Adductor canal blocks for postoperative pain treatment in adults undergoing knee surgery

Alexander Schnabel; Sylvia U Reichl; Stephanie Weibel; Peter K Zahn; Peter Kranke; Esther Pogatzki‐Zahn; Christine H Meyer‐Frießem

doi:10.1002/14651858.CD012262.pub2

. 2019 Oct 26;2019(10):CD012262. doi: 10.1002/14651858.CD012262.pub2

Adductor canal blocks for postoperative pain treatment in adults undergoing knee surgery

Alexander Schnabel ^1,^✉, Sylvia U Reichl ², Stephanie Weibel ³, Peter K Zahn ⁴, Peter Kranke ³, Esther Pogatzki‐Zahn ¹, Christine H Meyer‐Frießem ⁴

Editor: Cochrane Anaesthesia Group

PMCID: PMC6814953 PMID: 31684698

Abstract

Background

Peripheral regional anaesthesia techniques are well established for postoperative pain treatment following knee surgery. The adductor canal block (ACB) is a new technique, which can be applied as a single shot or by catheter for continuous regional analgesia.

Objectives

To compare the analgesic efficacy and adverse events of ACB versus other regional analgesic techniques or systemic analgesic treatment for adults undergoing knee surgery.

Search methods

We searched CENTRAL, MEDLINE, and Embase, five other databases, and one trial register on 19 September 2018; we checked references, searched citations, and contacted study authors to identify additional studies.

Selection criteria

We included all randomized controlled trials (RCTs) comparing single or continuous ACB versus other regional analgesic techniques or systemic analgesic treatment. Inclusion was independent of the technique used (landmarks, peripheral nerve stimulator, or ultrasound) and the level of training of providers.

Data collection and analysis

We used Cochrane’s standard methodological procedures. Our primary outcomes were pain intensity at rest and during movement; rate of accidental falls; and rates of opioid‐related adverse events. We used GRADE to assess the quality of evidence for primary outcomes.

Main results

We included 25 RCTs (1688 participants) in this review (23 trials combined within meta‐analyses). In 18 studies, participants underwent total knee arthroplasty (TKA), whereas seven trials investigated patients undergoing arthroscopic knee surgery. We identified 11 studies awaiting classification and 11 ongoing studies.

We investigated the following comparisons.

ACB versus sham treatment

We included eight trials for this comparison. We found no significant differences in postoperative pain intensity at rest (2 hours: standardized mean difference (SMD) ‐0.56, 95% confidence interval (CI) ‐1.20 to 0.07, 4 trials, 208 participants, low‐quality evidence; 24 hours: SMD ‐0.49, 95% CI ‐1.05 to 0.07, 6 trials, 272 participants, low‐quality evidence) or during movement (2 hours: SMD ‐0.59, 95% CI ‐1.5 to 0.33; 3 trials, 160 participants, very low‐quality evidence; 24 hours: SMD 0.03, 95% CI ‐0.26 to 0.32, 4 trials, 184 participants, low‐quality evidence). Furthermore, they noted no evidence of a difference in postoperative nausea between groups (24 hours: risk ratio (RR) 1.91, 95% CI 0.48 to 7.58, 3 trials, 121 participants, low‐quality evidence). One trial reported that no accidental falls occurred 24 hours postoperatively (low‐quality evidence).

ACB versus femoral nerve block

We included 15 RCTs for this comparison. We found no evidence of a difference in postoperative pain intensity at rest (2 hours: SMD ‐0.74, 95% CI ‐1.76 to 0.28, 5 trials, 298 participants, low‐quality evidence; 24 hours: SMD 0.04, 95% CI ‐0.09 to 0.18, 12 trials, 868 participants, high‐quality evidence) or during movement (2 hours: SMD ‐0.47, 95% CI ‐1.86 to 0.93, 2 trials, 88 participants, very low‐quality evidence; 24 hours: SMD 0.56, 95% CI ‐0.00 to 1.12, 9 trials, 576 participants, very low‐quality evidence). They noted no evidence of a difference in postoperative nausea (24 hours: RR 1.22, 95% CI 0.42 to 3.54, 2 trials, 138 participants, low‐quality evidence) and no evidence that the rate of accidental falls during postoperative care was significantly different between groups (24 hours: RR 0.20, 95% CI 0.04 to 1.15, 3 trials, 172 participants, low‐quality evidence).

Authors' conclusions

Plain language summary

Advantages and problems of a specific nerve block in adults undergoing knee surgery

Background

Postoperative pain following knee surgery continues to be a relevant healthcare problem. Combinations of different analgesics are the best way to treat postoperative pain. One way is to block specific nerves (called regional anaesthesia) that are responsible for pain development. For many years, blocking the femoral nerve, which is responsible for sensation (e.g. pain) and movement of the upper leg, was very important. In recent years, blocking only one specific part of this nerve (called adductor canal block), which does not influence movement of the upper leg, has become more interesting.

Review question

We investigated advantages and problems of the adductor canal block compared to sham treatment (patients received saline instead of drugs) and other regional anaesthesia for postoperative pain treatment in adults undergoing knee surgery.

Study characteristics

We included 25 clinical studies in which people are randomly put into one of two or more treatment groups (called 'randomized controlled trials'), with results reported from a total of 1688 participants (929 females, 759 males). Participants were 29 to 72 years old. Eight trials compared participants receiving adductor canal block against patients receiving saline. A total of 15 RCTs compared adductor canal block versus femoral nerve block. The evidence is current to October 2018. No trial was funded by industry.

Key results

We are uncertain whether patients treated with adductor canal block have lower pain intensity at rest or during movement (e.g. walking) compared with those who received only saline. It is unclear whether rates of adverse events after taking opioids (e.g. nausea) or after accidental falls during postoperative care are lower. It is also uncertain whether patients receiving adductor canal block show different postoperative pain intensity at rest and during movement compared to those treated with femoral nerve block. We noted no differences in adverse events after taking opioids and after accidental falls.

Quality of the evidence

We rated the quality of evidence for many outcomes as low or very low. In contrast, we rated pain at rest (at 24 hours) as high‐quality evidence.

Summary of findings

Background

Description of the condition

Knee surgery (e.g. knee arthroplasty, arthroscopic knee surgery) is very commonly performed in western countries (knee replacement: USA 650,000 (2010); Germany 156,000 (2012)). Major goals following knee surgery include providing sufficient postoperative pain treatment to assist early physical therapy and allowing patients to return early to their physical capacity and to be discharged early from the hospital. Patients suffer from moderate to severe postoperative acute pain (Gerbershagen 2013), and if this pain is insufficiently treated, it might become chronic (Althaus 2014; Pogatzki‐Zahn 2012). Recently published data demonstrate that the incidence of chronic pain in adults undergoing total knee replacement is 10% to 34% after three months to five years on follow‐up pain measurement (Beswick 2012), and around 20% of patients describe moderate to severe sleep disturbances and alterations in quality of life one year after surgery (Grosu 2015). Finally, clear evidence suggests that use of regional analgesia, especially in joint arthroplasty surgery (Guay 2017; Guay 2017a), is associated with superior postoperative outcomes (pulmonary compromise, pneumonia, infection, acute renal failure, mechanical ventilation, blood product transfusion) (Memtsoudis 2013), and it might reduce the risk of chronic postsurgical pain (Weinstein 2018).

Description of the intervention

In recent years, adductor canal block through selective block of sensory nerves has become an interesting new option for postoperative pain treatment following knee surgery. The knee is innervated by the femoral nerve (via three vasti branches and the saphenous nerve), the posterior branch of the obturator nerve, and genicular branches of the tibial and common peroneal branches of the sciatic nerve (Bendtsen 2014a). The adductor canal includes the femoral vessels, the saphenous nerve, a nervous branch to the vastus medialis muscle, and sometimes the posterior branch of the obturator nerve (Bendtsen 2014b). The adductor canal is roofed by continuous fascia starting with the vasoadductor membrane distally (Andersen 2015). Adductor canal block, which is performed most often via ultrasound, can be used as a single shot or as continuous nerve block provided through a catheter.

How the intervention might work

Postoperative pain following knee surgery can be managed with systemic analgesics or regional blockade (neuraxial blockade or peripheral nerve blocks). Neuraxial blocks (e.g. epidural catheters) are used less frequently for postoperative pain treatment following knee surgery; distal peripheral nerve blocks (e.g. femoral nerve blocks) are performed more frequently because they involve lower risk for severe adverse events (e.g. epidural bleeding) (Cozowicz 2015). For a long time, femoral nerve block was the gold standard regional analgesic technique for postoperative pain treatment following knee surgery (Chan 2014). However, adductor canal block might be associated with a lower degree of motor blockade than femoral nerve block, and might provide better conditions for early rehabilitation, quicker return to mobility, and less risk for accidental falls during hospital care compared with femoral nerve block (Mariano 2014). It must be mentioned that two other large studies have indicated that appropriate fall prevention strategies should be used for all hospitalized patients, even those not receiving regional blockade (Johnson 2014), and it is not clear whether regional analgesia definitively increases risk for inpatient falls (Memtsoudis 2014). After the femoral vessels have been identified, the saphenous nerve might be blocked typically at two locations: subsartorially, or more distally within the adductor canal. Cadaveric studies have demonstrated that dye is normally spread freely into the adductor canal after a subsartorial injection, so that the primary injection site might not be clinically relevant for clinical efficacy (Cowlishaw 2015; Tubbs 2007). Several cadaveric studies have revealed that a small amount of dye spreads to other nerves as well (e.g. sciatic, femoral), so that possible motor blockade cannot be definitively excluded (Andersen 2015; Cowlishaw 2015; Gautier 2015).

Why it is important to do this review

In patients undergoing knee surgery, femoral nerve block (and epidural catheter for special cases such as bilateral knee arthroplasty) is believed to be the gold standard for acute pain management because it provides better analgesia than is provided by systemic analgesic treatment for adults undergoing knee surgery (Chan 2014). However, this block might be associated with a higher degree of motor blockade, possibly increasing the risk for inpatient falls (Johnson 2013; Wasserstein 2013). As has been mentioned, evidence regarding use of femoral nerve block and risk for inpatient falls is currently inconclusive. Many RCTs published in recent years have compared analgesic efficacy and safety between adductor canal block and other regional analgesic techniques (particularly femoral nerve block). A quantitative systematic review has not been conducted to analyse analgesic efficacy and adverse effects of adductor canal block compared with other regional analgesic techniques or systemic analgesic treatment for patients undergoing knee surgery.

Objectives

To compare the analgesic efficacy and adverse events of adductor canal block versus other regional analgesic techniques or systemic analgesic treatment for adults undergoing knee surgery.

Methods

Criteria for considering studies for this review

Types of studies

We included all randomized controlled trials (RCTs) investigating adductor canal block in comparison with other regional analgesic techniques or systemic analgesic treatment. Cluster RCTs, cross‐over RCTs, and quasi‐RCTs were not included.

Types of participants

We included all adults (≥ 18 years old) undergoing knee surgery, irrespective of sex or type of surgery.

Types of interventions

We included all RCTs comparing single or continuous adductor canal block versus sham treatment (patients received saline instead of local anaesthetics), single or continuous femoral nerve block, or any other regional anaesthetic technique. Inclusion was independent of the technique used (landmarks, peripheral nerve stimulator, or ultrasound) and the level of training of providers.

Types of outcome measures

Primary outcomes

Mean difference in postoperative pain at rest/during movement (2 hours (within the postoperative care unit), 24 hours, 48 hours)
Rates of opioid‐related adverse events (nausea, vomiting, postoperative nausea and vomiting, pruritus, respiratory depression, sedation (2 hours (within the postoperative care unit), 24 hours, 48 hours))
Rate of accidental falls during postoperative care

Secondary outcomes

Cumulative mean morphine requirement (2 hours (within the postoperative care unit), 24 hours, 48 hours)
Degree of quadriceps muscle strength (2 hours (within the postoperative care unit), 24 hours, 48 hours)
Rate of chronic postsurgical pain (after 3 months, 6 months, 1 year)
Rates of block‐related adverse events (accidental vascular puncture, paraesthesia, motor blockade, failed block, neurological impairment)

We applied no restrictions regarding the scales that were used to measure pain and quadriceps muscle strength.

Search methods for identification of studies

Electronic searches

We searched for studies through systematic and sensitive search strategies, as described in the Cochrane Handbook for Systematic Reviews of Interventions, Chapter 6 (Higgins 2011). We applied no language, publication year, or publication status restrictions. We searched the following databases.

Cochrane Central Register of Controlled Trials (CENTRAL) (2018, Issue 8), in the Cochrane Library.
MEDLINE (Ovid SP, 1946 to 19 September 2018).
Embase (Ovid SP, 1974 to 19 September 2018).
Web of Science (1945 to 19 September 2018).

We developed a subject‐specific search strategy for MEDLINE and modified it appropriately for the other databases. When appropriate, we used the highly sensitive search strategy designed by Cochrane for identifying RCTs and controlled clinical trials, as described in theCochrane Handbook for Systematic Reviews of Interventions, Chapter 6 (Lefebvre 2011). Search strategies can be found in Appendix 1, Appendix 2, Appendix 3, and Appendix 4. Searches were last run 19 September 2018.

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), along with the World Health Organization International Clinical Trials Registry Platform (WHO ICTRP) (apps.who.int/trialsearch/), for unpublished and ongoing studies; Open Grey for grey literature (http://opengrey.eu/); and Google Scholar for additional trials (25 February 2018). When necessary, we contacted trial authors for additional information. We developed the search strategy in consultation with the Information Specialist.

Data collection and analysis

Three review authors (AS, CMF, SR) independently scanned article titles to exclude irrelevant studies.

Selection of studies

The same three review authors (AS, CMF, SR) identified studies that might be included in this review. We applied no restrictions according to publication type or language. If we encountered disagreements, we consulted a third review author (EPZ) and resolved all differences by discussion. We recorded the selection process in sufficient detail to complete a PRISMA flow diagram (Moher 2009), as well as a Characteristics of excluded studies table.

Data extraction and management

Four review authors (AS, SR, CMF, SW) independently extracted data using a standardized data extraction form developed by the review authors. If necessary, we tried to receive missing data by contacting the leading authors of 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 (CMF, SR) independently assessed risk of bias of included studies by using the Cochrane tool for assessing risk of bias (Review Manager 2014). Standard components of domains included adequacy of allocation generation (random sequence generation (e.g. computer‐generated table)); allocation concealment (e.g. SNOSE (sequentially numbered opaque sealed envelopes)); blinding of participants, personnel dealing directly with participants, and outcome assessors; completeness of outcome data (e.g. no missing outcome data, description of reasons for missing data); possible selected outcome reporting (reporting of primary outcome data (at least postoperative pain scores)); and any other potential sources of bias (e.g. extreme baseline imbalance). We assessed every component as having 'low risk' of bias, 'high risk' of bias, or 'unclear risk' of bias. Within the current review, we have provided a 'Risk of bias' graph as part of the Characteristics of included studies table and a 'Risk of bias' summary figure, which summarize risk of bias assessments for all included studies. Both responsible review authors resolved disagreements by discussion with a third review author (AS).

Measures of treatment effect

For proportions (dichotomous outcomes), we calculated the risk ratio (RR) with 95% confidence interval (CI), and for continuous data, we estimated the mean difference (MD) with 95% CI. For the outcome 'postoperative pain', we used the standardized mean difference (SMD) as a summary statistic in meta‐analysis because we did not transform results based on a numerical rating scale (NRS) or a visual analogue scale (VAS). For the outcome 'cumulative postoperative morphine consumption', we converted all reported opioids into intravenous morphine equivalents by using an opioid conversion table (http://opioidcalculator.practicalpainmanagement.com/). To estimate the statistical significance of these results, we calculated the 95% CI for each item. Furthermore, 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 (> 4 trials per outcome).

We considered a difference of 10% (increase or decrease) as the minimum clinically relevant difference, but for rare outcomes such as inpatient falls, we assumed that a difference of 1% was clinically relevant. For SMDs, we considered 0.2 a small effect, 0.5 a medium effect, and > 0.8 a large effect (Pace 2011).

The protocol reports a plan to perform a trial sequential analysis (TSA) to calculate the required information size (IS; number of participants needed in a meta‐analysis to detect or reject a certain intervention effect) and the sequential monitoring boundaries (testing for statistical significance before the IS has been reached) for primary dichotomous outcomes (rates of opioid‐related adverse events, rate of accidental falls). Both the IS and the monitoring boundaries provide information relevant to estimation of the level of evidence for the experimental intervention, as cumulative meta‐analyses are at risk of producing type I errors as a result of sparse data and repetitive testing of accumulating data (Brok 2008; Thorlund 2009; Wetterslev 2008; Wetterslev 2009). Given that all dichotomous outcomes of this review included only a small number of participants (< 400 participants) and estimated effects included the line of no effect in all cases, TSA does not provide any new information. We downgraded results for all dichotomous outcomes for imprecision by one level.

For the primary continuous outcome of pain (summary statistic: SMD), we calculated the optimal information size (OIS), which is similar to a sample size calculation for an individual trial, if more than 200 participants were included for that outcome (Brant 2005).

Dealing with missing data

If we identified missing data (patient dropouts, selective outcome reporting), we contacted relevant study authors to request further information. We performed sensitivity analyses focused on the possible influence of these missing data by inputting missing data as 'best case' or 'worst case' scenarios, if these data were rated as relevant. If missing data were randomly distributed between experimental and control groups, we included in the meta‐analysis only data on participants with known results. Finally, we analysed the possible influence of studies with incomplete outcome reporting within a sensitivity analysis. We calculated missing standard deviations from standard errors or CIs, as described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). If data were reported as median values with interquartile ranges, we assumed that the median was very similar to the mean when the distribution of data was symmetrical, and we used the median directly in the meta‐analysis and calculated the standard deviation from the interquartile range in accordance with Higgins 2011. We did not pool asymmetrical data for meta‐analysis.

Assessment of heterogeneity

We assessed clinical and methodological differences within included studies to decide whether studies were sufficiently homogeneous to be combined. Within subgroup analyses, we analysed the influence of clinical heterogeneity according to surgery (total knee replacement vs arthroscopic knee surgery), local anaesthetic dose, continuous versus single shot technique, and location of the adductor canal block (proximal vs distal). According to Higgins 2011, we performed subgroup analyses only if more than 10 trials were included for this outcome. We reported statistical heterogeneity using the I² statistic. We calculated this value for each of the outcomes listed above and assessed the extent of heterogeneity as low (< 25%), moderate (25% to 50%), or high (> 75%), depending on the value of the I² statistic (Higgins 2003).

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 or to detect possible reporting bias and small‐study effects. If asymmetry was suggested by visual assessment, we investigated by performing exploratory analyses (e.g. Arcsine test for binary data, Egger´s test for continuous data). To adjust for small‐study effects, we used Duval and Tweedie's trim and fill method. We performed all statistical tests for publication bias by using R software (R package: meta, metasens).

Data synthesis

For dichotomous data, we used the Mantel‐Haenszel method, and for continuous data, we used the inverse variance method in Review Manager 2014. 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 methods were judged sufficiently similar). When clinical heterogeneity was sufficient to suggest that underlying treatment effects differed between trials, or when we detected substantial statistical heterogeneity (> 50%), we used random‐effects meta‐analysis to produce an overall summary if an average treatment effect across trials was considered clinically meaningful. However, after taking into account that study weights were 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 (Cochrane Handbook for Systematic Reviews of Interventions, Section 16.9.2) (Higgins 2011). RevMan ignores zero/zero event trials and uses a constant continuity correction of 0.5 for studies with zero events in one arm. Excluding such trial data potentially creates the risk of inflating the magnitude of the pooled treatment effect. We included zero total event trials to take into account the sample sizes of these studies. To assess the robustness of estimated treatment effects, we will perform alternative non‐fixed zero‐cell corrections that have been explored by Sweeting and colleagues, including a correction proportionate to the reciprocal of the size of the contrasting study arm, which these investigators found preferable to the fixed 0.5 correction when arm sizes were not balanced (Sweeting 2004). We performed different types of continuity corrections using TSA software v0.9 Beta (Thorlund 2011), and we have presented these corrections in a sensitivity analysis.

We reported summary RRs, MDs, and SMDs along with 95% CIs. We considered RRs, with the range of lower and upper bounds of the 95% CI not crossing one, and MDs, respectively, as well as SMDs with the range of lower and upper bounds of the 95% CI not crossing zero, to be statistically significant (P < 0.05).

Subgroup analysis and investigation of heterogeneity

We investigated the influence of clinical and methodological heterogeneity. We performed subgroup analyses to calculate RR, MD, or SMD in conjunction with corresponding CI for each subgroup, if heterogeneity exceeds 50%. We used a random‐effects model Chi² test of heterogeneity to compare subgroups. Additionally, we considered non‐overlapping subgroup CIs as consistent with a statistically significant difference.

We analysed data pertaining to the following subgroups, if available.

Type of surgery (total knee replacement vs arthroscopic knee surgery).
Type of local anaesthetic (long‐ vs short‐lasting vs mixture of local anaesthetics).
Continuous versus single shot regional analgesia.
Location of adductor canal block (proximal vs distal).
Type of anaesthesia technique (general anaesthesia, neuraxial anaesthesia).
Use of perioperative non‐opioid analgesics.
Use of sciatic nerve block.

Sensitivity analysis

We performed sensitivity analyses focused on the following issues.

Influence of study quality, by excluding trials assessed as having high risk of bias for random sequence generation/allocation concealment and blinding.
Influence of incomplete outcome data reporting, by inputting missing participants in 'best case' versus 'worst case' scenarios.
Effect estimate under the fixed‐effect model.
Influence of inclusion of randomized trials with zero events.

'Summary of findings' table and GRADE

We used the GRADE approach to rate the quality of evidence and the grading strength of recommendations in healthcare associated with the following (primary) outcomes in our review (Guyatt 2011a; Guyatt 2011b).

Mean difference in postoperative pain.
Rates of opioid‐related adverse events.
Rate of accidental falls during postoperative care.

We constructed 'Summary of ﬁndings' tables using GRADE software (www.gradepro.org). Through the GRADE approach, we appraised the quality of evidence on the basis of the extent to which one can be conﬁdent that the estimate of effect reﬂects the item assessed. The quality of the body of evidence reﬂects within‐study risk of bias (methodological quality), indirectness, heterogeneity of the data (inconsistency), imprecision of effect estimates, risk of publication bias, and magnitude of effect.

For risk of bias, we judged the quality of evidence as adequate when most information was derived from studies at low risk of bias; we downgraded the quality by one level when most information was provided by studies at high or unclear risk of bias; and we downgraded the quality by two levels when the proportion of data from studies at high risk of bias was sufficient to affect interpretation of results (sensitivity analysis) (Guyatt 2011c).

For inconsistency, we downgraded the quality of evidence by one level when the I² statistic was 50% or higher without satisfactory explanation (subgroup analysis), and by two levels when the I² statistic was 75% or higher with no explanation (Guyatt 2011c).

We judged the quality of evidence for indirectness as adequate if outcome data were based on direct comparisons of interest, on the population of interest, and on the outcome of interest (not surrogate markers) (Guyatt 2011d).

If the 95% CI excluded a risk ratio of 1.0 or an SMD of 0.0, and the total number of participants exceeded the IS (RR) or OIS (SMD) criterion, precision was adequate (Guyatt 2011e); we did not downgrade if the 95% CI was narrow and included a risk ratio of 1.0 or an SMD of 0.0 (no appreciable difference between treatments), or if the total number of participants exceeded the IS or OIS criterion. We downgraded the quality of evidence for imprecision by one level when the confidence interval around the effect size was large or overlapped an absence of effect and failed to exclude an important benefit or harm, and when the number of participants was smaller than the required information size (IS or OIS), or the monitoring boundaries were not crossed (see TSA). We generally downgraded the evidence by one level if fewer than 400 patients were included for dichotomous outcomes and if 200 patients were included for continuous outcomes.

For publication bias (Guyatt 2011f), we downgraded the quality of evidence by one level if the statistical test for funnel plot asymmetry suggested publication bias, and if the adjustment for small‐study effects as assessed by Duval and Tweedie’s ﬁll and trim analysis changed the conclusion. We downgraded the level of evidence for publication bias by two levels if most trials were small and were industry sponsored.

The GRADE assessment resulted in one of four levels of 'quality'; these expressed our confidence in the estimate of effect (Balshem 2011).

High: further research is very unlikely to change our confidence in the estimate of effect.
Moderate: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low: any estimate of effect is very uncertain.

Results

Description of studies

Results of the search

We identified 846 related articles by searching electronic databases (Figure 1). After reviewing the titles, we selected 55 articles for abstract review, of which we excluded 19 articles and determined that 11 trials were very new trials currently awaiting assessment. Finally, 25 studies including 1688 participants met the inclusion criteria of this review. All studies were RCTs using a parallel group design. One group selected additionally a cross‐over design ‐ Memtsoudis 2015 (see Characteristics of included studies table).

Included studies

We included 25 RCTs. Please see the Characteristics of included studies tables for details.

Support

The RCTs were published between 2008 and 2017. Four RCTs were funded by a charitable organization, and nine by departmental resources. Four studies received no financial support. The remaining trials did not specify the source of funding.

Setting

The 25 included studies were performed in Canada (n = 3), China (n = 3), Egypt (n = 2), Denmark (n = 6), Germany (n = 1), India (n = 1), Iran (n = 1), Korea (n = 1), USA (n = 6), and Turkey (n = 1).

Study population

The number of participants in these studies varied from 30 to 159. Studies included significantly more female adults (females n = 929 vs males n = 759). In most studies, adductor canal block was performed in participants undergoing total knee arthroplasty (TKA) (18/25 studies). Only seven trials used block in patients scheduled for arthroscopic knee surgery (Abdallah 2016; Akkaya 2008; Espelund 2013; Espelund 2014a; Hanson 2013; Messeha 2016; Rahimzadeh 2017). The population undergoing TKA was similar regarding diagnosis and ranged from 42 to 83 years of age. In comparison, the group of participants with arthroscopic knee surgery was significantly younger on average (18 to 65 years).

Intervention

Included studies compared adductor canal block versus femoralis nerve block or placebo. Fifteen trials explored the analgesic efficacy of adductor canal block (ACB) and femoralis nerve block (FNB). Another eight trials compared the analgesic effect of ACB using perineural local anaesthetic (LA) or placebo (saline). Two studies compared ACB versus periarticular infiltration. Finally, one study compared the effect of ACB versus psoas compartment block (Messeha 2016).

Thirteen RCTs used the LA ropivacaine in different concentrations. Most trial authors used 0.5% to 0.75% ropivacaine, with the exception of four studies, which used 0.2% (Sztain 2015; Zhang 2014), 0.25% (Li 2017), or 0.375% ropivacaine (Wiesmann 2016). The other studies infiltrated lidocaine 2% (Machi 2015), levobupivacaine 0.25% (Akkaya 2008), and 0.125% (Rahimzadeh 2017), 0.25% (Macrinici 2017; Memtsoudis 2015; Nader 2016), or 0.5% bupivacaine (Messeha 2016). Four authors added ephedrine to LA (Abdallah 2016; Hanson 2013; Messeha 2016; Nader 2016). Sawhney 2016 used additional ketorolac and morphine in the infiltration solution.

Most trial authors performed the nerve block preoperatively. In five studies, researchers conducted the block procedure postoperatively (Jaeger 2013; Macrinici 2017; Rahimzadeh 2017; Zhang 2014; Zhao 2017). Eleven studies selected continuous postoperative administration of LA: ropivacaine 0.2% to 0.25% (Andersen 2013; Elkassabany 2016; Jaeger 2013; Jaeger 2014; Jenstrup 2012; Shah 2014; Sztain 2015; Wiesmann 2016; Zhang 2014; Zhao 2017) or lidocaine 2% (Machi 2015). The remaining trials used a single injection procedure.

Most trial authors performed an ultrasound‐guided injection nerve block technique. Only one trial author used the combination of nerve stimulation (NS) and ultrasound (Zhang 2014).

For surgical procedures, most participants received general anaesthesia. Eight studies performed spinal anaesthesia (Elkassabany 2016; Hegazy 2015; Jaeger 2013; Jenstrup 2012; Machi 2015; Nader 2016; Sawhney 2016; Shah 2014). Two performed combined spinal‐epidural anaesthesia (Memtsoudis 2015; Zhang 2014). Three trials reported that they additionally provided local infiltration analgesia (LIA) to both groups (Andersen 2013; Nader 2016; Sztain 2015).

Most trials (13 out of 25 studies) reported that an additional multi‐modal analgesic regimen was started preoperatively and was continued postoperatively (Elkassabany 2016; Espelund 2013; Espelund 2014a; Hanson 2013; Hegazy 2015; Jaeger 2014; Koh 2017a; Li 2017; Machi 2015; Macrinici 2017; Nader 2016; Sawhney 2016; Sztain 2015). Opioids were given as rescue analgesics in all studies. Seven out of 25 of the included trials reported the use of a prophylactic drug against postoperative nausea and vomiting (PONV) (Andersen 2013; Elkassabany 2016; Hanson 2013; Memtsoudis 2015; Sawhney 2016; Shah 2014; Wiesmann 2016).

Excluded studies

We excluded 19 studies. The reasons for their exclusion are given in the Characteristics of excluded studies table. We excluded three trials because volunteers were investigated (Jaeger 2013b; Kwofie 2013, Monahan 2016), and we excluded three trials because they performed only retrospective analysis (Grant 2017; Gwam 2017; Seo 2017). One trial investigated hindfoot and ankle surgery instead of knee surgery (Joe 2016). Four RCTs compared ACB within a cross‐over design and were therefore excluded (Espelund 2014b Grevstad 2014 Grevstad 2015; Sorensen 2016).

We excluded five studies because they used two different local anaesthetics (Beausang 2016), or they used different volumes of local anaesthetics within study groups (Henshaw 2016 Kim 2014 Ortiz‐Gomez 2017 Sogbein 2017). Some participants were treated differently than described in the protocol (Jaeger 2012). Another trial provided additional treatment that was not part of the original protocol (Hanson 2014).

We excluded Shah 2015 because it compared single versus continuous ACB blockade.

Studies awaiting classification

We have presented 11 studies that are awaiting classification. Please refer to the Characteristics of studies awaiting classification table for details.

Ongoing studies

Within www.clinicaltrials.gov and http://www.who.int/ictrp/en/, 11 ongoing potentially relevant trials are registered and are recruiting patients. Please refer to the Characteristics of ongoing studies table for details.

Risk of bias in included studies

The risk of bias graph and summary can be seen in Figure 2 and Figure 3. The graph displays review authors’ judgements about each risk of bias item presented as percentages across all included RCTs. The risk of bias summary shows review authors’ judgements about each risk of bias item for each included study.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Allocation

Sequence generation

We judged six trials as having unclear risk of bias as they did not provide clear information on how the randomization sequence was generated (Akkaya 2008; Hanson 2013; Memtsoudis 2015; Messeha 2016; Zhang 2014; Zhao 2017). We judged all other studies as having low risk of bias due to adequate randomization.

Concealment of allocation

Sixteen of the included studies described allocation concealment; we judged them as having low risk of bias (Andersen 2013; Elkassabany 2016; Espelund 2013; Espelund 2014a; Hegazy 2015; Jaeger 2013; Jaeger 2014; Jenstrup 2012; Koh 2017a; Li 2017; Machi 2015; Macrinici 2017; Memtsoudis 2015; Nader 2016; Sawhney 2016; Sztain 2015). Nine trials did not report the method of allocation concealment, and we judged them as having unclear risk of bias (Abdallah 2016; Akkaya 2008; Hanson 2013; Messeha 2016; Rahimzadeh 2017; Shah 2014; Wiesmann 2016; Zhang 2014; Zhao 2017).

Blinding

Sixteen out of 25 trials were performed as double‐blind studies, with the participant and the provider of the intervention blinded to therapy (Abdallah 2016; Andersen 2013; Elkassabany 2016; Espelund 2013; Espelund 2014a; Hanson 2013; Hegazy 2015; Jaeger 2013; Jaeger 2014; Jenstrup 2012; Li 2017; Macrinici 2017; Memtsoudis 2015; Nader 2016; Sawhney 2016; Wiesmann 2016). We assessed five studies as having unclear risk of bias because blinding was not mentioned (Akkaya 2008; Messeha 2016; Rahimzadeh 2017; Zhang 2014; Zhao 2017). We rated six trials as having unclear risk of bias because they did not describe blinding of outcome assessment (Akkaya 2008; Andersen 2013; Messeha 2016; Rahimzadeh 2017; Zhang 2014; Zhao 2017). We rated two studies as having high risk of bias due to total non‐blinding (Machi 2015; Sztain 2015).

Incomplete outcome data

Three trials did not adequately report all evaluation data (Akkaya 2008; Zhang 2014; Zhao 2017). The remaining trials reported that all participants were included in the analysis; we assessed them as having low risk of bias.

Selective reporting

Two studies did not report all secondary outcomes; we therefore judged them to be at high risk of bias for selective reporting (Abdallah 2016; Messeha 2016). Due to insufficient data sources, we rated one study as having unclear risk of bias (Zhao 2017). We judged all other trials as having low risk of bias because all outcomes were measured and reported in full length, as judged from study reports (methods sections).

Other potential sources of bias

We found no further potential sources of bias in 23 trials. We rated two trials as having unclear risk of bias because data sources were insufficient (Zhang 2014; Zhao 2017).

Effects of interventions

See: Table 1; Table 2

for the main comparison.

Adductor canal block compared with sham treatment for postoperative pain following knee surgery
Patient or population: adult participants undergoing knee surgery (arthroscopic knee surgery or total knee replacement) Settings: postoperative care in hospital, Turkey (one trial), Denmark (four trials), USA (one trial) Intervention: adductor canal block Comparison: sham treatment (saline injection)
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect  (95% CI)	No. of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Assumed risk	Corresponding risk
	Sham treatment	Adductor canal block
Postoperative pain at rest (VAS 0 to 100 mm, NRS 0 to 10)  (2 hours)		Mean postoperative pain at rest (2 hours postoperatively) in the intervention group was  0.56 standard deviations lower (1.2 lower to 0.07 higher)		208 (4)	⊕⊕⊝⊝  low^a	Standard deviation of 0.5 represents a moderate effect
Postoperative pain at rest (VAS 0 to 100 mm, NRS 0 to 10) (24 hours)		Mean postoperative pain at rest (24 hours postoperatively) in the intervention group was  0.49 standard deviations lower (1.05 lower to 0.07 higher).		272 (6)	⊕⊕⊝⊝  low^a	Standard deviation of 0.5 represents a moderate effect
Postoperative pain during movement (VAS 0 to 100 mm, NRS 0 to 10) (2 hours)		Mean postoperative pain during movement (2 hours postoperatively) in the intervention group was 0.59 standard deviations lower (1.5 lower to 0.33 higher)		160 (3)	⊕⊝⊝⊝  very low^b	Standard deviation of 0.5 represents a moderate effect
Postoperative pain during movement (VAS 0 to 100 mm, NRS 0 to 10) (24 hours)		Mean postoperative pain during movement (24 hours postoperatively) in the intervention group was 0.03 standard deviations higher (0.26 lower to 0.32 higher)		184 (4)	⊕⊕⊝⊝  low^c	Standard deviation of 0.2 represents a small effect
Postoperative nausea   (24 hours)	Two out of 61 participants in the sham group suffered from nausea	Five out of 60 participants in the adductor canal group suffered from nausea	RR 1.91 (95% CI 0.48 to 7.58)	121 (3)	⊕⊕⊝⊝  low^c
Accidental falls during postoperative care   (24 hours)	No patient out of 24 participants in the sham group suffered from an accidental fall	No patient out of 24 participants in the adductor canal group suffered from an accidental fall		48 (1)	⊕⊕⊝⊝  low^d	Only 1 small trial assessed this outcome
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; NRS: numerical rating scale; RR: risk ratio; VAS: visual analogue scale.
GRADE Working Group grades of evidence.  High quality: further research is very unlikely to change our confidence in the estimate of effect.  Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.  Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.  Very low quality: we are very uncertain about the estimate.

Adductor canal block compared with femoral nerve block for postoperative pain following knee surgery
Patient or population: adult participants undergoing knee surgery (arthroscopic knee surgery or total knee replacement) Settings: postoperative care in hospital, USA (seven trials), China (two trials) Germany (one trial), India (one trial), Iran (one trial), Denmark (one trial) Intervention: adductor canal block Comparison: femoral nerve block
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect  (95% CI)	No. of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Assumed risk	Corresponding risk
	Femoral nerve block	Adductor canal block
Postoperative pain at rest (VAS 0 to 100 mm, NRS 0 to 10) (2 hours)		Mean postoperative pain at rest (2 hours postoperatively) in the intervention group was 0.74 standard deviations lower (‐1.76 lower to 0.28 higher)		298 (5)	⊕⊕⊝⊝  low^a	Standard deviation of 0.8 represents a large effect
Postoperative pain at rest (VAS 0 to 100 mm, NRS 0 to 10) (24 hours)		Mean postoperative pain at rest (24 hours postoperatively) in the intervention group was 0.04 standard deviations higher (‐0.09 lower to 0.18 higher)		868 (12)	⊕⊕⊕⊕  high	Standard deviation of 0.2 represents a small effect
Postoperative pain during movement (VAS 0 to 100 mm, NRS 0 to 10) (2 hours)		Mean postoperative pain during movement (2 hours postoperatively) in the intervention group was 0.47 standard deviations lower (‐1.86 lower to 0.93 higher)		88 (2)	⊕⊝⊝⊝  very low^b	Standard deviation of 0.5 represents a moderate effect
Postoperative pain during movement (VAS 0 to 100 mm, NRS 0 to 10) (24 hours)		Mean postoperative pain during movement (24 hours postoperatively) in the intervention group was 0.56 standard deviations higher (‐0.00 lower to 1.12 higher)		576 (9)	⊕⊝⊝⊝  very low^b	Standard deviation of 0.5 represents a moderate effect
Postoperative nausea (24 hours)	Five out of 70 participants in the femoral nerve block group suffered from postoperative nausea 24 hours postoperatively	Six out of 68 participants in the adductor canal block group suffered from postoperative nausea 24 hours postoperatively		138 (2)	⊕⊕⊝⊝  low^c
Accidental falls during postoperative care (24 hours)	Six out of 84 participants in the femoral nerve block group suffered from an accidental fall	No patient out of 88 participants in the adductor canal block group suffered from an accidental fall		172 (3)	⊕⊕⊝⊝  low^c
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; NRS: numerical rating scale; VAS: visual analogue scale.
GRADE Working Group grades of evidence.  High quality: further research is very unlikely to change our confidence in the estimate of effect.  Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.  Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.  Very low quality: we are very uncertain about the estimate.

Reference	Quadriceps MVIC	Adductor MVIC	Quadriceps strength scale MMT	Adductor strength scale MMT	Duration of SLR	Modified Bromage Scale
Abdallah 2016	# within 60 minutes
Elkassabany 2016			# within first 24 hours postoperative
Jaeger 2013	# at 24 hours postoperative	‐ at 24 hours postoperative
Koh 2017a	‐ at 1 week postoperative		# within first 24 hours postoperative		# within first 24 hours postoperative
Li 2017			# within first 12 hours postoperative	‐ within first 72 hours postoperative
Macrinici 2017		# within first 24 hours postoperative
Memtsoudis 2015			‐ within first 48 hours postoperative
Rahimzadeh 2017						‐ within first 24 hours postoperative
Wiesmann 2016			# at 24 hours postoperative
Zhang 2014		# within first 48 hours postoperative
Zhao 2017	‐ within 48 hours postoperative

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Mean difference in postoperative pain at rest (2 hours)	4	208	Std. Mean Difference (IV, Random, 95% CI)	‐0.56 [‐1.20, 0.07]
2 Mean difference in postoperative pain at rest (24 hours)	6	272	Std. Mean Difference (IV, Random, 95% CI)	‐0.49 [‐1.05, 0.07]
3 Mean difference in postoperative pain during movement (2 hours)	3	160	Std. Mean Difference (IV, Random, 95% CI)	‐0.59 [‐1.50, 0.33]
4 Mean difference in postoperative pain during movement (24 hours)	4	184	Std. Mean Difference (IV, Fixed, 95% CI)	0.03 [‐0.26, 0.32]
5 Rate of postoperative nausea 2 hours	2	79	Risk Ratio (IV, Fixed, 95% CI)	1.75 [0.56, 5.49]
6 Rate of postoperative nausea 24 hours	3	121	Risk Ratio (IV, Fixed, 95% CI)	1.91 [0.48, 7.58]
7 Rate of postoperative vomiting 24 hours	2	79	Risk Ratio (IV, Fixed, 95% CI)	1.18 [0.56, 2.47]
8 Rate of postoperative nausea and vomiting 24 hours	2	111	Risk Ratio (IV, Fixed, 95% CI)	0.54 [0.29, 1.02]
9 Rate of postoperative sedation 2 hours	2	91	Risk Ratio (IV, Random, 95% CI)	0.51 [0.17, 1.52]
10 Rate of postoperative sedation 24 hours	2	73	Risk Ratio (IV, Random, 95% CI)	0.78 [0.20, 3.07]
11 Cumulative mean morphine requirement (until 24 hours postop)	5	232	Mean Difference (IV, Random, 95% CI)	‐15.88 [‐30.87, ‐0.89]
12 Rate of failed block	2	89	Risk Ratio (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Mean difference in postoperative pain at rest (2 hours)	5	298	Std. Mean Difference (IV, Random, 95% CI)	‐0.74 [‐1.76, 0.28]
2 Mean difference in postoperative pain at rest (24 hours)	12	868	Std. Mean Difference (IV, Fixed, 95% CI)	0.04 [‐0.09, 0.18]
3 Mean difference in postoperative pain at rest (48 hours)	9	626	Mean Difference (IV, Random, 95% CI)	0.25 [‐0.71, 1.21]
4 Mean difference in postoperative pain during movement (2 hours)	2	88	Std. Mean Difference (IV, Random, 95% CI)	‐0.47 [‐1.86, 0.93]
5 Mean difference in postoperative pain during movement (24 hours)	9	576	Std. Mean Difference (IV, Random, 95% CI)	0.56 [‐0.00, 1.12]
6 Mean difference in postoperative pain during movement (48 hours)	8	528	Std. Mean Difference (IV, Fixed, 95% CI)	0.07 [‐0.10, 0.24]
7 Rate of postoperative nausea 24 hours	2	138	Risk Ratio (IV, Fixed, 95% CI)	1.22 [0.42, 3.54]
8 Rate of accidental falls during postoperative care 24 hours	3	172	Risk Ratio (IV, Fixed, 95% CI)	0.20 [0.04, 1.15]
9 Rate of accidental falls during postoperative care 48 hours	2	75	Risk Ratio (IV, Fixed, 95% CI)	0.27 [0.01, 6.11]
10 Rate of postoperative nausea and vomiting (PONV) 24 hours	2	151	Risk Ratio (IV, Fixed, 95% CI)	0.68 [0.44, 1.04]
11 Rate of accidental falls during postoperative care 24 hours	3	172	Risk Ratio (IV, Fixed, 95% CI)	0.20 [0.04, 1.15]
12 Rate of accidental falls during postoperative care 48 hours	2	75	Risk Ratio (IV, Fixed, 95% CI)	0.27 [0.01, 6.11]
13 Cumulative mean morphine requirement (until 2 hours postop)	5	305	Mean Difference (IV, Fixed, 95% CI)	1.00 [‐0.79, 2.79]
14 Cumulative mean morphine requirement (until 24 hours postop)	6	418	Mean Difference (IV, Fixed, 95% CI)	‐1.03 [‐3.48, 1.41]
15 Rate of failed block	3	281	Risk Ratio (IV, Fixed, 95% CI)	1.46 [0.16, 12.99]
16 Rate of postoperative block‐related neurological impairment	4	385	Risk Ratio (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]

Methods	RCT Parallel design
Participants	100 patients (ACB: 31.6 years (28.9 to 34.3), 38 males; FNB: 33.3 years (30.7 to 35.9), 26 males), elective unilateral anterior cruciate ligament reconstruction, general anaesthesia
Interventions	Single shot Preoperative Study period: May 2013 to March 2015 ACB: 20 mL R 0.5% with epinephrine (52) FNB: 20 mL R 0.5% with epinephrine (48)
Outcomes	Pain at rest on VAS 2 hours, 24 hours Rate of postop nausea and vomiting Rate of accidental falls during postop care Cumulative mean morphine requirement 24 hours Rate of postop neurological impairment Rate of failed block Maximal voluntary isometric contraction (MVIC) postop within 60 minutes
Notes	Country: USA Conflict of interest: no conflicts of interest among primary researchers Funding: Merit Award Program, Department of Anesthesia, University of Toronto, Canada, University Health Network Innovation Fund Plan
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: “an investigator with no further involvement in the study generated a list of random numbers in varying block sizes by using an online computer randomization service (www.Randomization.com). The unique randomization code was used to randomize consenting study participants at a 1:1 ratio with no restrictions for either of the 2 study groups: ACB group or FNB group“
Allocation concealment (selection bias)	Unclear risk	Quote: “the results of the allocation were concealed in sealed opaque envelopes and kept with the research coordinator. On the day of surgery, the research coordinator handed an envelope to the attending anesthesiologist or a directly supervised regional anesthesia fellow in the block procedure room immediately before administering the study block to the participant”
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Quote: “the staff anesthesiologist or fellow performing the block had no further role in the study., […] patient and assessor‐blinded, […] blinded PACU nursing staff“
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Quote: ”[…] all outcome data were collected by a blinded research coordinator”
Incomplete outcome data (attrition bias)   All outcomes	Low risk	All patients were evaluated; no data were missing
Selective reporting (reporting bias)	High risk	Quote: "only minimal secondary outcome data were missing"
Other bias	Low risk	No apparent bias

Methods	RCT  Parallel design
Participants	62 patients (ACB: 63 ± 8 years, 9 males; FNB: 65 ± 8 years, 12 males), scheduled for primary unilateral TKA, ACB
Interventions	Continuous  Preoperative Study period: unclear ACB: 20 mL R 0.5% bolus and 8 mL/h 0.2% R after surgery (31) FNB: 20 mL R 0.5% bolus and 8 mL/h 0.2% R after surgery (31)
Outcomes	Pain at rest 24 hours, 48 hours (no scale mentioned ‐ pain is assessed before and after physical therapy sessions) Cumulative mean opioid requirement (24 hours, 48 hours; hydromorphone or fentanyl) Degree of quadriceps muscle strength (24 hours, 48 hours; MMT = manual muscle testing)
Notes	Country: USA Conflict of interest: no conflicts of interest among primary authors Funding: Education and Development funds, Department of Anesthesiology and Critical Care, University of Pennsylvania
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: “a computer‐generated randomization table was used for patient allocation to 1 of the 2 study groups: the FNB group or the ACB group. Randomization was performed in blocks of 10 patients”
Allocation concealment (selection bias)	Low risk	Quote: “patients’ assignments were written in a sealed envelope that was opened only after patient consent for the study”
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Quote: “the dressing over the catheter was made wide enough to conceal the difference of the catheter location between FNB and ACB groups. The nurses on the floor, the research coordinator, and the physical therapist were all blinded to the nature of patient assignment. All PT measurements were performed by the same physical therapist who was blinded to the nature of group assignment“
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Quote: “all PT measurements were performed by the same physical therapist who was blinded to the nature of group assignment. The questionnaires were administered by the same research assistant who was blinded to the group assignments“
Incomplete outcome data (attrition bias)   All outcomes	Low risk	All patients were evaluated; no data were missing
Selective reporting (reporting bias)	Low risk	Obviously, all primary and secondary outcome data were reported
Other bias	Low risk	No apparent bias

Methods	RCT  Parallel and cross‐over design
Participants	59 patients (age: 64.41 ± 7.36 years, 26 males), scheduled for bilateral TKA, SPA, and epidural catheter
Interventions	Single  Preoperative Study period: April 2012 to September 2013 ACB + FNB: 15 mL B 0.25%, 30 mL B 0.25% (30) left leg saphenous, right leg femoral FNB + ACB: 30 mL B 0.25%, 15 mL B 0.25% (29) right leg saphenous, left leg femoral
Outcomes	Pain at rest on VAS (24 hours, 48 hours) Pain during movement on VAS (24 hours, 48 hours) Cumulative mean morphine requirement (total) Degree of quadriceps muscle strength (24 hours, 48 hours; Lafayette manual muscle test system, standardized 0 to 5 motor‐strength scale)
Notes	Country: USA Conflict of interest: no conflicts of interest Funding: Hospital for Special Surgery, Department of Anesthesiology, New York, Anna‐Maria and Stephen Kellen Physician‐Scientist Career Development Award, New York
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: “the two extremities were randomised to receive either US guided subsartorial”
Allocation concealment (selection bias)	Low risk	Quote: “[…] using blinded envelopes prepared by a independent research assistant and only visible to the attending anaesthesiologist assigned"
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Quote: “patients, surgeons, physical therapists and research assistants performing the follow‐up were blinded to the randomisation. Subsequently, blocks were performed as randomised using a sterile technique"
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Quote: “[...] research assistants performing the follow‐up were blinded to the randomisation. All study data were collected and managed by using REDCap electronic data capture tools through the Clinical and Translational Science Center at Weill Cornell Medical College"
Incomplete outcome data (attrition bias)   All outcomes	Low risk	All patients were evaluated; no data were missing
Selective reporting (reporting bias)	Low risk	Obviously, all primary and secondary outcome data were reported
Other bias	Low risk	No apparent bias

Study	Reason for exclusion
Beausang 2016	Different use of LA in ACB and control groups
Espelund 2014b	Cross‐over design
Grant 2017	Retrospective analysis
Grevstad 2014	Cross‐over design
Grevstad 2015	Cross‐over design
Gwam 2017	Retrospective analysis
Hanson 2014	Some participants were treated differently than was described in the protocol with additional infiltrations
Henshaw 2016	Different amount of LA
Jaeger 2012	Some participants were treated differently than was described in the protocol
Jaeger 2013b	Volunteers
Joe 2016	Investigated hindfoot and ankle surgery instead of knee surgery
Kim 2014	Different amount of LA
Kwofie 2013	Volunteers were investigated
Monahan 2016	Volunteers were investigated
Ortiz‐Gomez 2017	Different amount of LA
Seo 2017	Retrospective analysis
Shah 2015	Comparison: single vs continuous ACB
Sogbein 2017	Comparison: ACB vs periarticular infiltration; different amount of LA
Sorensen 2016	Cross‐over design

Trial name or title	Efficacy of different lower extremity nerve block combined with general anaesthesia in knee arthroscopic surgery
Methods	RCT Parallel design
Participants	Unclear
Interventions	Total intravenous anaesthesia Adductor canal block combined with general anaesthesia Femoral nerve block combined with general anaesthesia Adductor canal and lateral femoral cutaneous nerve block combined with general anaesthesia Femoral nerve and lateral femoral cutaneous nerve block combined with general anaesthesia
Outcomes	Amount of anaesthesia in the operation Time for spontaneous breathing recovery Awakening time Extubation time Visual analogue scale at rest Visual analogue scale for active functional exercise Visual analogue scale for continuous passive movement Quadriceps strength Analgesic dosage within 24 hours after operation
Starting date	September 2018
Contact information	liu711029@hotmail.com
Notes	Declaration of interests: unclear Funding: unclear

Trial name or title	Analgesic efficacy of saphenous nerve blockade for outpatient knee anterior cruciate ligament surgery
Methods	RCT Parallel design
Participants	58 participants
Interventions	Saphenous nerve blockade 15 mL of levobupivacaine 0.5% Femoral nerve blockade 15 mL of levobupivacaine 0.5%
Outcomes	Readiness to discharge from day care centre in hours according to post‐anaesthetic discharge scoring system (PADSS) (time frame: 1 day) Motor block (time frame: 2 days) Pain (VAS) (time frame: 2 days) Sensory blockade extent (time frame: 1 day) Time to rescue analgesic and postoperative opioid consumption (time frame: 1 day) Overall benefit of analgesia score (OBAS) (time frame: 6 weeks) SF‐12 score (time frame: 12 weeks) KOOS‐score (time frame: 12 weeks) IKDC‐score (time frame: 12 weeks)
Starting date	March 2014
Contact information	w.tenhoope@amc.uva.nl
Notes	Declaration of interests: unclear Funding: unclear

Trial name or title	Does single injection adductor canal block improve postoperative analgesia in patients receiving periarticular local anesthesia injections for total knee arthroplasty?
Methods	RCT Parallel design
Participants	90 participants
Interventions	Adductor canal block + local infiltration Local infiltration Adductor canal block
Outcomes	Pain score (time frame: within 24 hours postoperative) Opioid use (time frame: within 24 hours postoperative) PACU opioid use (time frame: postoperative (while in PACU), expected average of 60 minutes) Daily opioid use (time frame: duration of hospital stay, expected average of 3 days) Average NRS pain score (time frame: within 24 hours postoperative) Length of stay (time frame: duration of hospital stay, expected average of 3 days)
Starting date	October 2014
Contact information	canalesc@uci.edu
Notes	Declaration of interests: unclear Funding: unclear

Trial name or title	Assessment of sensory and motor blockade of the adductor canal blockade performed for surgery of arthroscopic anterior cruciate ligament repair
Methods	RCT Paralell design
Participants	40 participants
Interventions	Ultrasound guided adductor canal blockade (20 mL of ropivacaine 0.75%) Ultrasound guided femoral nerve blockade (20 mL of ropivacaine 0.75%)
Outcomes	Cold sensitivity assessment (cold, very cold, no sensation) (time frame: from 30 to 60 minutes after nerve blockade), sensitivity description (cold, very cold, no sensation) Motor blockade assessment (dynamometer) (time frame: from 30 to 60 minutes after nerve blockade), motor blockade evaluation with dynamometer Postoperative pain assessment (visual analogue scale) (time frame: at 2, 4, 6 postoperative hours) Evaluation with visual analgesic scale
Starting date	October 2014
Contact information	jplecoq@chu.ulg.ac.be
Notes	Declaration of interests: unclear Funding: unclear

Trial name or title	Patient outcomes with periarticular liposomal bupivacaine injection vs adductor canal block after primary total knee arthroplasty
Methods	RCT Parallel design
Participants	250 participants
Interventions	Liposomal bupivacaine Adductor canal and tibial nerve block
Outcomes	Length of stay (LOS, in days) (time frame: participants will be followed for the duration of their hospital stay ‐ an expected average of 1.5 days) Time to ambulation (in hours) (time frame: 2 to 6 hours) Pain as assessed by visual analogue scale (VAS) on postoperative day 0 (time frame: 6 hours) Pain as assessed by visual analogue scale (VAS) on postoperative day 1 (time frame: 24 hours) Pain as assessed by visual analogue scale (VAS) on postoperative day 2 (time frame: 48 hours) Pain as assessed by visual analogue scale (VAS) on postoperative day 3 (time frame: 72 hours) Pain as assessed by visual analogue scale (VAS) on postoperative day 4 (time frame: 96 hours) Pain as assessed by visual analogue scale (VAS) on postoperative day 5 (time frame: 120 hours) Opioid consumption in oral morphine equivalents (OMEs, in milligrams) (time frame: participants will be followed for the duration of their hospital stay ‐ an expected average of 1.5 days) Postoperative complications and adverse events (time frame: 2 weeks)
Starting date	February /2016
Contact information	atorres@txortho.com
Notes	Declaration of interests: unclear Funding: unclear

Trial name or title	Femoral nerve block versus adductor canal nerve block for peri‐operative analgesia following anterior cruciate ligament reconstruction: evaluation of post‐operative pain and strength
Methods	RCT Parallel design
Participants	80 participants
Interventions	Adductor canal block Femoral nerve block
Outcomes	Visual analogue scale (time frame: postop day 0 to 5, every 4 hours) Opioid requirement (time frame: postop day 0 to 5) Thigh circumference (time frame: 2 weeks postoperative vs 6 months postoperative) Straight leg raise (time frame: 0 to 7 days postoperative)
Starting date	May 2016
Contact information	jlynch6@hfhs.org
Notes	Conflict of interests: unclear Funding: unclear

Trial name or title	Adductor canal block versus femoral nerve block with repeated bolus doses: postoperative analgesia and functional outcomes after total knee arthroplasty
Methods	RCT Parallel design
Participants	42 participants
Interventions	ACB: bupivacaine 0.25% will be applied to the catheter at 6 hours, when the first dose of catheter is inserted, and the catheter will be inserted with the peripheral nerve stimulator lateral to the femoral artery. Bupivacaine 0.25% will be applied to the catheter at 6 hours, when the first dose of catheter is inserted FNB: bupivacaine 0.25% will be applied to the catheter at 6 hours, when the first dose of catheter is inserted
Outcomes	Postoperative muscle strength ‐ quadriceps muscle strength scale (time frame: postoperative 48 hours) Postoperative analgesia ‐ visual analogue scale (time frame: 72 hours)
Starting date	April 2017
Contact information	kurtbeyogluseda@gmail.com
Notes	Declaration of interests: unclear Funding: unclear

Trial name or title	Epidural analgesia vs adductor canal block in bilateral TKA
Methods	RCT Parallel design
Participants	70 participants
Interventions	Bilateral single‐shot bilateral adductor canal block Continuous epidural block
Outcomes	Pain scores at rest (time frame: 48 hours postoperatively) Morphine consumption (time frame: 48 hours postoperatively) Pain scores on movement (time frame: 48 hours postoperatively) Side effects of interventions (time frame: 48 hours postoperatively)
Starting date	August 2017
Contact information	stangwiwat@yahoo.com
Notes	Declaration of interests: unclear Funding: unclear

Trial name or title	A prospective comparison of pain and quality of recovery in patients undergoing anterior cruciate ligament reconstruction with adductor canal or femoral perineural infusions
Methods	RCT Parallel design
Participants	50 participants
Interventions	Adductor canal catheter Femoral nerve catheter
Outcomes	Pain score (time frame: postoperative day 2) Quality of recovery (time frame: POD 2) Opioid use (time frame: POD 2) CPM compliance (time frame: POD 2) Quality of recovery (time frame: POD 1) Bolus dose usage (time frame: POD 2 ) Return to play (time frame: 3 months) Quadriceps circumference, percent of baseline
Starting date	June 2018
Contact information	hornj@stanford.edu
Notes	Conflict of interests: unclear Funding: unclear

Trial name or title	Early mobilization and postoperative analgesia after total knee arthroplasty, a prospective comparative study: adductor canal block vs femoral nerve block vs apex femoral triangle block
Methods	RCT Parallel design
Participants	126 participants
Interventions	Apex femoral triangle block Adductor canal block Femoral nerve block
Outcomes	Maximum voluntary isometric contraction (MVIC) (time frame: 6 hours postoperatively) Timed up and go (TUG) (time frame: 6, 24, and 48 hours postoperatively) 30' CST (30 seconds chair stand test) (time frame: 6, 24, and 48 hours postoperatively) Range of motion (ROM) (time frame: 6, 24, and 48 hours postoperatively) Daniels' test (time frame: 6, 24, and 48 hours postoperatively) 10‐PMS (10 point mobility scale) (time frame: 6, 24, and 48 hours postoperatively) Pain measurement through the visual analogue scale (VAS) (time frame: 6, 24, and 48 hours postoperatively) Quantity of opioids administered (time frame: 6, 24, and 48 hours postoperatively) APS‐POQ‐R (time frame: at 24 hours postoperatively) Patient satisfaction (time frame: 24 and 48 hours postoperatively) Length of hospital stay (time frame: at patient discharge, an average of 6 days postoperatively) Maximum voluntary isometric contraction (MVIC) (time frame: 24 and 48 hours postoperatively)
Starting date	April 2018
Contact information	csalvadores@vhebron.net
Notes	Declaration of interests: unclear Funding: unclear

Trial name or title	Comparison of two techniques of locoregional analgesia in total knee prosthesis surgery: block to the adductor channel versus peri‐articular local infiltrations
Methods	RCT Parallel design
Participants	120 participants
Interventions	Adductor canal block Periarticular infiltration
Outcomes	Cumulative consumption of morphine Pain at rest (time frame: resting pain will be measured every 4 hours in the first 24 hours, then every 8 hours between 24 and 72 hours) Pain in movement (time frame: at day 3) Quality of the analgesia offered by the ACB (time frame: within 48 hours after injection to the adductor channel) Quality of the analgesia offered by the perarticular infiltration (time frame: within 48 hours after the end of surgery) Quality of the analgesia (time frame: within 48 hours after the surgical incision time) Functional capabilities and rehabilitation (time frame: at day 3) Duration of hospitalization (time frame: at day 3) Patient satisfaction (time frame: at day 3) Secondary complications due to adductor channel block (time frame: day 0: at the time of the ACB in the pre‐induction room) Complications secondary to analgesic medications and surgery (time frame: at day 3)
Starting date	February 2018
Contact information	caroline.macabeo@chu‐lyon.fr
Notes	Declaration of interests: unclear Funding: unclear

PERMALINK

Adductor canal blocks for postoperative pain treatment in adults undergoing knee surgery

Alexander Schnabel

Sylvia U Reichl

Stephanie Weibel

Peter K Zahn

Peter Kranke

Esther Pogatzki‐Zahn

Christine H Meyer‐Frießem

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Summary of findings

Background

Description of the condition

Description of the intervention

How the intervention might work

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

Data extraction and management

Assessment of risk of bias in included studies

Measures of treatment effect

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Sensitivity analysis

'Summary of findings' table and GRADE

Results

Description of studies

Results of the search

1.

Included studies

Support

Setting

Study population

Intervention

Excluded studies

Studies awaiting classification

Ongoing studies

Risk of bias in included studies

2.

3.

Allocation

Sequence generation

Concealment of allocation

Blinding

Incomplete outcome data

Selective reporting

Other potential sources of bias

Effects of interventions

for the main comparison.

2.

Comparison 1: adductor canal block (ACB) versus sham treatment

Primary outcomes

Mean differences in postoperative pain at rest/during movement (2 hours (within the postoperative care unit), 24 hours, 48 hours)

1.1. Analysis.

1.2. Analysis.

1.3. Analysis.

1.4. Analysis.