Esmolol as an Adjunct in Multimodal Anesthesia: A Systematic Review and Meta-Analysis of Its Opioid-Sparing and Analgesic Effects

Abstract

BACKGROUND:

Esmolol, an ultra-short-acting β1-selective adrenergic antagonist, has been investigated for its potential opioid-sparing effects in multimodal anesthesia. Previous systematic reviews included trials with different control groups causing severe limitations to the generalization of the findings. This systematic review and meta-analysis exclusively synthesized placebo-controlled randomized trials to evaluate the impact of intraoperative esmolol infusion on opioid consumption and postoperative pain scores within the first 24 hours after surgery.

METHODS:

A systematic search was conducted in Medline, Embase, Cochrane Library, and Google Scholar to identify randomized placebo-controlled trials assessing the effects of continuous intraoperative esmolol infusion on opioid consumption and pain scores. The outcomes of interest were total intraoperative and postoperative opioid consumption, converted to intravenous morphine milligram equivalents (IV MME), and pain intensity, and assessed using either the visual analog scale (VAS) or the Numeric Rating Scale (NRS), both of which were standardized using validated methods to a common 0 to 10 scale. Meta-analyses were performed using a random-effects model, and heterogeneity was assessed using Cochran’s Q test and I² statistics. Meta-regression and subgroup analyses explored the effects of esmolol infusion rate, type of surgery, intraoperative anesthetic and hemodynamic management, and patient age as potential moderators.

RESULTS:

Nineteen randomized trials (1028 patients) were included, involving esmolol regimens with a loading dose ranging from 0.5 to 1.0 mg/kg and a maintenance infusion rate of 0.3 to 6 mg/kg/h. Surgical procedures ranged from minimally invasive to open intracavitary surgeries. Esmolol infusion significantly reduced in 32% the intraoperative opioid consumption (mean differences [MD], –12.89 IV MME; 95% Confidence Interval, 95% confidence interval [CI], –24.74 to –1.05; P < .001; I² = 93.6%) and 38.6% the postoperative opioid consumption (MD, –3.03 IV MME; 95% CI, –4.29 to –1.76; P < .001; I² = 89.9%). For pain scores, 3 analyses were performed at the following intervals: within 30 minutes (MD, −1.47; 95% CI, −2.02 to −0.93; P < .001), between 2 and 4 hours (MD, −0.67; 95% CI, −1.29 to −0.06; P = .032), and at 24 hours (MD, −0.48; 95% CI, −0.92 to −0.03; P =.038). Based on the weighted mean values for the placebo group in each pooled analysis, we observed reductions in pain scores of 27.3%, 15.8%, and 23.5%, respectively. In addition, esmolol significantly reduced intraoperative heart rate in most cases and lowered MAP at multiple time points in several studies. Despite this, no significant increase in hypotension or bradycardia was reported, and only 1 study noted higher ephedrine and atropine use.

CONCLUSIONS:

Esmolol infusion significantly reduces opioid consumption and postoperative pain, with a magnitude of effect that may have clinical significance. The observed effects remained consistent in subgroups where confounding variables, expected to bias the results toward the null, were present. However, the uncertainty in later pain outcomes and the persistent heterogeneity warrant further research into esmolol’s applicability across different surgical contexts and populations.

KEY POINTS.

• Question: Is multimodal anesthesia with esmolol as an adjuvant effective in reducing opioid consumption and postoperative pain?

• Findings: Meta-analyses of randomized placebo-controlled trials reveal a dose-dependent opioid-sparing effect of intraoperative esmolol, with significantly reduced postoperative pain scores, particularly in early recovery.

• Meaning: Evidence shows that esmolol can be used as an adjuvant in multimodal anesthesia, especially in medium-to-major and fast recovery surgeries.

See Article, page 879

Perioperative pain management and multimodal analgesia are essential in modern anesthesia, with guidelines recommending individualized approaches to enhance pain control and reduce opioid-related adverse effects.^1–3 These recommendations advocate incorporating various classes of drugs during anesthesia, including: acetaminophen, nonsteroidal anti-inflammatory drugs (NSAIDs), local anesthetics, alpha-2-adrenergic agonists, and N-methyl-d-aspartic acid (NMDA) receptor antagonists. Some of these medications were created with other objectives, like clonidine, originally synthesized in 1960 to treat hypertension, and today successfully integrated into anesthetic practice.^4–6 As previously observed with clonidine, novel strategies focusing on pain modulation via adrenergic system receptors have gained significant attention in recent decades. Experimental studies in animal models have identified pathways of pain modulation involving beta-adrenoreceptors.^7–12

In humans, research into the analgesic potential of beta-blockers has grown steadily,^13–16 with 3 previous meta-analysis highlighting the role of esmolol in perioperative pain management.^17–19 Interest in investigating this beta-1-adrenergic antagonist in anesthesia stems from its favorable safety profile, rapid metabolism via plasma esterases, and its ability to provide enhanced hemodynamic control.^20–23 The most recent meta-analysis, by Gelineau et al,¹⁷ highlighted an intriguing opioid-sparing effect in patients receiving continuous esmolol infusion during surgery. However, the study has limitations, as it combines findings from placebo-controlled trials, opioid-controlled trials, and studies utilizing peripheral nerve blocks as an adjunct technique, making the synthesis less robust. In addition, all of these studies were identified through a restricted search strategy. Since then, new randomized placebo-controlled trials have been published, providing additional evidence of esmolol’s potential in perioperative pain management.^24–27 Despite this, limited robustness of the available evidence has led to the most recent and influential consensus to exclude esmolol from recommendations for multimodal anesthesia.^1,3,28 One of the key remaining uncertainties is whether the observed reduction in opioid use with esmolol reflects a true analgesic effect or merely a blunted autonomic response to surgical stimuli.^17,29

To address the need for strategies to reduce perioperative opioid consumption without compromising analgesia quality,^30–32 and the lack of an updated meta-analysis focused solely on placebo-controlled trials, we conducted a systematic review to evaluate the role of esmolol in multimodal anesthesia, analyzing opioid consumption and pain levels.

METHODS

This systematic review adhere to the guidelines described in Cochrane Handbook for Systematic Reviews of Interventions³³ and followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines.³⁴ Protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) as CRD42025631415. No new human or animal data were collected.

Eligibility Criteria

Eligible studies were randomized, placebo-controlled trials with esmolol administered as a loading dose followed by continuous intraoperative infusion. The population of interest included adult patients (aged ≥18 years) undergoing any type of surgery. The intervention of interest was defined as an esmolol bolus dose followed by continuous infusion maintained until the end of the surgical procedure, ensuring patients were under beta-blocker effect from the onset of the painful stimulus. Only studies comparing esmolol to a placebo (normal saline or ringer lactate) control were included. Studies were included if they reported at least one of the following outcomes: intraoperative opioid dose, and/or postoperative opioid consumption at any time-point within 24 hours postoperatively, and/or pain intensity at rest within 24 hours after surgery, assessed using the VAS and/or NRS. Studies that exclusively reported hemodynamic parameters were excluded. Studies with peripheral nerve blocks, which could confound results, or without relevant outcomes were excluded. Peripheral nerve blocks provide substantial analgesia, potentially masking or reducing the ability to detect differences in analgesic efficacy of the studied drug. No restrictions on language or publication date were applied, and gray literature was considered if methodological details were sufficient. Corresponding authors were contacted in cases where missing outcome data or reporting bias was suspected.

Information Sources

Two reviewers (E.D.S.N. and S.N.A.) conducted a systematic search of Medline, Embase, Cochrane, and Google Scholar using terms related to esmolol. We searched the databases from inception to November 30, 2024. Detailed search strategy, outlined in Supplemental Digital Content 1, https://links.lww.com/AA/F518 focused on surgical populations, intraoperative esmolol infusion, and placebo controls. Records were independently screened by both reviewers according to predefined eligibility criteria. Discrepancies were resolved through discussion, with a third reviewer (M.C.S.) providing input when needed to achieve consensus.

Data Collection and Synthesis Methods

Information on the baseline characteristics, hemodynamic management, study methodology, esmolol dose, opioid regimen, adverse events, and outcomes of interest for the review were extracted by a single investigator (E.D.S.N.) and verified by 2 authors (S.N.A. and M.C.S.). All opioid usage data were standardized to intravenous morphine milligram equivalents (IV MME) using the conversion criteria (Supplemental Digital Content 2, Supplemental Table S1, https://links.lww.com/AA/F519),³⁵ to ensure consistency across studies. To synthesize pain scores expressed in VAS or NRS, we followed specific guidelines to standardize the results into a common metric ranging from 0 to 10.³⁶ If the values were reported in medians and interquartile range, estimated mean and standard deviation were obtained adequately through Luo et al’s³⁷ and Wan et al’s³⁸ methods. Data points of the outcomes of interest were extracted from published figures using WebPlotDigitizer (Version 4.6), ensuring accurate digitization of values presented in graphical format.³⁹

Quality Assessment

Two reviewers (S.N.A. and E.D.S.N.) independently assessed methodological quality in the included randomized trials using Cochrane Risk of Bias 2 tool,⁴⁰ a standardized method for evaluating bias in such studies. Discrepancies between reviewers were resolved through discussion, involving a third reviewer when necessary. Two reviewers(S.N.A. and E.D.S.N.) evaluated overall quality of evidence using GRADEpro GDT (Guideline Development Tool), developed by McMaster University and Evidence Prime, following the principles of the Grading of Recommendations Assessment, Development, and Evaluation framework.⁴¹

Analyses

Selection of studies for each synthesis followed a comprehensive data collection process, which involved systematically extracting and tabulating information from all eligible studies. This approach facilitated the integration of consistent and relevant data for synthesis, allowing for a robust and transparent summary of results.

Analyzed outcomes included were as follows: (1) total IV MME related at any time-point within 24 hours postoperatively; (2) total intraoperative opioid dose converted to IV MME; and (3) VAS or NRS for pain at rest (0–10) at 30 minutes, between 2 and 4 hours, and at 24 hours postoperatively. Intraoperative opioid dosage and pain assessment within 30 minutes post-surgery provide an opportunity to evaluate the potential direct analgesic effects of esmolol.^14–16 Despite esmolol’s short half-life, findings in animal models suggest potential modulation of pain pathways for up to 24 hours, and thus we evaluated this timeframe in our study.^8–12

Effect Measures and Quantitative Synthesis

Aggregating pain-related outcomes required methodological adaptation due to substantial variability in measurement tools, assessment timing, and reporting practices across trials, a recognized challenge in pain research by the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT).^42–45 All opioid doses were converted to intravenous morphine milligram equivalents (IV MME) using standard equianalgesic potency ratios, and mean differences were calculated for the statistical synthesis. Percentage reductions were calculated based on the weighted mean opioid consumption in the placebo group. In the pooled analyses of pain scores, consistency in outcome measurement methods and standardization of scales allowed for synthesis using mean differences (MD).

Investigation of Heterogeneity

Based on the primary analysis of each outcome, we subsequently explored sources of heterogeneity across studies. Meta-regression analysis was conducted in meta-analyses with 10 or more studies.^33,43 For moderators with essential discrepancies in the number of studies per category (<10 studies per category), we opted for subgroup analysis. We evaluate the effects of surgery type (minor versus medium/major), patient age (continuous), intraoperative opioid type (remifentanil versus nonremifentanil) on the amount of IV MME administered intra and postoperatively, and pain scores.⁴⁴ Procedures that involve penetrating fascial layers, opening the peritoneum, or removing internal organs were classified as medium/major surgeries. In contrast, endoscopic and hysteroscopic procedures, septoplasties, tooth extractions, and skin tumor removals were considered minor surgeries.⁴⁶ Measurement point-time (continuous) of the outcome and intraoperative hemodynamic management regimen (opioid-only versus adjuncts) were specific moderators for postoperative morphine consumption and intraoperative opioid dose, respectively. Patients managed with opioid-only regimens rely exclusively on opioids for intraoperative analgesia, using blood pressure and heart rate as primary targets for dose titration. This practice may make the opioid-sparing effects of esmolol more evident, as patients managed solely with opioids are more exposed to these drugs. The time point of measurement is included as a moderator because the short half-life of esmolol may result in its opioid-sparing effects diminishing at later time points. Early time points may better reflect its residual effects.

We conducted post hoc subgroup analyses to explore potential modifiers of esmolol’s effect. For infusion rate, studies were stratified by dose (≥3 vs <3 mg/kg/h), and a meta-regression was performed with dose as a continuous moderator to assess a potential dose–response relationship. Additional subgroup analysis was conducted based on reported use of adjunct analgesics (eg, NSAIDs, ketamine, dexamethasone, dipyrone, clonidine, dexmedetomidine), with studies classified as “adjuncts reported or administered” versus “not reported or absent.” In addition, post hoc subgroup analysis was performed to assess the potential impact of missing intraoperative opioid dosing data on the primary outcome of postoperative pain scores within 30 minutes after surgery. Sensitivity analyses were performed using the R package find.outliers to identify and assess the influence of potentially influential studies, comparing results with and without these observations to evaluate robustness.

Meta-Analytical Model

This analysis utilized a mixed-effects model with the restricted maximum-likelihood (REML) estimator for (τ²).⁴⁷ Cochran Q test and I² statistics were used to assess heterogeneity; I² > 50% were considered significant for inconsistency.⁴⁸ Heterogeneity was rigorously assessed using tests for residual heterogeneity(residual I²) and between-study variability(R²).⁴⁹ Results were considered statistically significant when P < .05.⁵⁰ Publication bias was assessed using Egger’s test and funnel-plots.⁵¹ Statistical analyses were performed using R Statistical Software (v4.4.2; R Core Team 2021).⁵²

RESULTS

We found 1327 results after executing the search strategy. After deletion of duplicate studies, 743 records were screened by title and abstract, of which 36 studies were retrieved for eligibility assessment. One of the studies could not be retrieved, even after repeated attempts.⁵³ Subsequently, 17 studies were excluded for not reporting outcomes of interest, lack of information about randomization, administering esmolol only before extubating, or performing peripheral nerve block as an adjunct technique. Nineteen randomized placebo-controlled studies evaluating outcomes of interest were included and analyzed (Figure 1).³⁴ These 19 studies included 1028 patients, with 515 allocated to the esmolol group.^{24–26,54–69} Studies involved patients undergoing various surgical procedures and different baseline characteristics, summarized in the Table. Findings of interest can be found in Supplemental Digital Content 3, Supplemental Table S2, https://links.lww.com/AA/F520. Of the 19 studies, 7 were identified as having a low risk of bias, 11 as having some concerns, and one as having a high risk of bias (Supplemental Digital Content 4, Supplemental Figure S1, https://links.lww.com/AA/F521). Forest plots for all subgroup analyses are provided in Supplemental Digital Content 5, Figures S2–S13, https://links.lww.com/AA/F522. Furthermore, funnel plots did not reveal significant publication bias (Supplemental Digital Content 6, Supplemental Figure S14, https://links.lww.com/AA/F523). A detailed evidence profile, analyzed using the GRADE tool, is presented in Supplemental Digital Content 7, Supplemental Table S3, https://links.lww.com/AA/F524.

Figure 1.

Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) flow diagram of study screening and selection. The search strategy in Medline, Embase, Google Scholar, and Cochrane Library yielded 1327 studies, of which 36 were fully reviewed for inclusion and exclusion criteria. Nineteen studies were included in the meta-analysis.

Table.

Baseline Characteristics of Included Studies

Study	No. of patients E/C	Age mean in years E/C [range]	Hemodynamic management agents	Esmolol regimen	Control	Not analyzed	Type of surgery
Bhawna et al55	25/25	46.6 ±7.80/ 44.4 ±8.55	Fentanyl	0.5 mg/kg before induction over 5min then 3 mg/kg/h	Normal saline	-	Lower abdominal surgeries
Celebi et al56	30/30	27.4 ±7.90/ 29.1 ±9.50	Remifentanil	0.5 mg/kg then 3 mg/kg/h	Normal saline	-	Septorhinoplasty
Chia et al57	49/48	48.5 [30–79]/ 49.8 [27–75]	Isoflurane and fentanyl	0.5 mg/kg before induction then 3 mg/kg/h	Normal saline	-	Abdominal total hysterectomy
De Oliveira et al58	29/29	38.0 ±7.00/ 41.0 ±9.00	Fentanyl	0.5 mg/kg then 0.3–0.9 mg/kg/h	Normal Saline	-	Gynecologic hysteroscopic surgery
Dhir et al54	30/30	45.3 ±13.7/51.2 ±13.6	Isoflurane and fentanyl	0.5 mg/kg then 3 mg/kg/h	Normal saline	-	Laparoscopic cholecistectomy
Haghighi et al60	41/41	37.5 ±13.4/ 38.7 ±12.9	Not available	0.5 mg/kg after induction 30 min then 0.3 mg/kg/h	Normal saline	-	Tibial fracture fixation
Hwang et al61	28/28	39.3 ±9.5/ 38.4 ±11.0	Remifentanil	0.5 mg/kg then 1.8 mg/kg/h	Normal saline	-	Laparoscopic gynecologic surgery
Kavak Akelma et al62	16/16	47.0 ± 14.0/ 51.0 ± 14.0	Sevoflurane and fentanyl	1 mg/kg before induction then 3 mg/kg/h	Normal saline	Lidocaine	Laparoscopic cholecistectomy
Kim et al63	13/13	53.7 ±12.6/ 53.4 ±13.8	Sevoflurane and remifentanil	0.5 mg/kg after induction then 1.8 mg/kg/h	Normal saline	Subclinical esmolol	Laparoscopic gastrectomy
Lee et al65	20/20	51.9 ±14.1/ 43.2 ±12.3	Nicardipine	0.5 mg/kg then 0.6 mg/kg/h	Normal saline	Ketamine	Laparoscopic cholecistectomy
Lee et al64	30/30	36.3 ±7.70/ 34.5 ±6.10	Remifentanil	1 mg/kg after induction then 0.6 mg/kg/h and 1 mg/kg before extubation	Normal saline	-	Laparoscopic appendectomy
Mahrose et al26	30/30	31.2 ±8.11/ 32.4 ±6.92	Sevoflurane and fentanyl	1 mg/kg after induction 30 min then 0.6 mg/kg/h	Normal saline	-	Laparoscopic bariatric
Mendonça et al24	32/34	53.3 ±10.9/ 51.2 ±11.7	Fentanyl	0.5 mg/kg 10min then 6 mg/kg/h	Normal saline	-	Mastectomy
Moon et al66	27/27	37.9 ±11.7/ 40.9 ±11.0	Sevoflurane	0.5 mg/kg then 1.8 mg/kg/h	Normal saline	-	Laparoscopic gynaecological surgery
Morais et al25	20/20	35.8 ±10.9/ 33.2 ±8.70	Remifentanil	0.5 mg/kg after induction then 0.9 mg/kg/h	Normal saline	-	Laparoscopic gastroplasty bypass
Ozturk et al67	20/20	62.7[46–70]/ 61.3 [44–70]	Alfentanil	1mg/kg before induction then 0.3–0.6 mg/kg/h	Ringers Lactate	-	Laparoscopic cholecistectomy
White et al54	15/15	41.0 ±11.0/ 37.0 ±6.00	Esmolol and desflurane	50 mg then 0.3 mg/kg/h	Normal saline	Esmolol + nicardipine	Gynecological laparoscopic surgery
Srivastava et al68	29/28	47.5 ±10.1/ 42.3 ±8.57	Nitroglycerine	1 mg/kg after induction 5 min then 0.5 mg/kg/h	Normal saline	Dexmedetomidine	Laparoscopic cholecistectomy
Sultan et al69	30/30	37.3 ± 3.30/ 38.0 ± 3.10	Propofol and hydralazine	1 mg/kg before induction then 1.8 mg/kg/h	Normal saline	-	Gynecological laparoscopic surgery

± standard deviation.

“Not analyzed” indicates the drugs used in the third arm of individual studies, which were not included in the statistical synthesis.

Abbreviations: C, control; E, esmolol; E/C, esmolol/control.

Type of Surgery

The included studies involved different types of surgery, which may increase heterogeneity. Patients underwent minor surgeries in 2 studies: hysteroscopic surgery and septoplasty.^56,58 The remaining studies involved medium to major surgeries, including: laparoscopic cholecystectomy,^{59,62,65,67,68} laparoscopic gynaecological surgeries,^54,61,66,69 laparoscopic bariatric surgery,²⁶ laparoscopic gastrectomy,⁶³ laparoscopic gastroplasty bypass,²⁵ mastectomy,²⁴ laparoscopic appendectomy,⁶⁴ lower abdominal surgeries,⁵⁵ abdominal total hysterectomy,⁵⁷ and tibial fracture fixation.⁶⁰ However, it is worth noting that 13 studies used laparoscopic techniques, which typically involve less intense pain stimuli.

Intervention Dosing Regimens

All studies initiated continuous esmolol infusion immediately after a bolus dose. Esmolol infusion continued until the end of surgery in all studies (Table). Eighteen studies calculated the bolus dose based on the patient’s weight, ranging from 0.5 to 1.0 mg/kg. Bolus doses were 0.5 and 1 mg/kg in 12 and 6 studies, respectively. Only 1 study administered the same bolus dose (50 mg IV) to all intervention group patients.⁵⁴ This same study adjusted the esmolol infusion rate to maintain a heart rate between 65 and 75 bpm. Infusion rates ranging from 0.3 to 6 mg/kg/h. In 1 study, the corresponding author was contacted due to a suspected data discrepancy concerning the esmolol infusion rate (reported as 0.5 mg/kg/min).⁵⁵ The corresponding author confirmed that the correct rate administered in the trial was 0.05 mg/kg/min (3 mg/kg/h). In 1 study, the infusion started at 0.6 mg/kg/h and was increased to 1 mg/kg/h just before extubation.⁶⁴

Intraoperative Opioid Dosing Regimens

All studies administered opioids to both the esmolol and placebo groups (Supplemental Digital Content 3, Table S2 https://links.lww.com/AA/F520). Opioid regimens met the same parameters for both the esmolol and placebo groups, except in 1 study that performed an additional bolus of 10 µg of remifentanil for the control group during induction.⁶² Most frequent opioid regimen was bolus fentanyl,^{24,26,54,55,57,59,60,68,69} followed by continuous infusion of remifentanil,^{25,56,61–66} and only 1 study used continuous infusion of alfentanil.⁶⁷ In 9 studies, fentanyl boluses ranging from 1 to 3 µg/kg were administered at induction. Only 1 study utilized fentanyl exclusively for intraoperative analgesia maintenance, without bolus administration during induction.⁵⁸ Another study administered fentanyl only during induction, without increasing intraoperative opioids.⁶⁹ One study employed a fixed rescue interval every 60 minutes.⁵⁵ One study did not specify the parameters for administering additional fentanyl doses.⁵⁷ In 2 studies, opioid maintenance was not reported.^54,60 Continuous infusion of remifentanil was performed in 7 studies, of which 3 were initiated after induction and 4 were initiated during induction. Four studies performed target-controlled remifentanil infusion,^61,63,65,66 only 1 did not report the initial target concentration.⁶³ Three studies employed weight-based initial infusion rates of remifentanil, ranging from 0.05 to 0.5 µg/kg/min.^25,56,64 One study performed a bolus of 10 µg/kg alfentanil during induction and alfentanil continuous infusion for maintenance (0.5–3.0 µg/kg/min).⁶⁷

Five studies had third arms, data not included in this systematic review. The medications administered in the third arm of each study are listed under the “Not Analyzed” column in the Table.^{54,62,63,65,68}

Anesthetic and Hemodynamic Management

Included studies utilized various intraoperative hemodynamic management regimens (Table). Volatile anesthetics, including isoflurane,^57,59 sevoflurane,^26,62,63 and desflurane,⁵⁴ were used in 7 studies. Fentanyl (7 studies)^{24,26,55,57–59,62} and remifentanil (5 studies)^{25,56,61,63,64} were the most commonly used opioids for analgesia or hemodynamic stability. For hemodynamic parameters, fentanyl alone was used in 3 studies^24,55,58 and remifentanil alone in 4 studies.^25,56,61,64 One study did not specify whether hemodynamic parameters were used to manage anesthesia.⁶⁰ Additionally, nicardipine⁶⁵ and nitroglycerin⁶⁸ were each administered in 1 study for blood pressure control. One study used alfentanil,⁶⁷ one used propofol or hydralazine,⁶⁹ and another used esmolol and desflurane.⁵⁴ The approach to anesthetic and hemodynamic management during the intraoperative period may be a source of variability across studies. Patients with hemodynamic parameters managed exclusively with opioids may have exhibited a more pronounced opioid-sparing effect from esmolol, as their greater exposure to opioids makes any reduction more evident.

Perioperative Non-Opioid Analgesic Adjuncts

Of the 8 studies reporting perioperative analgesic adjuncts, 5 administered them intraoperatively and 3 postoperatively, using identical regimens for both esmolol and placebo groups (Supplemental Digital Content 8, Supplemental Table S4, https://links.lww.com/AA/F525). Only 1 study showed an imbalance in adjunctive analgesic administration between study arms.⁶⁴ Among the intraoperative protocols, 2 studies employed NSAIDs alone: diclofenac 75 mg intramuscularly 15 minutes before the end of surgery,¹⁴ and ketorolac 30 mg intravenously 30 minutes before the end of surgery.⁶⁵ One study used parecoxib 40 mg IV combined with dipyrone 2 g IV before induction.²⁵ One study administered a combination of diclofenac 75 mg IM and dexamethasone 4 mg IV at induction,⁶⁷ while another administered dexamethasone post-induction only in patients who experienced more than 3 postoperative episodes of emesis.⁵⁷ Postoperative adjunct protocols included ketorolac 30 mg IV in the PACU,⁵⁸ diclofenac administered as a rescue medication (used more frequently in the control group),⁶⁴ dexamethasone for patients who experienced more than 3 postoperative episodes of emesis,⁵⁷ and dipyrone 1 g IV every 6 hours.²⁴

Postoperative Pain Assessment

Of the 10 studies reporting postoperative pain using VAS scores,^{26,55–57,59,60,62–65} 9 assessed pain at multiple time points and one at a single time point (Supplemental Digital Content 3, Supplemental Table S2, https://links.lww.com/AA/F520).⁶³ Only 1 study assessed both VAS on movement and rest.⁵⁷ Among the 6 studies reporting postoperative pain using the NRS,^{24,25,58,61,66,69} 2 assessed pain at multiple time points.^24,25 One study reported postoperative pain outcome using area under NRS score curve (AUC) over time.⁵⁸ Only 1 study performed both NRS score under stress and rest.²⁴ Only 1 study reported the VAS score on a scale of 0 to 100, requiring standardization and conversion of values to a scale of 0 to 10 for statistical synthesis.⁶⁴

Analgesic Rescues and Time of First Rescue

Of the 16 studies pre-established postoperative analgesic administration, 13 provided quantitative values of this outcome (Supplemental Digital Content 3, Table S2 https://links.lww.com/AA/F520). Six studies reported the outcome in milligrams of morphine,^{24,25,55–58} 5 reported postoperative fentanyl consumption,^{61,62,65,66,69} and 1 related the pethidine dose at PACU.⁶⁰ Additionally, one of these studies also reported the consumption of tramadol, ketorolac, and diclofenac between 1 and 6 hours postoperatively.⁶⁵

Analgesic rescue was administered via patient-controlled analgesia (PCA) in 3 studies,^56,57,62 2 using morphine and 1 using fentanyl.⁶² Intravenous opioid boluses were employed for rescue analgesia in 13 studies: 5 used fentanyl,^{54,61,65,66,69} 3 utilized morphine,^24,25,55 3 administered tramadol,^26,59,67 1 study used hydromorphone,⁵⁸ and another employed pethidine.⁶⁰ Additionally, 1 study exclusively used intramuscular diclofenac.⁶⁴ Across these studies, 1 employed rescue metamizole before resorting to morphine,²⁴ and another used rescue diclofenac before tramadol for managing lower levels of pain.⁶⁷ Seven of these studies reported a statistically significant reduction in patients requiring rescue analgesics.^{24,26,54,56,59,66,67}

Among the 6 studies evaluating the time required for first opioid rescue, only 2 showed a statistically significant increase in esmolol group,^56,69 while remaining studies found no difference.^24,25,58,60 In the study that utilized metamizole in addition to morphine for analgesic rescue, a significant increase was observed in the time required for first rescue with metamizole in esmolol group, whereas no difference was noted between groups for morphine rescue.²⁴ Detailed evaluations of each study reporting time to first analgesic rescue and number of patients requiring additional analgesia are provided in Supplemental Digital Content 9, Supplemental Table S5, https://links.lww.com/AA/F526.

META-ANALYSIS AND META-REGRESSION

Intraoperative Opioid Dose

Primary Analysis

Intraoperative esmolol infusion significantly reduced opioid consumption compared to placebo (MD, –12.89 IV MME; 95% confidence interval [CI], –24.74 to –1.05; P < .001; I² = 93.6%), based on data from 649 participants across 11 studies (Figure 2A). The absolute reduction of 12.89 IV MME, representing a 32% relative decrease in intraoperative opioid consumption, supports the clinical significance of the intervention. The certainty of evidence was rated as moderate due to baseline imbalances and considerable heterogeneity.

Figure 2.

A, Forestplot of intraoperative opioid consumption, standardized as IV MME. B, Forestplot of opioid consumption within 24 h postoperatively, standardized as IV MME. C, Forestplot of pain scores within 30 min postoperatively. D, Forestplot of pain scores between 2 and 4 h postoperatively. E, Forestplot of pain scores at 24 h postoperatively. CI indicates confidence interval; MD, mean differences; MME, morphine milligram equivalent; SD, standard deviation; IV, intravenous.

Subgroup Analysis and Meta-Regression

Esmolol reduced postoperative opioid use under both intraoperative strategies, but the effect was greater when only opioids were used (MD, –25.57 IV MME; 95% CI, –51.19 to –0.05; P =.049; I² = 94.6%) compared to when adjuncts were included (MD, –1.49; 95% CI, –2.34 to –0.45; P =.003; I² = 31.4%), with no significant difference between groups (P =.064). Meta-regression showed the type of hemodynamic strategy as an factor influencing the effect (P =.035), explaining 22.9% of the variability across studies (R² = 22.9%), although high heterogeneity remained (I² = 90.2%; P =.010). This suggests that esmolol may be especially beneficial in settings where opioid use is more intensive.

In subgroup analysis by surgical type, esmolol significantly reduced intraoperative opioid use in medium-to-major procedures (MD, –7.98 IV MME; 95% CI, –13.83 to –2.13; P < .001; I² = 92.0%), with a 24% relative decrease, considering the mean opioid dose in the placebo group (33.26 IV MME). This effect may help reduce opioid-related adverse events in patients undergoing more extensive surgery. No significant effect was observed in minor procedures (MD, –39.32 IV MME; 95% CI, –117.71 to 39.07; P =.326; I²= 98.2%), likely due to imprecision from the small number of studies (n = 2). The test for subgroup differences was not significant (P =.434), suggesting that esmolol may reduce opioid use regardless of surgical type, though the benefit appears more consistent in larger procedures.

Intraoperative opioid type (P =.081), patient age (P =.096), and esmolol infusion rate (P =.972) showed no significant impacts.

Sensitivity Analysis

The exclusion of 1 outlier was performed,⁵⁶ supporting the robustness of results (MD, −7.08 IV MME; 95% CI, −12.94 to −1.68; P =.010; I²= 91.1%). Visual inspection of the funnel plot, along with Egger’s test (P =.055), did not suggest significant asymmetry, indicating a low likelihood of publication bias.

Postoperative Opioid Consumption

Primary Analysis

The synthesis included 695 participants from 12 studies (Figure 2B). Patients receiving intraoperative esmolol required significantly fewer opioids postoperatively compared to placebo (MD, –3.03 IV MME; 95% CI, –4.29 to –1.76; P < .001; I² = 89.9%), with a mean absolute reduction of 3.03 IV MME, representing a 38.6% relative decrease. Despite substantial heterogeneity (I² = 89.9%), the high certainty of evidence is supported by the strength of the association and the persistence of the effect even in the presence of adjunctive analgesics, a confounder expected to diminish between-group differences.

Subgroup Analysis and Meta-Regression

Esmolol significantly reduced perioperative opioid consumption regardless of adjunctive analgesic use, with consistent effects when adjuncts were not reported (MD, –2.37 IV MME; 95% CI, –3.45 to –1.29; P < .001; I² = 89.2%) and when administered (MD, –3.61 IV MME; 95% CI, –6.06 to –1.17; P =.008; I² = 87.2%). The absolute reduction was slightly greater in the presence of the confounding factor that typically biases the effect toward the null, despite no significant difference between subgroups (P =.361).

Subgroup analysis based on esmolol infusion rate revealed a possible dose effect, with higher infusion rates (≥3 mg/kg/h) associated with a greater reduction in postoperative opioid consumption. The effect was twice as large in the high-dose subgroup (MD, –4.42 IV MME; 95% CI, –6.04 to –2.80; P =.003; I² = 74.2%) compared to the lower-dose group (MD, –1.96 IV MME; 95% CI, –3.36 to –0.56; P < .001; I² = 80.1%), with the between-group difference demonstrating statistical significance (P =.024). A meta-regression analysis was conducted to explore the relationship between esmolol infusion rate and effect size on postoperative opioid consumption. Although the bubble plot (Figure 3) visually suggests a pattern of greater effect with higher doses, the association was not statistically significant (MD, –0.40 IV MME; 95% CI, –1.21 to 0.41; P =.331; I² = 90.9%). This analysis was potentially limited by the small number of studies, outlier’s presence, methodological variability in outcome measurement, and the use of similar exposure levels to intervention across trials, which may have hindered the detection of dose-related effects. Given its exploratory nature, this analysis should be interpreted with caution and regarded as hypothesis-generating rather than confirmatory.

Figure 3.

Bubble plot illustrating the influence of esmolol infusion rate (mg/kg/h) on mean differences (MD) across individual studies in the meta-regression analysis. Light beige bubbles represent outliers studies, while brown bubbles correspond to nonoutliers studies.

In the subgroup analysis by surgical type, esmolol significantly reduced postoperative opioid use in patients undergoing medium/major procedures (MD, –3.22 IV MME; 95% CI, –3.33 to –2.10; I² = 90.7%; P < .001), representing a 37.9% decrease relative to the weighted mean opioid dose in the placebo group, reinforcing the clinical relevance of esmolol’s opioid-sparing effect. No significant reduction was found in minor surgeries (MD, –1.92 IV MME; 95% CI, –8.99 to 5.14; I² = 88.6%; P =.594), likely due to the limited number of studies (n = 2). Given the limited data and high variability in minor procedures, these findings supports the use of esmolol primarily in medium/major surgeries.

No significant associations with effect size were observed in meta-regression for patient age (P =.520), time of measurement (p= 0.089), opioid type (P =.529), or the use of adjunct analgesics (P =.455).

Sensitivity Analysis

The exclusion of 3 outlier studies,^57,58,69 improved consistency and supported the robustness of the primary analysis, yielding a pooled effect size of MD –3.23 IV MME (95% CI, –3.99 to –2.48 IV MME; P < .001) with moderate heterogeneity (I² = 51.7%). Visual inspection of the funnel plot, along with Egger’s test (P =.058), did not suggest significant asymmetry in the primary synthesis, indicating a low likelihood of publication bias.

Pain Score

Primary Analysis

Several studies reported this outcome at multiple time points using VAS or NRS scores. Five of these studies related pain scores without this allowed data to be synthesized in 3 meta-analysis at different intervals (Figure 2C–2E). Analysis included 12 (n = 626),^{24–26,55,60–66,69} 7 studies(n = 417),^{24–26,57,60,62,65} and 6 studies (n = 355),^{24–26,57,62,64} respectively. Esmolol significantly reduced pain scores at all assessed time points: within 30 minutes (MD, –1.47; 95% CI, –2.02 to –0.93; P < .001), at 2–4 hours (MD, –0.67; 95% CI, –1.29 to –0.06; P =.032), and at 24 hours postoperatively (MD, –0.48; 95% CI, –0.92 to –0.03; P =.038). These correspond to relative reductions of 27.3%, 15.8%, and 23.5%, respectively, based on weighted mean placebo scores. These reductions are consistent with established thresholds for clinical significance and underscore the reliability of esmolol’s analgesic effect. Certainty of evidence was high at 30 minutes due to the strong association, but lower at later time points due to inconsistency and imprecision, with confidence intervals approaching null effect.

Subgroup Analysis and Meta-Regression

All 3 analyses exhibited substantial inconcistence (I² = 97.8%, 91.2%, and 88.8%, respectively). However, only the synthesis within the first 30 minutes postoperatively included a sufficient number of studies (more than 10) to allow for a reliable assessment of heterogeneity.

Esmolol significantly reduced postoperative pain within 30 minutes across subgroups, regardless of adjunctive analgesic use or intraoperative opioid type. In studies without reported adjuncts, the effect was significant (MD, –1.48; 95% CI, –2.36 to –0.61; I² = 98.3%), and even more consistent when adjuncts were used (MD, –1.54; 95% CI, –1.80 to –1.29; I² = 25.2%), despite no significant subgroup difference (P =.88). In addition, esmolol reduced pain both with nonremifentanil opioids (MD, –1.43; 95% CI, –1.76 to –1.10) and remifentanil (MD, –1.61; 95% CI, –2.62 to –0.60), with no significant interaction (P =.74). These findings, suggest that esmolol’s early analgesic effect is robust across analgesic strategies and anesthetic regimens, enhancing its clinical applicability.

Meta-regression analysis did not identify a significant influence of mean patient age (P =.069) or esmolol infusion rate (P =.539) on the variability of effect sizes across studies. Similarly, subgroup analysis revealed no significant difference in pain reduction between groups receiving higher versus lower doses of esmolol (P =.052). The type of surgery could not be analyzed as a moderator because no included study met the criteria for the “minor surgery” category.

Sensitivity Analysis

The exclusion of 2 outliers from the primary analysis within the first 30 minutes postoperatively,^26,55 supported the robustness of results in the sensitivity analysis, yielding an MD of –1.51 (95% CI, –1.65 to –1.36; P < .001) and substantially improved consistency (I² = 16.2%). Five studies included in the primary synthesis within the 30-minute time frame reported postoperative pain scores but did not report the total intraoperative opioid dose,^{55,60,62,63,69} despite describing identical opioid regimens for both the placebo and esmolol groups. To address the potential influence of reporting bias, we conducted a subgroup analysis isolating these studies. The effect size in the subgroup lacking total intraoperative opioid dose reporting was slightly smaller (MD, –1.18; 95% CI, –2.02 to –0.33; P =.006) than in the subgroup that reported it (MD, –1.67; 95% CI, –2.38 to –0.97; P < .001), but this difference was not statistically significant (test for subgroup differences: P =.376). Visual inspection of the funnel plots, together with Egger’s tests for pain assessments at 30 minutes (P =.146), 2–4 hours (P =.930), and 24 hours (P =.308), revealed no significant asymmetry, suggesting a low risk of publication bias.

Adverse Event Outcomes

Of the 19 studies, 12 compared intraoperative HR and MAP between groups. Among these, 8 reported significant HR reductions in the esmolol group,^{24,26,54,55,57,59,64,68,69} while 3 found no significant differences.^56,62,65 Two studies reported a significant difference only during time points corresponding to intubation and/or extubating favorable to esmolol,^57,64 1 evidenced significant reduction in multiple time points for esmolol group,⁶⁹ and 1 reported it during intubation and skin suturing in patients that received esmolol.⁵⁴ Intraoperative MAP was lower in the esmolol group in multiple time points for 6 studies,^{24,26,59,64,68,69} while 6 found no significant differences.^{54–57,62,65} Only 1 of these studies demonstrated a significant difference exclusively during intubation and extubation.⁶⁴ In this systematic review, there were no reports of significant differences in the occurrence of hypotension or bradycardia. One study reported a higher use of ephedrine and 1 atropine bolus in the esmolol group, though neither was statistically significant.²⁴

Among the 12 studies that reported incidence of nausea or vomiting,^{24–26,54,56–58,61,62,64,66,67} only 1 demonstrated a significant reduction for esmolol group.²⁶ The number of patients who required an additional antiemetic dose was reported by 6 studies, with no statistically significant differences between the groups.^{24,54,61,64,66,67} Therefore, the esmolol infusion results in little to no difference in PONV and antiemetic rescues.

DISCUSSION

In this systematic review, meta-analysis, and meta-regression of 19 randomized placebo-controlled trials and 1028 participants, we assessed effects of continuous intraoperative esmolol infusion in outcomes related to the central objectives of multimodal anesthesia. Pooled analysis revealed the following key outcomes among patients receiving esmolol: (1) significantly decreases intraoperative opioid use; (2) a significant reduction in postoperative opioid dose, even in settings where adjunctive analgesics were routinely administered; (3) higher esmolol dosing was associated with greater postoperative opioid sparing compared to lower dosing strategies; and (4) significantly lower pain scores at 3 time points within the first 24 hours postoperatively. Findings on opioid consumption are consistent with previous meta-analysis by Gelineau et al, although with notable differences in key aspects, available in Supplemental Digital Content 10, Supplemental Table S6, https://links.lww.com/AA/F527.

Direct analgesic effects are highly desirable in drugs considered for multimodal anesthesia.⁷⁰ Our analysis revealed a significant reduction in intraoperative opioid requirements in the esmolol group, suggesting that this beta-blocker may possess direct analgesic properties. However, the level of certainty in our findings is moderate, primarily due to inconsistencies among the included studies. This limitation can be partly attributed to variations in hemodynamic management protocols, as the use of volatile anesthetics and vasodilators in some studies that may have underestimated the true opioid-sparing effects of esmolol.^{26,57,58,65,66,68} Despite the high variability, our results are corroborated by studies demonstrating direct pain reduction during rocuronium and propofol infusion in patients premedicated with esmolol.^13–15 The mechanism underlying esmolol’s effect may be associated with a reduction in sympathetic nociceptive excitability and an enhancement in the release of inhibitory neurotransmitters by neurons involved in central pain modulation, as demonstrated in animal models.^10,71–73 Furthermore, it is possible that esmolol exerts effects similar to lidocaine on dorsal root ganglion neurons by inhibiting tetrodotoxin-resistant sodium channels (TTX-r Na+), which play a critical role in the conduction of nociceptive signals, as evidenced in rat models.^9,74–77

Reducing postoperative opioid consumption to mitigate opioid-related adverse events is a core objective of multimodal anesthesia guidelines.^1,3,28 This meta-analysis highlights the significant potential of multimodal anesthesia with esmolol to reduce opioid consumption within the first 24 hours postoperatively, which may be related to possible modulatory effects on pain sensitization. Although heterogeneity was substantial, the overall certainty of evidence remained high, supported by the magnitude of the effect and its persistence despite potential confounding. Six studies in our synthesis reported the use of adjunctive analgesics such as NSAIDs, dipyrone, or dexamethasone: 5 applied identical regimens across treatment arms, and only 1 showed imbalance in diclofenac dosing.⁶⁴ This relative balance likely introduced a conservative bias, attenuating between-group differences and shifting effect estimates toward the null. Nevertheless, esmolol consistently reduced postoperative opioid use, indicating that it remains effective even when standard adjuncts are part of the anesthetic plan.

Subgroup analysis suggests that a higher esmolol infusion rates (≥3 mg/kg/h) may be associated with a greater reduction in postoperative opioid consumption. Although this association was not statistically significant in meta-regression based on infusion rate as a continuous moderator, the directional trend and proximity of the confidence interval to the null value raise the hypothesis that infusion rate may influence the magnitude of effect. This hypothesis is biologically plausible given esmolol’s pharmacokinetic attributes: higher infusion rates cause proportional and uniform increases in the level of beta blockade.²³ In addition, may also be linked to the dose-dependent relationship of esmolol with tetrodotoxin-resistant sodium channels (TTX-r Na⁺) present in dorsal root ganglion neurons.⁹ Furthermore, reduced selectivity of esmolol at higher doses may contribute to modulatory effects mediated by β2-adrenoreceptors.²³ Studies in animal models have also demonstrated that blocking these receptors can reduce the release of pain-related proinflammatory cytokines by macrophages and mast cells, as well as attenuate the action of catecholamines in central nervous system, particularly in microglial regions associated with pain sensitization.^7,78–80 This observation, while exploratory, supports the need for adequately powered dose-finding trials to confirm whether higher esmolol exposure enhances analgesic outcomes.

The 3 analyses conducted at different time points revealed a significant reduction in postoperative pain scores among patients who received esmolol. However, this analgesic effect appeared to diminish progressively over time. Finding of greater magnitude of the effect in the first minutes after surgery may reflect a residual effect of esmolol that was not observed in subsequent intervals due to its short half-life.²³ Within the first 30 minutes postoperatively, esmolol reduced pain by 1.47 points, representing a 27.3% relative decrease compared to placebo. This magnitude approaches the threshold for moderate clinical relevance, as defined by IMMPACT guidelines.⁴² While group-level estimates do not capture individual variability, such early analgesia may enhance patient comfort and reduce immediate opioid needs in the post-anesthesia care unit.

Among the 12 studies evaluating postoperative opioid use, 7 involved laparoscopic abdominal procedures, increasing to 7 of 10 in the medium-to-major subgroup. Under these surgical conditions, meta-analyses have shown that nonopioid adjuncts such as magnesium,⁸¹ lidocaine,⁸² ketamine,⁸³ acetaminophen,⁸⁴ and dexmedetomidine⁸⁵ reduce 24-hour postoperative opioid requirements by 3.85 to 18.13 IV MME (absolute) and 5.4% to 38.6% (relative). In bariatric laparoscopic surgery, reductions range from 2.2 to 6.4 IV MME.⁸⁶ The opioid-sparing effect of esmolol observed in our primary analysis (–3.03 IV MME) and in medium-to-major procedures (–3.22 IV MME) falls within this range. Given accumulating evidence linking higher perioperative opioid use to adverse outcomes (eg, nausea, sedation, ileus), along with increased costs and length of stay, esmolol represents a promising addition to multimodal analgesia strategies.^87–89

The intraoperative opioid-sparing effect of esmolol was associated with a 32% relative reduction overall and 24% in medium-to-major surgeries, based on placebo group means. Although the pooled mean intraoperative opioid consumption in the medium-to-major surgery subgroup of our synthesis was higher (33.26 IV MME), it remained within the range reported in a large retrospective cohort of 153,902 surgical patients (mean: 20.12 IV MME; range: 8–41 IV MME), providing a useful external benchmark.⁹⁰ Extrapolating from this reference, esmolol could reduce intraoperative opioid use by approximately 39% in broader surgical populations. This effect may be particularly valuable in ambulatory surgery, where even modest increases in opioid exposure have been linked to higher rates of nausea, sedation, and delayed discharge.^90–92 By minimizing these complications, esmolol may support faster recovery, greater patient satisfaction, and reduced healthcare costs in fast-track pathways.

Perioperative beta-blocker safety remains a valid concern, particularly after the POISE trial linked metoprolol to increased cardiovascular complications.⁹³ However, in several studies, esmolol’s ultra-short-acting profile and rapid titratability have been shown to offer a safer alternative in the intraoperative setting, even in patients with cardiovascular risk.^94,95 In this meta-analysis, esmolol was not associated with a significant increase in bradycardia or hypotension in any of the included studies. Moreover, its opioid-sparing and analgesic effects appeared to be independent of hemodynamic adverse events, demonstrating that its benefits extend beyond cardiovascular modulation.

The high heterogeneity and uncertainty of effects in later time periods (2–4 hours and 24 hours) indicate that the efficacy of esmolol may decrease after discontinuation of its intraoperative administration. While esmolol showed consistent opioid-sparing effects in medium-to-major surgeries, its impact was not statistically significant in minor procedures, suggesting limited applicability in less painful surgical settings. Further studies are needed to assess continuous postoperative esmolol regimens and its potential in prolonged pain, minor surgeries, and specific populations, such as pediatric and geriatric patients. In addition, the use of esmolol infusion may pose an economic barrier to routine implementation, given the relatively high cost of the drug and the additional clinician time required for preparation and titration.

The heterogeneity investigation revealed that conventional moderators such as the type of intraoperative opioid and the mean age of participants did not account for the variability in effect magnitude across studies for any of the evaluated outcomes. In contrast, the anesthetic and hemodynamic management strategy emerged as a clinically relevant factor, significantly influencing the between-study differences in effect size. Furthermore, sensitivity analysis excluding statistical outliers in the postoperative opioid consumption and pain score outcomes markedly improved consistency (I²= 51.7% and I² = 16.2%, respectively), indicating that those studies were key contributors to the observed heterogeneity.

Implications for Research

Given these findings, we can highlight important recommendations for future research: (1) esmolol should be preferentially employed in anesthetic regimens involving short-acting opioids, given the higher susceptibility to opioid-induced hyperalgesia associated with these agents; (2) given esmolol’s potential dose-effect and favorable safety profile, future trials should compare placebo with multiple esmolol doses across different groups, exploring the therapeutic range of 1.2 to 9 mg/kg/h within established safety limits;²³ (3) to address the high heterogeneity observed in pain outcome analyses,⁴⁴ future studies should include detailed baseline data on participant’s gender, psychological status, chronic comorbidities, and socioeconomic factors;^96–99 (4) ensure consistent reporting outcomes at standardized intervals during first 24 postoperative hours; and (5) encourage further animal studies to elucidate mechanisms linking pain modulation and beta-adrenoreceptors.

Strengths and Potential Limitations

Our analysis included only randomized, placebo-controlled studies, minimizing confounding, comparability, and expectation biases. Studies involving peripheral nerve blocks were excluded to isolate esmolol’s specific effects. A comprehensive search across multiple databases (Medline, Embase, Cochrane, and Google Scholar) and gray literature ensured broad coverage. Meta-regression, subgroup and sensitivity analyses explained much of the heterogeneity in opioid consumption outcomes. To enhance consistency, opioid data were converted to intravenous morphine milligram equivalents (IV MME), and pain scores (VAS and NRS) were standardized using validated methods, improving the reliability of the results.

Substantial heterogeneity remained in several outcomes, particularly in pain scores at later time points, likely influenced by esmolol’s short half-life and limited long-term data. Some subgroup analyses and meta-regressions were constrained by the small number of studies, reducing statistical power. Even after an extensive search, we were unable to retrieve one of the articles screened for eligibility,⁵³ which could generate nonreporting bias. Additionally, 1 study lacked standard deviation data, preventing its inclusion in statistical synthesis.²⁵ The small sample size and wide CIs suggest low statistical power, reducing precision and increasing uncertainty. Sensitivity analyses were conducted to mitigate this imprecision. Only 3 studies published protocols,^24,25,58 highlighting the need for greater transparency in future research to reduce incomplete reporting. The inclusion of multiple outcomes and pain assessments at several time points, without designation of primary versus secondary endpoints, inherently increases the risk of type I error. This should be considered when interpreting the findings. A key limitation of our analysis is that postoperative pain scores were not consistently accompanied by intraoperative opioid dosing data. Given the direct influence of opioid administration on pain perception, the absence of this information may limit the interpretability of pain score comparisons across studies. Another limitation of this review is that a Medical Information Specialist (MIS) was not involved in the development or execution of the literature search strategy, which may have increased the risk of missing potentially relevant studies. Lastly, aggregate data meta-analyses are inherently limited by the potential for ecological fallacy.

This meta-analysis suggests that esmolol as an adjunct in multimodal anesthesia, reduces opioid consumption and postoperative pain, with a potential dose effect on postoperative opioid requirements. However, its analgesic benefits appear to diminish over time, and heterogeneity limits certainty in later outcomes. The long-term effects remain unclear, and further research is needed to define optimal dosing and assess its role in multimodal analgesia.

DISCLOSURES

Conflicts of Interest: None. Funding: None. This manuscript was handled by: Anna Woodbury, MD.

Supplementary Material

ane-142-882-s001.pdf ^{(96.7KB, pdf)}

ane-142-882-s002.pdf ^{(9.7KB, pdf)}

ane-142-882-s003.pdf ^{(206.8KB, pdf)}

ane-142-882-s004.pdf ^{(2.6MB, pdf)}

ane-142-882-s005.pdf ^{(129.7KB, pdf)}

ane-142-882-s006.pdf ^{(609.8KB, pdf)}

ane-142-882-s007.pdf ^{(142.5KB, pdf)}

ane-142-882-s008.pdf ^{(11.2KB, pdf)}

ane-142-882-s009.pdf ^{(289.4KB, pdf)}

ane-142-882-s010.pdf ^{(118KB, pdf)}