Intraoperative Methadone Reduces Pain and Opioid Consumption in Acute Postoperative Pain: A Systematic Review and Meta-analysis

Abstract

BACKGROUND: 

Methadone is a potent opioid exerting an analgesic effect through N-methyl-d-aspartate receptor antagonism and the inhibition of serotonin and noradrenaline reuptake. It has also been used in several procedures to reduce postoperative pain and opioid use. This meta-analysis aimed to determine whether the intraoperative use of methadone lowers postoperative pain scores and opioid consumption in comparison to other opioids.

METHODS: 

Double-blinded, controlled trials without language restrictions were included from MEDLINE, Embase, LILACS, The Cochrane Central Register of Controlled Trials (CENTRAL), and CINAHL via EBSCOhost. The included studies tracked total opioid consumption, postoperative pain scores, opioid-related side effects, and patient satisfaction until 72 hours postoperatively. Mean difference (MD) was used for effect size.

RESULTS: 

In total, 476 articles were identified and 13 were considered eligible for inclusion in the meta-analysis. In 486 patients (7 trials), pain at rest (MD, 1.09; 95% confidence interval (CI), 1.47–0.72; P < .00001) and at movement (MD, 2.48; 95% CI, 3.04–1.92; P = .00001) favored methadone 24 hours after surgery. In 374 patients (6 trials), pain at rest (MD, 1.47; 95% CI, 3.04–1.02; P < .00001) and at movement (MD, 2.03; 95% CI, 3.04–1.02; P < .00001) favored methadone 48 hours after surgery. In 320 patients (4 trials), pain at rest (MD, 1.02; 95% CI, 1.65–0.39; P = .001) and at movement (MD, 1.34; 95% CI, 1.82–0.87; P < .00001) favored methadone 72 hours after surgery. A Trial Sequential Analysis was performed and the Z-cumulative curve for methadone crossed the monitoring boundary at all evaluations, additionally crossing Required Information Size at 24 and 48 hours at rest. Methadone group also showed lower postoperative opioid consumption in morphine equivalent dosage (mg) at 24 hours (MD, 8.42; 95% CI, 12.99–3.84 lower; P < .00001), 24–48 hours (MD, 14.33; 95% CI, 26.96–1.91 lower; P < .00001), 48–72 hours (MD, 3.59; 95% CI, 6.18–1.0 lower; P = .007) postoperatively.

CONCLUSIONS: 

Intraoperative use of methadone reduced postoperative pain scores compared to other opioids, and Trial Sequential Analysis suggested that no more trials are required to confirm pain reduction at rest until 48 hours after surgery. Methadone also reduced postoperative opioid consumption and led to better patient satisfaction scores through 72 hours postoperatively compared to other opioids.

See Editorial, p 1456

KEY POINTS

• Question: How effective is intraoperative methadone administration in reducing postoperative pain scores and opioid consumption compared to other opioids?
• Findings: Pain scores until 72 hours postoperatively, postoperative opioid consumption, and patient satisfaction favor methadone over other opioids, and the incidence of opioid-related side effects was not increased with methadone use.
• Meaning: Intraoperative methadone is effective in reducing postoperative pain scores and opioid consumption.

GLOSSARY

CENTRAL = The Cochrane Central Register of Controlled Trials; CI = confidence interval; ERAS = Enhanced Recovery After Surgery; GRADE = Grading of Recommendations, Assessment, Development and Evaluation; I² = heterogeneity index; MD = mean difference; MED = morphine equivalent dose; MOR = μ-opioid receptor; NMDA = N-methyl-d-aspartate; NRS = Numerical Rating Scale; PACU = postanesthesia care unit; PCA = patient-controlled analgesia; PONV = postoperative nausea and vomiting; PRISMA = Preferred Reporting Items for Systematic review and Meta-analysis Protocol; PROSPERO = International Prospective Register of Systematic Reviews; RCT = randomized controlled trial; RevMan = Review Manager software; RIS = required information size; RR = relative risk; SD = standard deviation; VAS = Visual Analog Scale; VRS = Verbal Rating Scale

Acute postoperative pain is an important event related to increased morbidity, hospital costs, and mortality. Approximately 60% of patients undergoing surgical intervention experience moderate to severe postoperative pain.¹ Inadequately managed acute postoperative pain results in several deleterious consequences, mainly in the initial postoperative days, considered the most painful phase of recovery.^2,3 These include delayed ambulation; higher cardiovascular, respiratory, and thromboembolic morbidity; as well as mortality. Consequently, slower recovery in the postoperative period also increases hospital costs.^1–3 Opioids are frequently used in the perioperative period to manage moderate to severe pain. However, side effects can be significant including nausea, vomiting, pruritus, constipation, respiratory depression, and urinary retention, which may cause a negative impact in the quality of postoperative recovery.¹

Methadone is a strong μ-opioid receptor (MOR) agonist that exerts its activity through activation of MORs. Methadone also antagonizes N-methyl-d-aspartate (NMDA) receptors and inhibits the reuptake of serotonin and noradrenaline in the central nervous system. In addition, its half-life ranges from 13 to 50 hours.^4,5 Therefore, methadone may be more effective in acute pain management compared with other opioids mainly because of its long duration of effect, but also by modulating both pain stimuli propagation and analgesic descending pathways. Its pharmacological properties make methadone a unique opioid analgesic for anesthesia use, which can contribute to analgesia in the early postoperative period.

Some studies have compared intravenous methadone used in the intraoperative period with other opioids commonly used in anesthesia and have proven the analgesic benefit of this medication in a wide range of procedures, such as cardiac surgery, spine surgery, laparoscopy, abdominal surgery, and major surgery in children.^6–9 There is evidence that methadone use during anesthesia can significantly reduce postoperative opioid consumption, lower postoperative pain scores, and result in improved patient satisfaction with pain management.^6–9

Although methadone appears to be an effective opioid for the perioperative period, several important questions remain to be addressed. There are no published systematic reviews or meta-analyses on the subject. This systematic review and meta-analysis aimed to evaluate the efficacy and safety of intraoperative methadone compared with other commonly used opioids for the control of acute postoperative pain and postoperative opioid consumption.

METHODS

The present meta-analysis adhered to the recommendations of the Preferred Reporting Items for Systematic review and Meta-analysis Protocol (PRISMA)¹⁰ statement and was recorded in the International Prospective Register of Systematic Reviews (PROSPERO) database (CRD 42018108629). The entire review process followed the protocol, and the information was extracted and stored in Review Manager software (RevMan) version 5.3 (The Nordic Cochrane Center, The Cochrane Collaboration, Copenhagen, Denmark; 2014).

Search Strategy and Selection Criteria

The PRISMA guidelines were followed in this systematic review, meta-analysis, and trial sequential analysis. The first search was for previously published meta-analyses in the following databases: The Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE via Ovid SP, Embase via Ovid SP, CINAHL via EBSCOhost, and LILACS. Key words included: “methadone,” “opioids,” “surgery,” “postoperative,” “analgesia,” and the syntax elements (Supplemental Digital Content 1, Text 1, https://links.lww.com/AA/C924). Free text words and controlled vocabulary/MeSH terms were combined without any limitation in the search period. The MEDLINE search terms were adapted for each database. There were no language or status of publication (published or unpublished) limitations.

Trials that compared intravenous methadone used intraoperatively versus other opioids according to each study protocol were included. Studies that enrolled adults or pediatric population of any age undergoing any surgical procedure were included.

Outcomes

The primary end point was “pain management,” including 2 coprimary outcomes: (1) intensity of postoperative pain assessed 24, 48, and 72 hours after the procedure. Data were collected from Visual Analog Scale (VAS), Numerical Rating Scale (NRS), or Verbal Rating Scale (VRS) and were normalized ranging from 0 to 10 to grade the intensity of pain; with subgroup analyses for pain at rest and pain after movement; (2) postoperative opioid consumption, defined as the equivalent cumulative morphine dose used, was assessed in postanesthesia care unit (PACU), and 24, 48, and 72 hours after anesthesia.¹¹

The following secondary outcomes were compared among groups: incidence of postoperative nausea and vomiting (PONV), incidence of respiratory side effects, incidence of urinary retention, and incidence of cardiovascular side effects. Also, patient satisfaction with analgesia was evaluated at 24, 48, and 72 hours, ranging from 0 to 100. All categorical variables classified as “yes” or “no” were defined according to the authors of the studies.

Data Collection and Quality Assessment

Two authors performed the search, selected the relevant articles according to the eligibility criteria, and performed data extraction and content analysis independently. Disagreements were discussed with a third author.

When the information from the studies was insufficient or unclear to perform the data analysis, attempts were made to contact the study author to obtain additional information. The quality of evidence was evaluated using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) system for each outcome. GRADE ranks the level of evidence as high, moderate, low, and very low, considering 5 domains: risk of bias, inconsistency, indirectness, imprecision, and publication bias.^12,13 Evidence from each domain was subjected to downgrade following the classification: no (no downgrade), serious (downgrade 1 point), and very serious (downgrade 2 points), being punctuated by reviewers according to interferences detected in these items.¹⁴

Risk of Bias

Risk of bias assessment was performed using RevMan5.3 software according to the following criteria for each study included in the present review: selection bias (random sequence generation, allocation concealment), performance bias (blinding of participants and personnel), detection bias (blinding of outcome assessment), attrition bias (incomplete outcome data), reporting bias (selective reporting), and others.¹⁵ According to the Cochrane tool, the graduation established to assess the risk of bias is divided into high, low, and unclear. A high risk of bias was considered when any of the items evaluated in the studies were not performed or reported. A low risk of bias was considered when the items were accessed adequately. Risk of bias was considered unclear when the information available in the article was insufficient to classify each item as high or low risk of bias or was not properly reported in the article.^16–18

Statistical Analyses

Statistical analysis was performed using RevMan version 5.3. Dichotomous outcomes were defined as the presence or absence of an event (eg, occurrence of PONV). Results are expressed as relative risk (RR) with 95% confidence interval (CI). Mean difference (MD) and standard deviation (SD) were used for continuous outcomes (postoperative pain intensity, postoperative opioid consumption, and patient satisfaction). MD and SD were recovered from study data when available and were estimated when median, minimum and maximum values, and/or first and third quartiles were reported.¹⁹ Random-effects method was used for studies with high heterogeneity index (I² > 30%), and fixed-effects method was used for studies with lower heterogeneity (I² < 30%).

A 1-stage joint hypothesis testing was performed for the primary outcome (pain scores and postoperative opioid consumption). Intervention was considered more effective than control if there was superiority on all outcomes, as an intersection-union test with 2 hypothesis.¹¹ Both null hypothesis (pain scores and opioid consumption) needed to be rejected to claim a significant effect for methadone use. Correction for multiple testing was not required because the alternative hypothesis was an intersection (“AND”) of multiple tests. Significance level was set at an α = .025 for each outcome.

The heterogeneity in each meta-analysis was evaluated using the I². Heterogeneity was considered to be substantial if I² exceeded 25% or whether there was clear substantial inconsistency in the direction or magnitude of the effects as judged by visual inspection.²⁰ Additionally, a Trial Sequential Analysis was performed for pain scores to correct for the increased type I error resulting from multiple comparisons. An overall type I risk of 5% was maintained, with a power of 90%. We calculated the Required Information Size (RIS) considering a difference of 1.1 in pain scales to be clinically significant, according to recent literature in postoperative pain.^21,22 Small-study effect was accessed using funnel plot because there were not enough studies in each comparison to statistically test for publication bias using other methods.

RESULTS

A total of 476 articles were retrieved from the databases according to the literature search strategy of the current review. The steps involving the selection and exclusion stages of the articles are represented in the flow diagram proposed in the PRISMA statement (Figure 1). After initial evaluation, in which duplicate articles were excluded and the remainders were subjected to title and abstract analysis, 26 articles were read in their entirety. Twelve studies were excluded from review because they did not meet the proposed eligibility criteria. Thus, 14 randomized controlled trial (RCTs) were reviewed,^{6–9,23–32} involving a total of 929 patients. One study compared outcomes not evaluated in this review and was excluded from meta-analysis,²⁷ and only 1 study evaluated a pediatric population.⁹

Figure 1.: Flow diagram of the selection steps of the identified articles according to PRISMA. CENTRAL indicates The Cochrane Central Register of Controlled Trials; PRISMA, Preferred Reporting Items for Systematic review and Meta-analysis Protocol.

Thirteen studies were included in the meta-analysis, of which 9 studies used methadone in analgesic bolus doses ranging from 0.1 to 0.3 mg·kg⁻¹ at the beginning of the procedure^{6–9,23,25,26,29,32}; 3 studies used methadone 20 mg after anesthesia induction,^24,26,31 and 1 other used methadone 0.1 mg·kg⁻¹ at the end of the procedure.²⁸ Seven studies used methadone as sole opioid during anesthesia,^{6,8,9,24,29,31,32} while 6^{7,23,25,26,28,30} studies used methadone associated with other opioids in the intervention group. Control groups received morphine, fentanyl, sufentanil, or remifentanil. Surgical interventions varied among the studies; therefore, the cumulative dose of opioids in control groups also showed a wide variation. The included studies’ characteristics are summarized in Supplemental Digital Content 2, Table 1, https://links.lww.com/AA/C924. A summary of the risk for bias in the included studies is included in Supplemental Digital Content 3, Figure 1, https://links.lww.com/AA/C924.

Joint Hypothesis Testing: Postoperative Pain and Analgesic Requirements

Pain intensity in the postoperative period and postoperative opioid consumption was reported in 8 included articles. Meta-analysis was performed for pain at 24, 48, ad 72 hours after the procedure, providing a subgroup analyses in each assessment comparing the intensity of the pain at rest and at movement. The use of postoperative opioids was also evaluated in through the first 72 hours after surgery as the morphine equivalent dose (MED) in milligrams.

Pain at rest 24 hours after surgery was assessed in 7 included studies (486 patients).^{6–8,23,28,30,31} A MD in pain scores of 1.09 was obtained favoring methadone use, with a statistical difference (95% CI, 1.47–0.72 lower; P < .00001; I² = 0). A subgroup analysis of pain at movement 24 hours after surgery was performed with 3 studies (292 patients).^6–8 Methadone use showed a significant difference when compared to other opioids, with a reduction of 2.48 in pain scores favoring methadone (95% CI, 3.04–1.92 lower; P = .00001; I² = 20%) (Figure 2A). Methadone group also showed a MD in postoperative opioid consumption of 8.42 mg MED lower than control groups at 24 hours (7 studies, 460 patients, 95% CI, 12.99–3.84 lower; P < .00001; I² = 94%).^{6–8,23,29,30,32} A joint hypothesis testing for pain at rest, pain at movement, and opioid consumption 24 hours after surgery rejected all null hypothesis; therefore, intervention was considered more effective than control.

Figure 2.: Forest plot for comparison of pain scores at 24 (A), 48 (B), and 72 h (C) postoperatively. The table displays the study, mean, SD, sample size (total), difference in means with 95% CI, heterogeneity, overall effect, and P values. CI indicates confidence interval; df, degrees of freedom; I ², heterogeneity index; IV, intravenous; SD, standard deviation.

Six trials (374 patients)^{6–8,29–31} analyzed pain at rest 48 hours after the surgical procedure, showing a MD of 1.47 in pain scores favoring methadone (95% CI, 3.04–1.02 lower; P < .00001; I² = 70%). The subgroup analysis for pain at movement at 48 hours after surgery included 3 studies (295 patients),^6–8 and a MD of 2.03 in pain scores was obtained favoring methadone (95% CI, 3.04–1.02 lower; P < .00001; I² = 70%) (Figure 2B). In addition, methadone group showed a mean opioid consumption of 14.43 mg MED lower than control groups at 24–48 hours (4 studies, 332 patients, 95% CI, 26.96–1.91 lower; P < .00001; I² = 89%).^6–8,30 A joint hypothesis testing for pain at rest, pain at movement, and opioid consumption 24–48 hours after surgery rejected all null hypotheses; therefore, intervention was considered more effective than control.

Four studies (320 patients)^6–8,30 were included in the meta-analysis for pain at rest after 72 hours postoperatively, showing a MD of 1.02 in pain scores favoring methadone (95% CI, 1.65–0.39 lower; P = .001; I² = 0). The subgroup analysis for pain at movement at 72 hours after surgery included 3 studies (291 patients)^6–8 with a MD in 1.34 in pain scores obtained favoring methadone (95% CI, 1.82–0.87 lower; P < .00001; I² = 26%) (Figure 2C). Furthermore, methadone group showed a mean opioid consumption of 3.59 mg MED lower than control group at 48–72 hours (4 studies, 332 patients, 95% CI, 6.18–1.0 lower; P = .007; I² = 0%).^6–8,30 A joint hypothesis testing for pain at rest, pain at movement, and opioid consumption 48–72 hours after surgery rejected all null hypotheses; therefore, intervention was considered more effective than control.

Figure 3.: Forest plot for methadone versus control for opioid consumption at PACU, first 24, 24–48, and 48–72 hours. Data are displayed as morphine equivalent dose (MED) in milligrams. The table displays the study, mean, SD, sample size (total), difference in means with 95% CI, heterogeneity, overall effect, and P values. CI indicates confidence interval; df, degrees of freedom; I ², heterogeneity index; IV, intravenous; PACU, postanesthesia care unit; PO, postoperatively; SD, standard deviation.

There was no significant statistical difference between groups in MED consumption during the PACU period. Forest plot for opioid consumption is shown in Figure 3. The quality of evidence for pain at rest was classified as “very low” and for pain with effort was classified as “moderate” according to the GRADE system; while opioid consumption outcome was derived from a “low quality” level of evidence. A detailed quality evaluation of evidence level is included in Supplemental Digital Content 4, Text 2, https://links.lww.com/AA/C924.

Trial Sequential Analysis

All trials were included in the Trial Sequential Analysis for postoperative pain because most trials were considered to have high risk of bias. The estimated RIS for pain at rest 24 hours after surgery was 312, which was surpassed by recovered evidence in this meta-analysis. In addition, the cumulative Z-curve for pain 24 hours after surgery crosses the Trial Sequential Analysis monitoring boundary for benefit (Figure 4A), indicating there are sufficient data to support the pain score reduction for methadone use and no more trials are necessary to evaluate this effect. Trial Sequential Analysis–adjusted CI ranged from −3.79 to −1.79.

Figure 4.: Trial sequential analysis of the effect of intravenous methadone versus other opioids for postoperative pain control 24 h postoperatively at rest (A) and 48 h postoperatively at rest (B). Risk of type I error maintained at 5% with a power of 90%. Variance calculated from data deriving from included trials. Clinically significant reduction in pain scores was set at 1.1/10. Red lines on the left represent trial boundaries for efficacy (upper) or harm (lower). Blue line shows cumulative Z-statistics for included studies; each black dot representing 1 trial. Region inside meeting red lines at the far right indicates the futility region. Horizontal green lines indicate efficacy and harm boundaries with no adjust to repeated testing (conventional P = .05). Red vertical line is RIS. RIS indicates Required Information Size.

The estimated RIS for pain at rest 48 hours after surgery was 226, which was also surpassed by accumulated evidence. In addition, the cumulative Z-curve for pain 48 hours after surgery crosses the Trial Sequential Analysis monitoring boundary for benefit (Figure 4B), indicating there are sufficient data to support the pain score reduction for methadone use and no more trials are necessary to evaluate this effect. Trial Sequential Analysis–adjusted CI ranged from −2.10 to −0.83.

Cumulative Z-curve at Trial Sequential Analysis for other pain evaluations crossed the monitoring boundary for benefit but did not reached the estimated RIS. These results suggest a statistically significant difference favoring methadone over other opioids in the postoperative period evaluated after 24 hours at movement (adjusted CI, 3.79 to −1.39, reaching 69.9% RIS); 48 hours at movement (adjusted CI, 2.92 to −0.97, reaching 40% RIS); 72 hours at rest (adjusted CI, 1.92 to −0.12, reaching 55.6% RIS); and 72 hours at movement (adjusted CI, 1.92 to −0.77, reaching 97.7% RIS) (respectively: Supplemental Digital Content 5, Figure 2, https://links.lww.com/AA/C924).

Secondary Outcomes

Nine studies (573 patients)^{6–8,26,28–32} evaluated the incidence of PONV, indicating no difference between both groups (RR, 0.89 [95% CI, 0.66–1.20]; P = .44; I² = 46%) (Figure 5). The quality of the level of evidence was classified as low.

Figure 5.: Forest plot for methadone versus control for PONV. The table displays the study, mean, SD, sample size (total), difference in means with 95% CI, heterogeneity, overall effect, and P values. CI indicates confidence interval; df, degrees of freedom; I ², heterogeneity index; M–H, Mantel-Haenszel; PONV, postoperative nausea and vomiting; SD, standard deviation.

Regarding respiratory adverse effects, 6 studies (462 patients)^{6,7,9,28–30} analyzed this outcome, demonstrating no difference between the groups (RR, 1.24 [95% CI, 0.68–2.25]; P = .49; I² = 0%). Cardiovascular complications were evaluated in 4 studies (183 patients)^6–8,30 and showed no difference (RR, 1.18 [95% CI, 0.72–1.91]; P = .51; I² = 18%) between methadone and control groups. These outcomes were classified as a very low level of evidence and respective forest plots are shown in Supplemental Digital Content 6, Figure 3, https://links.lww.com/AA/C924. No studies described the occurrence of urinary retention.

Figure 6.: Forest plot for methadone versus control for patient satisfaction. The table displays the study, mean, SD, sample size (total), difference in means with 95% CI, heterogeneity, overall effect, and P values. CI indicates confidence interval; df, degrees of freedom; I ², heterogeneity index; IV, intravenous; PO, postoperatively; SD, standard deviation.

Patient satisfaction was assessed in 3 trials^6–8 and favored methadone in 24-, 48-, and 72-hour evaluations after surgery as a low-quality level of evidence (MD, 10.52 [95% CI, 4.44–16.61], P < .0007, I² = 80%, 293 patients; MD, 8.29 [95% CI, 4.99–11.60], P < .00001, I² = 36%, 291 patients; and MD, 6.75 [95% CI, 3.40–10.10], P < .00001, I² = 54%, 291 patients), respectively (Figure 6).

Small-Study Effect

The small-study effect was examined with funnel plot because there were insufficient studies for further testing. Funnel plots for pain at rest at 24, 48, and 72 hours showed symmetrical distribution, as well as pain at movement at 24 and 72 hours. Only pain at movement in the 48-hour analysis showed asymmetry, pointing to a small-study effect. Funnel plot for opioid consumption was symmetrical at PACU, 24–48 and 48–72 hours; with asymmetry detected in opioid consumption until 24 hours postoperatively (Supplemental Digital Content 7, Figure 4, https://links.lww.com/AA/C924).

DISCUSSION

This is the first meta-analysis to address intraoperative methadone use and postoperative pain management. Our primary outcomes were postoperative pain and opioid consumption. Both coprimary outcomes favored methadone in this joint-hypothesis testing. There were statistically significant lower pain scores at rest during the first postoperative 72 hours, favoring the methadone group. When a subgroup analysis was performed to assess the difference between groups in pain intensity at movement, pain score reduction showed a higher difference between methadone and other opioids. Evidence favoring methadone use for pain intensity at rest, particularly at 24 and 48 hours after surgery, surpassed the RIS and crossed the boundary for benefit at Trial Sequential Analysis. This effect suggests that conducting more trials to evaluate these effects is superfluous and that the minimal clinically important difference was achieved with methadone use.^16,21,22

Methadone plays a lasting role in pain control.⁵ Especially in the second postoperative day, adequate pain control at movement represents an important concern in the patient’s recovery because this period is considered the most painful due to onset of ambulation.^2,3,33 Thus, lower levels in pain at movement at this moment may represent more functional ambulation and recovery of gastrointestinal function. The quality of evidence was categorized as moderate according to GRADE, downgraded for indirectness due to a difference in cumulative doses of opioids in control groups, which could result in different pain scores.

The methadone group also evidenced significantly lower opioid consumption. This is particularly important, especially regarding the tendency for lower incidence of opioid-related complications, inversely related to opioid consumption.^34,35 The most interesting benefit may have come during the second postoperative day because patients should be encouraged to move. Pain usually decreases on the third postoperative day and the difference in opioid consumption also decreased. This evidence was also derived from studies in which the quality of the evidence was considered low, mainly because analgesia was not standardized in the postoperative period with an equivalent dose of opioids varying according to procedures and/or institutional protocols. Three^7,23,30 studies used patient-controlled analgesia (PCA), while the others used opioid on demand, so this outcome was downgraded for inconsistency.

No significant differences were found with respect to PONV, respiratory, or cardiovascular complications between groups. These results suggest that methadone use can reduce postoperative opioid consumption but does not significantly alter the incidence of these postoperative opioid-related side effects. This evidence was also derived from quality of the evidence considered as very low, mainly due to the problems with imprecision and the small number of participants assessed for these complications.

Methadone use significantly improved patient satisfaction. Overall patient satisfaction in this review was high, favoring the methadone group, representing an interesting intraoperative approach because an intraoperative opioid choice could significantly affect patient satisfaction. It is important to highlight that this effect may be secondary to more stable postoperative blood levels of methadone compared to other shorter-acting opioids, elevated levels of serotonin or norepinephrine, or possibly NMDA receptor antagonism. In addition, satisfaction assessments were performed with nonvalidated tools and questionnaires, and while this outcome is considered critical for reviewers, this finding was obtained from only 3 studies^6–8 and provided a moderate level of evidence.

The methadone favorable outcomes in postoperative opioid consumption and pain intensity can be explained by a combination of factors. Methadone has a longer elimination half-life (13–50 hours), which may lead to a longer analgesic effect, evidenced by lower pain scores, lower postoperative opioid consumption, and higher patient satisfaction until 72 hours postsurgery. In addition, this opioid, besides being a MOR agonist, also antagonizes glutamate by blocking the NMDA receptor and inhibits reuptake of serotonin and noradrenaline.^4,5,36 These other effects in pain transmission and modulation can enhance the analgesic potential of methadone when used during anesthesia, decreasing the effects of nociceptive stimulus of surgery trauma. This is especially important nowadays that Enhanced Recovery After Surgery (ERAS) protocols are being implemented. One of ERAS goals is to use analgesic drugs and techniques to reduce postoperative opioid consumption, and methadone can behave as an opioid with an opioid-sparing effect.^5,36,37

The trials included in this review were conducted in different countries and continents, expanding the population of study and included a wide range of procedures, improving the external applicability of the findings. Following the methodology established in the Cochrane Handbook for Systematic Reviews of Interventions minimized biases in this review process. The evidence of this review came from a detailed search process that included published and unpublished articles and imposed no restrictions with respect to language. Nonetheless, it is possible that potentially eligible studies published in nonindexed journals were not recovered, or there may be a time-lag bias (studies that had been completed but had not yet been published).

However, there were several limitations to this meta-analysis: (1) the consistency of some results may be impaired due to included studies in which the quality of evidence was classified as very low because of design flaws; (2) patients underwent different types of surgeries, which represent different nociceptive stimuli, representing inconsistency biases; (3) the cumulative dose of opioids in methadone and control groups showed a wide variation that may add problems of indirectness in pain intensity assessment; (4) other significant outcomes such as delay in anesthetic awakening and altering of QT interval in electrocardiogram^38,39 could not be evaluated in this meta-analysis due to lack of data; (5) most trials included in this meta-analysis had small sample sizes; (6) follow-up in the included studies ranged from 24 hours to 3 months, and the relatively short-term follow-up may underestimate the complication rate; (7) there were multiple analgesic and anesthetic approaches between studies; and (8) small-study effect could only be accessed with funnel plot, given the limited number of included studies, which can indicate possible publication bias.

Further investigation could include the use and safety of methadone for regional analgesia in pain management. Also, the QT interval should be actively monitored to guarantee the safe use of this opioid in future research. Long postoperative follow-up periods could be stipulated to evaluate the effect of methadone on chronic postoperative pain compared to other opioids. Finally, evaluating pharmacoeconomics involved in the choice of methadone for analgesia might be of interest. Further large, multicenter RCTs of high methodological quality are required to solidify the evidence and clarify the questions that have arisen from this meta-analysis.

DISCLOSURES

Name: Felipe C. Machado, MD, PhD.

Contribution: This author helped design the review, organize the data recovery, apply the eligible criteria, insert the data in Review Manager, perform the statistical analysis, interpret and analyze the data, and write and elaborate the process of meta-analysis.

Name: Joaquim E. Vieira, PhD.

Contribution: This author helped interpret and analyze the data and helped with statistical inferences.

Name: Flávia A. de Orange, PhD.

Contribution: This author helped interpret and analyze the data, helped with statistical inferences, and helped coordinate the process of elaborating this review.

Name: Hazem A. Ashmawi, PhD.

Contribution: This author helped apply the eligible criteria, insert data in Review Manager, and interpret and analyze the data; helped with statistical inferences; and helped coordinate the process of elaborating this review.

This manuscript was handled by: Honorio T. Benzon, MD.