Intraoperative Methadone in Next-day Discharge Outpatient Surgery: A Randomized, Double-blinded, Dose-finding Pilot Study

A single-center, randomized, double-blind, parallel-group, dose-escalation, dose-finding feasibility and pilot study^31–33  was performed. Feasibility assessed enrollment, conduct of anesthesia, assessments and follow-up. Pilot objectives were dose-finding and preliminary, exploratory assessment of outcomes. The protocol and informed consent document were approved by the Washington University in St. Louis Institutional Review Board (St. Louis, Missouri), and the research was registered at clinicaltrials.gov, November 18, 2014 (NCT02300077, Investigator: Dr. Helga Komen), before patient enrollment. All patients provided written informed consent. Potential participants were identified by investigators and research staff from operating room schedules. Inclusion criteria were age 18 to 65 yr and undergoing elective next-day discharge outpatient surgery (intended hospital stay lasting one but less than two midnights) under general anesthesia. Exclusion criteria included a history of liver or kidney disease, pregnant or nursing females, and potentially opioid-tolerant patients (i.e., daily methadone, more than 20 mg/day oxycodone or hydrocodone, or fentanyl transdermal patch). Patients were enrolled from December 2014 to October 2016, and from August 2017 to January 2018 (after additional funding).

Before surgery, patients were asked to complete a short questionnaire regarding the past 7 days average pain intensity (0 to 10 scale) and pain interference (0 to 4 scale), as well as overall physical-mental health (based on National Institutes of Health [Bethesda, Maryland] patient-reported outcomes measurement information system [PROMIS]-29 Profile v2.0). Demographic data were recorded. On the day of surgery, participants were also asked preoperatively (1) their self-reported baseline pain score (0 to 10), (2) what level of pain they expected after surgery, and (3) what is an acceptable level of pain after surgery?

Patients were randomized by blocks of three, in a 2:1 ratio (methadone:control), using a published randomization table, to receive either intraoperative single-dose methadone or usual practice with intraoperative short-duration opioid (choice of fentanyl, sufentanil, hydromorphone, or morphine, with opioid choice and dosing ad libitum at practitioners’ discretion). The initial cohort of 20 methadone patients received 0.1 mg/kg ideal body weight, and successive cohorts received 0.2, 0.25 and 0.3 mg/kg ideal body weight (dose rounded to the nearest mg). The highest methadone dose was 20 mg. Ideal body weight (kg) was calculated as (2.3 × No. inches height > 60) + 50 (males) or 45.5 (females). Patients received methadone as a single dose, in the operating room at, or up to 5 to 10 min before, induction.

Anesthesia providers were given a numbered sealed envelope containing the randomization assignment, which was prepared by an otherwise uninvolved individual. Investigators, patients, and research staff evaluating the patients were blinded to drug group assignment. Anesthesia providers and postanesthesia care unit (PACU) nurses were unblinded. PACU staff were unaware of the study objectives, design, and outcomes. Anesthesia and surgical care were not changed for study purposes, except for the following: (1) patients in the methadone groups received methadone as their intraoperative opioid rather than leaving the choice and doses of intraoperative opioid to the provider; (2) practitioners were asked to use no additional nonmethadone short-duration opioid during the procedure in subjects receiving methadone; (3) all patients could receive additional (short-duration) opioid for pain as needed during or after wound closure and emergence to ensure their comfort, and (4) no controls (short-duration opioid) received methadone.

Patient premedication was at the discretion of the anesthesia care team. Anesthesia was induced with propofol or etomidate, the designated opioid (single-dose methadone or unrestricted short-duration opioids), and muscle relaxant was given at anesthesia providers’ discretion. Anesthetic maintenance was with sevoflurane, desflurane or propofol at provider’s discretion. Less commonly and more variably used drugs (nitrous oxide, ketamine, dexmedetomidine) were generally avoided, to reduce variability. Antiemetic prophylaxis (ondansetron, 4 mg IV) was given per usual practice to each patient toward the end of surgery. Other prophylactic antiemetic use was not recorded.

Postoperative care was per institutional practice and not altered for study purposes. Patients in the PACU could receive short-duration opioids (fentanyl or hydromorphone), with dosing (for mild, moderate, or severe pain based on patient reports [0 to 10 verbal scale] of 1 to 3, 4 to 6, or 7 to 10, respectively), per institutional standing orders. Additional antiemetic (ondansetron or diphenhydramine) was given if needed. Additional oral analgesics (oxycodone, hydrocodone, or acetaminophen) after discharge from the PACU and until discharge from the hospital, if needed, were prescribed by the surgical team. Postdischarge opioids were prescribed by the surgical team, and included oxycodone (5 mg), oxycodone (5 mg)/acetaminophen (325 mg), and hydrocodone (5 mg)/acetaminophen (325 mg).

Patients were assessed postoperatively for pain, sedation, and adverse events after PACU admission, every 15 min for the first hour, hourly for the next 4 h, and at bedtime and before discharge. Assessments were by a trained member of the research team, blinded to randomization allocation. A standard protocol was used for assessing pain intensity (at rest, with coughing and deep breathing, and with activity) using a 0- to 100-mm visual analog scale. Sedation (Modified Observer’s Assessment of Alertness/Sedation, 0 to 5) was recorded concurrently with pain assessments. Adverse events recorded were respiratory depression (respiratory rate less than 8/min), oxygen desaturation (less than 90% for more than 1 min; less than 85% for more than 30 s), excessive sedation (Modified Observer’s Assessment of Alertness/Sedation, 0 to 2), pain/sedation mismatch (Modified Observer’s Assessment of Alertness/Sedation, 0 to 2 and pain score greater than 5),³⁴  and reintubation. Antiemetic use was recorded from the electronic medical record. Patients reported opioid side effects before discharge using the Opioid-related Symptom Distress Scale.³⁵  The Opioid-related Symptom Distress Scale uses 4-point Likert scales to characterize 12 opioid side effects (nausea, vomiting, constipation, difficulty urinating, difficulty concentrating, drowsiness, dizziness, confusion, fatigue, itching, dry mouth, and headache) according to frequency, severity, and bothersomeness.

Patients were followed for 30 days after surgery, using a paper diary to record their daily average pain (at rest, with coughing and with activity using a 0 to 100 visual analog scale), and daily opioid analgesic use. On postoperative days 7, 14 and 30 they also recorded pain interference with four activities of daily living (0 to 4 scale) using questions based on the patient-reported outcomes measurement information system (PROMIS) Pain Behavior and Pain Interference item banks, and opioid side effects (Opioid-related Symptom Distress Scale).

The feasibility and dose-finding protocol involved interim assessments of each methadone dose cohort after it was completed. Methadone group sample size (approximately 20 per group) for this feasibility and pilot investigation was based on previous studies,³⁶  and 2:1 randomization (methadone:control) was used to ensure adequate numbers of controls, anticipating at least two methadone doses. Methadone dose-escalation ceased when subjects required a clinically meaningful minimum of opioid in the PACU and postoperative day 1, based on group means, and without an increase in adverse events.

The main feasibility outcome was use of single-dose methadone for outpatient next-day discharge surgery. The primary pilot outcomes were methadone dose and in-hospital (intraoperative and postoperative) opioid utilization. Secondary outcomes were in-hospital postoperative pain scores, postoperative 30-day opioid consumption and pain scores, in-hospital and postdischarge opioid side effects, leftover unused take-home opioids, and pain interference after surgery. All opioid use was converted to IV milligram morphine equivalents (MME) for analysis.³⁷  For conversion from IV to oral morphine equivalents, multiply the results presented by three.

Contemporary perspectives and consensus statements,^32,33,38  including the 2010 Consolidated Standards of Reporting Trials (CONSORT) extension to pilot and feasibility studies,³⁹  state that formal hypothesis testing for clinical efficacy or effectiveness outcomes is not recommended for pilot studies, because they are underpowered to do so and lack statistical power to detect significant differences. Therefore, formal hypothesis testing was not an aim of the investigation. Limited exploratory statistical analysis was, however, performed for the primary outcomes. Groups were compared using the Kruskal-Wallis test followed by Dunn’s method for individual comparisons. Leftover home opioid use was compared using the chi-square test. Statistical significance was assigned at P < 0.05. Analyses were performed using SigmaPlot 14.5 (Inpixon/Systat Software, USA). Normally distributed data are presented as mean ± SD, and nonnormally distributed data are presented as the median and 25th and 75th quartiles.

The pilot data were evaluated visually and numerically for possible signals. A signal is “information that arises from one or multiple sources (including observations and experiments), that suggests a new potentially causal association, or a new aspect of a known association, between an intervention and an event or set of related events, either adverse or beneficial, that is judged to be of sufficient likelihood to justify further action to verify.”⁴⁰ 

A total of 402 patients were assessed for eligibility; 148 consented, 139 were randomized, and 129 received the allocated intervention. Of these, five were excluded from analysis (fig. 1). The remaining 124 were included in the final data analysis of the in-hospital portion of the study. Postoperative 30-d follow-up diaries were returned by 88 patients (table 1), and included in the final follow-up data analysis. The dose escalation trial ended when the planned number of patients in each dose cohort was enrolled and analysis confirmed an effective methadone dose.

Demographic data and anesthetic characteristics were comparable in the five groups with respect to age, actual body weight, ideal body weight, American Society of Anesthesiologists (ASA) Physical Status, anesthesia duration, and types of surgeries (table 1). Across all groups, actual body weight (average ± SD) was 87 ± 24 kg, ideal body weight was 57 ± 7 kg, and the majority of participants were ASA Physical Status 2 and female due to the nature of the commonly performed next-day discharge surgical procedures (hysterectomy). Among all participants, anesthesia duration was 172 ± 53 min, intraoperative fluid administration was 1,910 ± 620 ml, estimated blood loss was 90 ± 90 ml, and any differences between treatment groups were negligible, if present. As expected, there was less overall interindividual variability in ideal body weight (12% coefficient of variation) than actual body weight (27%). On the baseline assessment survey, among all participants, self-reported pain intensity over the past 7 days (0 to 10 scale) was 2 ± 3. On the day of surgery, self-reported baseline pain score (0 to 10) was 1 ± 2, the level of pain expected after surgery (0 to 10) was 6 ± 3, and the acceptable level of pain after surgery (0 to 10) was 5 ± 2. Differences, if any, between treatment groups were negligible.

Average intraoperative methadone doses were 6, 11, 14, and 17 mg in the 0.1, 0.2, 0.25, and 0.3 mg/kg ideal body weight groups, respectively (table 2). The highest methadone dose was 20 mg. There was little within-group variation in actual methadone dose, because of dosing to ideal body weight. Calculated methadone doses averaged 0.07, 0.13, 0.18, and 0.21 mg/kg actual body weight in the 0.1, 0.2, 0.25, and 0.3 mg/kg ideal body weight groups, respectively (table 2). In contrast to dosing to ideal body weight, there was far greater interindividual variability when methadone dosing was calculated relative to actual body weight. Controls all received fentanyl or fentanyl plus hydromorphone. In controls, average doses were 260 ± 80 µg fentanyl (26 ± 8 IV MME, n = 40) and 1.0 ± 0.7 mg hydromorphone (7 ± 5 IV MME, n = 25). In subjects receiving methadone, practitioners were asked to use no additional opioid during the procedure, but could use additional opioid for pain as needed on emergence (during or after wound closure). Only a small number of subjects receiving methadone (0 to 3 per dose group) also received small amounts of fentanyl (100 to 150 µg total) or hydromorphone (0.75 µg total), showing excellent protocol compliance.

Study feasibility evaluated whether single-dose methadone could be assessed in a research protocol, and could be used for outpatient next-day discharge surgery. Operationally, patients were willing to enroll and be randomized, anesthesiology practitioners were willing to execute the protocol with a high degree of compliance, and completion rates of in-hospital patient follow-up, patient assessments and patient-reported outcome surveys were high. Completion and return of 30-day take-home patient paper diaries overall were less, at 72%. Clinically, anesthesia with single-dose intraoperative methadone together with propofol or a volatile anesthetic for unconsciousness was feasible, effective, and readily accepted by anesthesia providers.

Dose-escalation was guided by PACU opioid use, which was assessed after each methadone dose cohort to determine whether to stop or escalate further. PACU opioid use (almost exclusively hydromorphone) was (in IV MME) 5.6 ± 6.9 mg in controls and 5.9 ± 5.0 mg in patients receiving methadone 0.1 mg/kg, so the methadone dose was escalated. Higher dose cohorts were sequentially studied. PACU opioid use was 5.1 ± 7.6, 4.8 ± 5.5, and 4.1 ± 5.0 IV MME in the 0.2 mg/kg, 0.25 mg/kg, and 0.3 mg/kg ideal body weight methadone groups, respectively. The intraoperative methadone dose was not increased further, and the study was stopped.

Opioid use after surgery (PACU, post-PACU in hospital) is shown in figure 2. There was considerable interindividual variability in opioid use. Visual inspection suggested a signal of less opioid use on the day of surgery (fig. 2, A to C), day after surgery (fig. 2D), and less total in-hospital opioid use (fig. 2E), at the higher methadone doses. Exploratory statistical analysis found significantly less total in-hospital opioid use (fig. 2E) in patients receiving higher doses of methadone compared to controls (short-duration opioid; P < 0.001 in all methadone groups).

Assessments of patient recovery in the PACU and for 4 h after surgery included alertness or sedation (Modified Observer’s Assessment of Alertness/Sedation score, 0 to 5), recovery (Aldrete score, 0 to 10), pain (visual analog scale, 0 to 100), discharge time, antiemetic use, and adverse events. PACU alertness scores increased over the first hour (fig. 3, medians). Considerable interindividual variability in alertness scores is apparent from the SD, shown along with the means (Supplemental Figure 1, means, https://links.lww.com/ALN/D173). There was a possible signal of slightly lower median scores (more PACU sedation) for the first hour in patients receiving 0.3 mg/kg ideal body weight methadone. Aldrete scores determined 1 h after PACU admission (median, 25th to 75th quartiles) were 10 (9 to 10), 9 (9 to 10), 10 (10 to 10), 10 (9 to 10), and 8 (8 to 10) in the control, and 0.1, 0.2, 0.25 and 0.3 mg/kg methadone groups, respectively.

Pain at rest and with coughing or deep breathing were assessed in the PACU and for 4 h after surgery, at bedtime, and the next day at hospital discharge (fig. 4). Pain with activity was also assessed at bedtime on the day of surgery and at hospital discharge (fig. 4). An evident signal of lower median pain scores in the patients receiving methadone was seen, particularly at the higher intraoperative methadone doses.

Patients were generally ready for PACU discharge after 1 to 1.5 h, with actual discharge after 1.5 to 2.5 h (table 3). Antiemetic administration, on the day of surgery and the next day, is shown in table 4. Antiemetics were commonly administered, particularly on the day of surgery after PACU discharge, but also on the day after surgery, with no obvious signals regarding greater use in any group. Adverse events in the PACU and for 4 h after surgery were very rare (table 3). A single episode of both oxygen desaturation to less than 90% (but not less than 85%) and respiration less than 8 while asleep was observed, shortly after admission to the PACU, and which resolved spontaneously, in a patient who received 0.3 mg/kg methadone.

Paper diaries to record postdischarge oral opioid use, pain, and recovery parameters for 30 days were returned by 72% of patients. The most common discharge opioids prescribed were oxycodone 5 mg (n = 28), oxycodone 5 mg with acetaminophen 325 mg (n = 75), and hydrocodone 7.5 mg with acetaminophen 325 mg (n = 4). Number of pills dispensed per patient averaged 37 ± 12 (range, 15 to 60), median 30 (interquartile range, 30 to 40). Opioid prescribing was comparable between opioid treatment groups.

There was considerable interindividual variability in opioid use. Some patients used no postoperative opioids while others used them for the full 30-day monitoring period. Daily opioid use and pain scores for individual patients are shown in Supplemental Figure 2 (https://links.lww.com/ALN/D174). Daily opioid use for each opioid group is shown in Supplemental Figure 3 (https://links.lww.com/ALN/D175), for both pill counts and MME. Overall median duration of postoperative opioid use was 3 days. Duration of use (median and interquartile range) was 3 (0 to 7) days in controls and 3 (1 to 11), 3 (0 to 8), 3 (1 to 10), and 2 (0 to 5) days after 0.1, 0.2, 0.25, and 0.3 mg/kg methadone, respectively. Total 30-day home opioid use after hospital discharge in MME is shown in figure 5A, and in opioid pills used in Supplemental Figure 4A (https://links.lww.com/ALN/D176). There was a signal suggesting less take-home opioid use in the patients receiving methadone at the higher doses. This finding was also observed in Supplemental Figure 3, A and B (https://links.lww.com/ALN/D175).

We evaluated the disposition of take-home opioids that were used and unused for 30 days. Among the 86 patients who reported their postoperative opioid use, 3,375 pills were prescribed, only 766 (23%) were used, and 2,610 (77%) were unused. The number and fraction used informs on just that—use (possibly reflecting pain); however, not all prescriptions were the same size, and the risk to patients and to society is the absolute number of pills left over. The fraction of dispensed pills used and unused by each patient was calculated, and the median fractions used (fig. 5B), and leftover unused (Supplemental Figure 4B, https://links.lww.com/ALN/D176) are shown. The sum total of pills used and unused in each group, expressed as a fraction of the total dispensed, is in Supplemental Figure 4C (https://links.lww.com/ALN/D176). This was 26% (264 pills) used in controls, and 42% (151 pills), 19% (164 pills), 20% (149 pills), and 10% (38 pills) with methadone 0.1, 0.2, 0.25, and 0.3 mg/kg, respectively. Unused pills were 74% (741 pills) in controls, and 58% (209 pills), 81% (701 pills), 80% (607 pills), and 90% (352 pills) with methadone 0.1, 0.2, 0.25, and 0.3 mg/kg, respectively. Figure 5 and Supplemental Figure 4 (https://links.lww.com/ALN/D176) indicate a signal of fewer take-home opioids used in patients receiving methadone, particularly at the higher doses, and more leftover pills in these patients. Exploratory statistical analysis (Supplemental Figure 4C, https://links.lww.com/ALN/D176) found a significantly different fraction of opioids used in patients receiving methadone compared to short-duration opioids (P < 0.01 in all groups).

Several indicators of postoperative pain, recovery, and potential opioid-related side effects were assessed daily or weekly as patient-reported outcomes. Patients rated their pain daily for 30 days, at rest, with coughing and deep breathing, and with activity. Summed 30-day pain scores are shown in figure 6. Pain scores were higher with greater activity (fig. 6, B and C). There was a possible signal of less pain with activity in the higher-dose methadone groups compared with controls. Individual pain scores with coughing and with activity are shown in Supplemental Figure 5 (https://links.lww.com/ALN/D177). Several dimensions of pain interference with daily activities were queried weekly, and two are shown in figure 7. A possible signal of less interference with some activities was observed in the higher-dose methadone groups compared with controls, on postoperative day 7. Median interference scores after day 7 were generally 0. Opioid-related side effects, likely reflecting side effects related to home oral opioid use, were measured using the opioid-related symptom distress scale (fig. 8). Side effect scores decreased over time, commensurate with diminishing opioid use.

Outpatient surgery is increasingly common. Approximately 70% of procedures in the United States (more than 50 million annually) are performed in an outpatient setting.^29,30  This feasibility and pilot investigation aimed to determine the feasibility of single-dose intraoperative methadone for next-day discharge outpatient surgery, identify an optimally analgesic and well-tolerated dose, and explore whether intraoperative anesthesia with single-dose methadone would result in less postoperative opioid use compared with conventional short-duration opioids. It was conducted in preparation for a possible larger-scale randomized trial of intraoperative methadone for next-day discharge outpatient surgery.

Both the clinical protocol and the research protocol were feasible. Anesthesia with single-dose intraoperative methadone together with a sedative hypnotic was effective and readily accepted by anesthesia providers. There was excellent practitioner compliance with the research protocol, and completion of in-hospital patient follow-up assessments and patient-reported outcome was high. Completion and return of 30-day take-home paper diaries overall were less, 72% overall. However, this resembled the anticipated dropout rate of 25%.

We aimed to determine a single intraoperative dose of methadone that minimized pain and PACU opioid requirement, without untoward side effects. The goal was not to increase the dose further, in order to find a maximally tolerable dose, as that could have caused unwanted side effects. Methadone dose-escalation targeted an anticipated dose-dependent lessening for PACU opioid. Based on monitoring of serial group mean PACU opioid doses, this was not observed, nor were adverse events observed, so methadone doses were sequentially escalated. Once reaching 0.3 mg/kg ideal body weight (17 ± 2 mg), the dose was not escalated further, to not exceed that (20 mg) previously reported in inpatients,^24,25  and the study was stopped.

Aggregate pilot data suggest an optimal methadone dose for further evaluation. Regarding therapeutic effects, visual and exploratory statistical analysis found, at the higher doses of methadone (0.25, 0.3 mg/kg ideal body weight) and compared with short-duration opioid, numerically less opioid use on the day of surgery, day after surgery, total in-hospital use, and less 30-day take-home opioid use. Similarly, at the higher methadone doses and compared with short-duration opioid, there was a signal of less pain in the hospital and after discharge. Compared with short-duration opioids, methadone was not associated with more adverse effects (nausea, emesis, respiratory depression) or delayed PACU discharge. Regarding side effects, there was a possible signal of sedation for the first few minutes in the PACU at the highest methadone dose (0.3 mg/kg ideal body weight). Conceptually, because elimination clearance determines duration of methadone analgesia, it is advantageous to use the highest possible effective and safe bolus dose to maximize duration.²¹  Based on all these considerations, we conclude that for next-day discharge outpatient surgery in adults undergoing operations comparable to those herein, the intraoperative methadone dose appearing to best combine opioid sparing, analgesia, minimal adverse events, and advantages compared with short-duration opioids is 0.25 mg/kg ideal body weight. Study patients at this dose received 14 ± 2 mg (median, 14; 25th to 75th percentiles, 13 to 16 mg) intraoperative methadone, and a median 2.0 (0 to 6.7) IV MME in the PACU. Because ideal body weight range is narrow, so too was the range of methadone doses in each cohort. Therefore, rather than dosing to body weight (actual or ideal, vide infra), for simplicity (and safety), comparable patient populations undergoing similar procedures might receive a single 15-mg induction dose of methadone. This dose is conservative, and is also now our routine in both research and clinical practice. As a conservative dose, it may be supplemented with additional PACU methadone (our research and clinical practice) if needed.

Obesity can complicate weight-based drug dosing.⁴¹  Conventional intraoperative short-duration opioids (e.g., fentanyl, sufentanil, remifentanil) are often dosed relative to body weight, while methadone has been dosed variously in absolute amounts or weight-adjusted.^22–26,42  Practitioners may be confused as to whether weight-based dosing recommendations refer to actual weight or ideal body weight. Drugs with recommended doses based on ideal body weight, but dosed to actual body weight, can result in an unsafe overdose in obese patients. Therefore, absent readily available ideal body weight calculators, methadone 0.25 mg/kg ideal body weight may be more safely given as a 15-mg fixed dose.

The optimal methadone dose needed by next-day discharge patients in this investigation (median, 14 to 18 mg) is only somewhat less than the 20-mg dose first reported for inpatients.^24,25  This finding is consistent with observations that many seemingly minor- to medium-level procedures do cause substantial pain, and laparoscopic surgery may not cause less pain than open surgery.^43–45 

If methadone dosing is too low, and concentrations are below the minimal effective concentration, clinical benefit will not ensue, even with slow methadone elimination. Patients who received only 0.1 mg/kg ideal body weight methadone, and thus also much less intraoperative opioid compared with controls, showed signals of using more PACU, postoperative and take-home opioid, and greater postoperative pain. This result shows the need for sufficient analgesia, the potential consequence of undertreating intraoperative pain, and the need for sufficient methadone dosing. Insufficiently treated postoperative pain can impair and prolong recovery, delay discharge after surgery, reduce mobilization from bed and ambulation, impair postoperative rehabilitation,⁶  and has been associated with greater postoperative surgical complications, surgical site and urinary tract infections, postdischarge emergency department visits and readmission,^46,47  as well as chronic postoperative pain.^7–11 

One noteworthy observation from this small pilot study was that patients who received intraoperative methadone, at the higher doses, used numerically less postdischarge take-home opioid, which may have been related to less pain. If one goal of anesthesiology is to devise regimens, optimize drug selection, and deliver care which provides optimal intraoperative and immediate postoperative analgesia, but ideally also confers longer-lasting benefit,²  then use of intraoperative methadone may facilitate that goal.^26,27 

A second corollary was the substantial interindividual variability in pain and opioid use. Variability appeared less, however, in patients who received higher methadone doses. If a goal is to identify more uniform postoperative opioid dosing regimens, then less variability with methadone may facilitate adequate responses among a greater fraction of patients.

Another corollary was the large amount of unused postoperative take-home opioid. Among patients who recorded postoperative opioid use, 77% of dispensed pills were unused. This mirrors the frequent observation that only a fraction of postdischarge opioids is used. We previously reported that across three different surgical specialties, less than 25% of postdischarge opioids prescribed were used, and requested opioid refill rates were only 3%.¹⁷  Overprescribing of postdischarge take-home oral opioids by surgeons is common and a public health problem.^18,19,48  This creates an “opioid pool” that constitutes potential harms.¹⁸  Reduced postoperative opioid overprescribing and adequate pain relief are both laudatory and necessary goals,^14,49  and are not necessarily mutually exclusive. In this study, across each opioid group, the fraction of unused opioids was 74% in controls, and 81%, 80% and 90% with 0.2, 0.25 and 0.3 mg/kg ideal body weight methadone, respectively. If intraoperative methadone deceases postoperative pain and opioid use, then postdischarge opioid prescriptions can be diminished, along with the potential for leftover opioids.

Opioid-sparing effects of methadone in this pilot study of next-day discharge surgery are similar to those of a dose-escalation pilot in same-day discharge surgery.²⁶  In that investigation, compared with conventional short-duration opioids, patients receiving IV methadone (0.15 mg/kg ideal body weight; median, 9 mg) needed less intraoperative and PACU opioid, and had less 30-day pain and opioid use. For simplicity, our routine induction dose in both research and clinical practice for same-day discharge patients is 10 mg methadone, supplemented with additional methadone in the PACU if needed.

There are potential limitations of this investigation. This was a feasibility and pilot study, with small numbers of patients, and not all patients returned the 30-day diaries. A predominance of women was enrolled, reflecting the surgical population. A larger comparative effectiveness trial would be needed to fully evaluate all of the primary and secondary outcomes, adverse events, and enhance generalizability.

In summary, this feasibility and pilot investigation found that anesthesia for next-day discharge outpatient surgery using single-dose intraoperative methadone (median, 14 mg), compared with conventional short-duration opioids, was feasible, and resulted in numerically less opioid use without greater side effects.

Supported by grants from the National Institutes of Health (Bethesda, Maryland; R01 DA042985 to Dr. Kharasch), Barnes-Jewish Hospital Foundation (St. Louis, Missouri; 7957-77 to Dr. Komen), and the Washington University in St. Louis Department of Anesthesiology, Division of Clinical and Translational Research (St. Louis, Missouri; to Dr. Komen).

The authors declare no competing interests.

Full protocol available at: evan.kharasch@duke.edu. Raw data available at: evan.kharasch@duke.edu.

Supplemental Figure 1: Postoperative alertness and sedation, https://links.lww.com/ALN/D173

Supplemental Figure 2: Daily home opioid use and pain after hospital discharge, https://links.lww.com/ALN/D174

Supplemental Figure 3: Daily home opioid use after hospital discharge, https://links.lww.com/ALN/D175

Supplemental Figure 4: Home opioid use after hospital discharge, https://links.lww.com/ALN/D176

Supplemental Figure 5: Daily pain scores for 30d postoperatively, https://links.lww.com/ALN/D177