postoperatively Opioid consumption during the first 24 hours (converted to morphine equivalents) Incidence of PONV
Notes	—

Abbreviations:

ASA: American Society of Anesthesiology
CI: confidence interval
BMI: body mass index
ESPB: erector spinae plane block
IQR: interquartile range
IV: intravenous
NRS: numerical rating scale
PACU: post‐anaesthesia care unit 
PCA: patient‐controlled analgesia 
PCNL: percutaneous nephrolithotomy
PECS: pectoralis plane
PONV: postoperative nausea and vomiting
PVB: paravertebral block
QLB: quadratus lumborum block
RLB: retrolaminar block
SD: standard deviation
STAPB: subcostal transversus abdominis plane block
TAPB: transversus abdominis plane block
TEA: thoracic epidural analgesia
TLIP: thoracolumbar interfascial plane block
US: ultrasound
VAS: visual analogue scale
VATS: video‐assisted thoracic surgery

Characteristics of ongoing studies [ordered by study ID]

Sarkar 2022.

Study name	Comparison of quadratus lumborum block and erector spinae plane block on postoperative analgesic effect in patients undergoing laparoscopic nephrectomy ‐ a randomized controlled trial
Methods	Randomised controlled trial using computer‐generated randomisation and sequentially numbered, sealed, opaque envelopes. Participants and outcome assessors are blinded.
Participants	Setting: All India Institute of Medical Science, Bhubaneswar Sample size: 62 adult patients Participant (baseline) characteristics: Age 18 to 80 years ASA I‐II functional status Laparoscopic nephrectomy under general anaesthesia Inclusion criteria: adult patients with ASA I‐II functional status undergoing laparoscopic nephrectomy under general anaesthesia under the Department of Urology in AIIMS, Bhubaneswar. Patients who can understand and provide a valid consent for the procedure and also can understand and are willing to report the pain using a NRS score. Exclusion criteria: patients who are currently suffering from chronic pain syndromes and are already under treatment with daily analgesics. Patients with local anaesthetic allergy. Patients in whom neuraxial/regional blocks are contraindicated due to coagulopathies. Patients in whom the sonoanatomy of the anaesthetic injection site is not clearly delineable for any reason.
Interventions	Intervention group: ESPB Comparator group: QLB
Outcomes	Unknown
Starting date	15 September 2021 ‐ not yet recruiting
Contact information	Monalisa Sarkar Address: Department of Anesthesiology and Critical Care, 751019 Khordha, ORISSA, India Email: mitra.jayanta@gmail.com Affiliation: All India Institute of Medical Science, Bhubaneswar
Notes	Website source:http://www.ctri.nic.in/Clinicaltrials/pmaindet2.php?trialid=59471

Singh 2022.

Study name	Efficacy of ultrasound‐guided erector spinae plane block on postoperative quality of recovery and analgesia after lumbar spine surgery.
Methods	This is a randomised, parallel group, placebo controlled trial, using computer‐generated randomisation and sequentially numbered, sealed, opaque envelopes. Participants and investigators were blinded.
Participants	Setting: Government Medical College and Hospital, Sector 32, Chandigarh, India. Sample size: 70 adult patients Participant (baseline) characteristics: Age 18 ‐ 65 years ASA I‐II Inclusion criteria: Patients undergoing lumbar surgeries in general anaesthesia. Patients with an American Society of Anesthesiologists (ASA) Grade I or II and a body mass index >18 and <35. Exclusion criteria: Patients with a history of any drug allergies, a history of substance abuse or chronic analgesia use. Pregnant women. Patients with severe cardiovascular, respiratory, neurological or metabolic disease. Infection at the planned site. Psychological inability of the patient to understand the QoR ‐15 questionaire or VAS scale or inability to operate a Patient controlled analgesia (PCA) pump.
Interventions	Intervention group: ESPB Comparator group: no block
Outcomes	Unknown
Starting date	20 June 2021 ‐ not yet recruiting
Contact information	Dr Jasveer Singh Address: Department of Anaesthesia and Intensive Care, Level‐5, D Block, Government Medical College and Hospital, Sector 32, Chandigarh 160030 Chandigarh, CHANDIGARH, India Email: drjassy18@gmail.com Affiliation: Government Medical College and Hospital, Chandigarh
Notes	—

Zengin 2022.

Study name	Comparison of erector spinae plane block and combination of deep and superficial serratus anterior plane block
Methods	Parallel‐group randomised controlled trial
Participants	Setting: Ankara Atatürk Chest Disease and Chest Surgery Training and Research Hospital Sample size: 60 adult patients Participant (baseline) characteristics: Age 18 to 65 years ASA I‐III BMI 18 to 30 Inclusion criteria: patients aged 18 to 65 years old, undergoing elective video‐assisted thoracoscopic surgery. Patients with ASA grade I‐III and BMI 18 to 30. Exclusion criteria: patient refusal of the procedure or emergency surgery; history of chronic analgesic or opioid therapy; history of local anaesthetic allergy; infection in the intervention area
Interventions	Intervention group: ESPB Comparator group: SAPB
Outcomes	Unknown
Starting date	6 June 2021
Contact information	Musa Zengin Ankara Atatürk Chest Disease and Chest Surgery Training and Research Hospital, Kecioren, Ankara, Turkey, 06000
Notes	—

Abbreviations:

ASA: American Society of Anesthesiology
BMI: body mass index
ESPB: erector spinae plane block
QLB: quadratus lumborum block
SAPB: Serratus anterior plane block

Differences between protocol and review

We adhered to the protocol as much as possible while putting together this review but in a few instances we have had to deviate slightly from protocol (Schnabel 2020). The differences between this review and the registered protocol are as follows:

1. Data extraction and synthesis as well as risk of bias assessment were performed by two authors (LO, AS).

2. We excluded trials with asymmetrical data distribution from the meta‐analysis.

3. We had to exclude one study that we were unable to translate from Chinese.

4. One of our prespecified outcomes was chronic pain at three and six months following an operation, however this was not reported in the included studies.

5. We contacted study authors whenever possible to ask about missing data or other potentially unreported data in the included studies, however we have only received two replies so far.

6. Given the recent concerns about trial authenticity, we performed a search to identify any retracted or altered studies, and found one retracted study (Tulgar 2019), which we subsequently excluded from the review.

7. We restricted 95% prediction interval (95% PI) and zero event analyses, which we only performed for outcomes with more than four trials.

8. Similarly, we restricted subgroup analysis to only one group that had more than 10 trials. In the protocol we planned to perform subgroup analysis focusing on risk of bias, attrition over 15%, fixed‐effect versus random‐effects meta‐analysis and zero events. We did not perform subgroup analysis for zero events due to an insufficient number of trials, and we performed zero event analysis in R instead. We also did not perform subgroup analysis focusing on attrition because only one study had attrition over 15% (Shim 2020).

9. In the protocol, we did not define a clinically significant difference in morphine consumption, and we retrospectively agreed on a 10 mg oral morphine equivalent.

10. For postoperative pain outcomes we decided to use the visual analogue scale (VAS) in the review, since this was the most commonly used scale in the included studies. All included scales were 1‐ to 10‐point scales, so we used the mean difference (MD) instead of the standardised mean difference (SMD). Due to high heterogeneity we used random‐effects meta‐analysis.

11. We only created funnel plots for primary outcomes with more than 10 trials, in order to help with GRADE assessment (Grading of Recommendations, Assessment, Development and Evaluations).

12. We did not search the grey literature as planned in the protocol; instead we focused our search on the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, Embase and Web of Science.

13. We changed the label of the first comparison from erector spinae plane block (ESPB) versus opioid treatment to ESPB versus no block because all patients received opioids as part of the standard care. We also changed the label of the second comparison (placebo instead of sham block).

Contributions of authors

Alexander Schnabel (AS), Stephanie Weibel (SW), Esther Pogatzki‐Zahn (EPZ), Christine Meyer‐Friessem (CMF), Lisa Oostvogels (LO).
Conceiving the review: AS.
Co‐ordinating the review: AS.
Writing the protocol: AS, SW.
Securing funding for the review: AS.
Performing previous work that served as the foundation of the present study: AS, EPZ.
Serving as guarantor for the review (one review author): AS.
Taking responsibility for reading and checking the review before submission: AS, SW, EPZ, LO.
Writing the review: LO, AS, SW, CMF.
Statistics using R: SW

Sources of support

Internal sources

• Internal source of support, Other

  Intradepartmental funding only

External sources

• None, Other

  No external sources of support

Declarations of interest

Alexander Schnabel: none known.
Stephanie Weibel: none known.
Christine Meyer‐Frießem: none known.
Esther Pogatzki‐Zahn: has no conflicts of interest regarding the topic of this review. She received lecture fees from Grünenthal, Metronic and Novartis and research support from Grünenthal outside the submitted work (all payments were made to her institution, not to EPZ personally); EPZ currently receives scientific support (unrelated and outside the submitted work) from the DFG, the BMBF, the ERANET ‐ Neuro (EU), the G‐BA and the Innovative Medicines Initiative 2 Joint Undertaking under grant agreement No 777500 (IMI‐PainCare). This Joint Undertaking receives support from the European Union’s Horizon 2020 research and innovation programme and EFPIA. All funding was unrelated to the present research.
Lisa Oostvogels: none known.

New