Intraoperative Methadone and Postoperative Anesthesia Care Unit Outcomes

Methadone has been used in perioperative anesthesia practice for at least four decades, and there is increasing interest in perioperative methadone use and even more so recently in both inpatients and outpatients.  Intraoperative methadone, compared with short-duration opioids, results in less postoperative opioid use, less postoperative pain, and greater patient satisfaction. There may also be long-term benefits to patients after discharge. The clinical advantage of methadone is the long elimination half-life, which confers long-duration analgesia and opioid sparing but which may engender theoretical concerns about adverse effects. Studies to date have found no difference between anesthesia with methadone versus short-duration opioids in the frequency of opioid-related adverse events. Nonetheless, most of these studies have been small and do not inform broadly on the safety of perioperative methadone. The need for such evidence, particularly with larger studies, has recently been highlighted. 

To better understand methadone postoperative clinical outcomes, opioid-related adverse events and safety, a retrospective analysis of an institutional medical records database was performed. The goal was to determine whether intravenous intraoperative methadone dose influenced (1) length of postanesthesia care unit (PACU) stay, (2) postoperative nausea and vomiting, and (3) postoperative PACU respiratory depression. The study was approved by the Washington University in St. Louis Institutional Review Board (St. Louis, Missouri). The outcomes were (1) PACU phase 1 length of stay, (2) PACU antiemetic administration as a surrogate indicator of postoperative nausea and vomiting, and (3) PACU naloxone administration as a surrogate indicator of respiratory depression. Data were obtained from 99,244 records in a MetaVision perioperative database of adult inpatients undergoing noncardiac surgery between January 2013 and November 2015. Methadone doses received were 0 mg (97,263), 1 to 5 mg (n = 68), 6 to 10 mg (n = 1,122), 11 to 15 mg (n = 187), 16 to 20 mg (n = 557), 21 to 29 mg (n = 9), and greater than or equal to 30 mg (n = 38); median and average doses were 10 and 13 mg, respectively. Data were analyzed by nonparametric one-way analysis of variance (Kruskal–Wallis) with post hoc multiple comparisons for multiple groups, and rank-sum or chi-square tests for two groups (SigmaPlot 15.1, Inpixon, USA). Significance was assigned at P < 0.05.

The relationship between phase 1 PACU length of stay and methadone dose is shown in figure 1A. There were no significant differences in length of stay between individual methadone dose groups and those who did not receive methadone, including the absence of a methadone dose–effect relationship. However, in aggregate analysis combining all methadone dose groups, patients who received methadone had longer PACU stays (median, 2.2 h; 25th and 75th percentiles, 1.6 and 2.8) than those who did not (median, 1.8 h; 25th and 75th percentiles, 1.2 and 2.5; P < 0.001). The relationship between phase 1 PACU antiemetic drug administration and methadone dose is shown in figure 1B. The median antiemetic doses and 25th and 75th percentiles were zero in all groups. The means are shown in the figure. There were no significant differences between individual methadone dose groups and those who did not receive methadone, including the absence of a methadone dose–effect relationship, in the number of antiemetic drug doses administered in the PACU. The antiemetics were ondansetron (58%), diphenhydramine (19%), prochlorperazine (13%), dexamethasone (6%), haloperidol (2%), or droperidol (2%). In an aggregate analysis combining all methadone dose groups, patients who received methadone received statistically more PACU antiemetics than those who did not (P < 0.001), although the median and 25th and 75th percentile doses were zero in the two groups. There was no significant difference in the fraction of patients receiving PACU naloxone between those who did or did not receive intraoperative methadone (fig. 1C). None of the 1,981 patients who received intraoperative methadone received naloxone, while 112 (0.12%) patients not receiving methadone did receive naloxone in the PACU.

These results show that intraoperative methadone administration was not associated with dose-dependent longer PACU length stay or greater PACU postoperative nausea and vomiting (as evidenced by antiemetic administration) or greater postoperative PACU respiratory depression (as evidenced by naloxone administration). The incidence of naloxone use in nonmethadone patients (0.12%) was similar to the reported incidence of postoperative opioid overdose in patients undergoing major elective inpatient surgery (0.1%). It is notable that no patient who received intraoperative methadone suffered a PACU respiratory event needing naloxone. While the MetaVision database does not inform on post-PACU events, the hours immediately after surgery is the time of greatest risk for opioid-induced respiratory depression. Limitations of this analysis are, by the nature of the database used, the retrospective approach, the convenience sample with some small dose groups, absence of an a priori sample size justification, some use of surrogate markers, and unknown patient-level case mix, risk factors for postoperative nausea and vomiting, and use of preoperative opioids. Additionally, the PACU-specific and deidentified database did not permit a risk-adjusted patient and surgical factor analysis.

The absence of a univariate association between intraoperative methadone administration and both postoperative respiratory depression and postoperative nausea and vomiting is consistent with previous clinical trials and summative reports. The current retrospective analysis adds some data, from records of 99,244 patients, of whom 1,981 received methadone, to previous evidence showing no greater risk of intraoperative methadone compared with short-duration opioids and which also showed that methadone is both pain and opioid sparing.  A fuller analysis, with even greater patient numbers and including risk adjustment for patient and surgical factors, would be informative.