ÌÇÐÄvlog

Key Points

QuestionÌý Does limiting opioid prescription to patients who request it (ie, opt-in program) decrease the quantity of opioids prescribed and consumed compared with patients given a routine prescription after outpatient cervical endocrine surgery?

FindingsÌý In this randomized clinical trial of 102 adults undergoing thyroidectomy or parathyroidectomy, patients randomized to the opt-in program were prescribed less opioids without reporting worse postoperative pain scores or health-related quality of life. No patients who opted out required a rescue opioid prescription.

MeaningÌý In this randomized clinical trial, an opt-in strategy for opioid prescriptions after endocrine surgery was noninferior to use of routine prescriptions with respect to postoperative pain and reduced overprescription of opioids.

ImportanceÌý Historically, opioid pain medications have been overprescribed following thyroid and parathyroid surgery. Many narcotic prescriptions are incompletely consumed, creating waste and opportunities for abuse.

ObjectiveÌý To determine whether limiting opioid prescriptions after outpatient thyroid and parathyroid surgery to patients who opt in to narcotic treatment reduces opioid consumption without increasing postoperative pain compared with usual care (routine narcotic prescriptions).

Design, Setting, and ParticipantsÌý A randomized clinical trial of Postoperative Opt-In Narcotic Treatment (POINT) or routine narcotic prescription (control) was conducted at a single tertiary referral center from June 1 to December 30, 2020. A total of 180 adults undergoing ambulatory cervical endocrine surgery, excluding patients currently receiving opioids, were assessed for eligibility. POINT patients received perioperative pain management counseling and were prescribed opioids only on patient request. Patients reported pain scores (0-10) and medication use through 7 daily postoperative surveys. Logistic regression was used to determine factors associated with opioid consumption.

InterventionsÌý Patients in the POINT group were able to opt in or out of receiving prescriptions for opioid pain medication on discharge. Control patients received routine opioid prescriptions on discharge.

Main Outcomes and MeasuresÌý Daily peak pain score through postoperative day 7 was the primary outcome. Noninferiority was defined as a difference less than 2 on an 11-point numeric rating scale from 0 to 10. Analysis was conducted on the evaluable population.

ResultsÌý Of the 180 patients assessed for eligibility, the final study cohort comprised 102 patients: 48 randomized to POINT and 54 to control. Of these, 79 patients (77.5%) were women and median age was 52 (interquartile range, 43-62) years. A total of 550 opioid tablets were prescribed to the control group, and 230 tablets were prescribed to the POINT group, in which 23 patients (47.9%) opted in for an opioid prescription. None who opted out subsequently required rescue opioids. In the first postoperative week, 17 POINT patients (35.4% of survey responders in the POINT group) reported consuming opioids compared with 27 (50.0%) control patients (P = .16). Median peak outpatient pain scores were 6 (interquartile range, 4-8) in the control group vs 6 (interquartile range, 5-7) in the POINT group (P = .71). In multivariate analysis, patients with a history of narcotic use were 7.5 times more likely to opt in (95% CI, 1.61-50.11; P = .02) and 4.8 times more likely to consume opioids (95% CI, 1.04-1.52; P = .01). Higher body mass index (odds ratio, 1.11; 95% CI, 1.01-1.23; P = .03) and highest inpatient postoperative pain score (odds ratio, 1.24; 95% CI, 1.04-1.52; P = .02) were also associated with opioid consumption.

Conclusions and RelevanceÌý In this trial, an opt-in strategy for postoperative narcotics reduced opioid prescription without increasing pain after cervical endocrine surgery.

Trial RegistrationÌý ClinicalTrials.gov Identifier:

The opioid epidemic has increased scrutiny on the surgeon’s contribution to the overprescription of narcotic pain medications.^1-3 Postoperative pain following cervical endocrine operations has generally been overestimated, and patients are prescribed far more opioids than they consume.^4-7 Not only is overprescribing wasteful, it also places patients and their families at risk for addiction and opioid abuse.⁸ Studies have reported that 5% to 10% of patients prescribed postoperative narcotics had persistent opioid use beyond the perioperative period.^9-13 In a recent prospective study of 216 patients who underwent outpatient cervical endocrine surgery, Ruffolo et al¹⁴ reported that most patients chose to forgo opioids after implementation of an opt-in program, but lack of follow-up pain assessment raised questions of adequate pain treatment after discharge.

Quiz Ref IDWe conducted a randomized clinical trial comparing perioperative pain management counseling with opioid prescription only on patient request with usual care with routine opioid prescription. The purpose of the trial was to determine whether allowing patient choice could reduce opioid consumption without having a negative effect on patients’ postoperative pain levels and quality of life. The primary outcome was daily postoperative pain score through postoperative day 7. Secondary outcomes included quantity of opioids prescribed and consumed, health-related quality of life (HRQOL), and clinical factors associated with opioid request and consumption. We hypothesized that postoperative pain would be equivalent between opt-in and routine opioid prescribing strategies and that an opt-in strategy would result in reduced opioid consumption.

We conducted a randomized clinical trial from June 1 to December 30, 2020, to compare pain scores and opioid consumption between an opt-in strategy and routine narcotic prescription following cervical endocrine surgery at a single institution. The protocol is available in Supplement 1. All adult (age, ≥18 years), English-speaking patients scheduled for outpatient surgery (<24 hours inpatient stay) were screened for eligibility. We excluded patients who were actively taking opioid pain medications. Patients were informed of the study during their preoperative clinic visit and subsequently enrolled over the telephone using a standardized script after informed consent was obtained. Participants received financial compensation if they completed all surveys and the HRQoL questionnaire (eAppendix 1 in Supplement 2). This study was approved by the University of California, Los Angeles Institutional Review Board and followed Consolidated Standards of Reporting Trials () reporting guideline for randomized clinical trials.

A random number–generated randomization schedule in a 1:1 allocation ratio was created before the start of the study by a research team member (D.T.N.) uninvolved in patient interactions. Treatment assignments were concealed in sealed, numbered envelopes until enrollment. Patients randomized to the intervention group were enrolled in the Postoperative Opt-In Narcotic Treatment (POINT) program. The program entailed a 5-minute preoperative pain counseling session using a standardized script (eAppendix 2 in Supplement 2), a 2-minute video review of the pain counseling session shown after surgery and before discharge (eAppendix 2 in the Supplement 2), and a discharge brochure for pain management strategies. The counseling sessions discussed the risks and adverse effects associated with opioids but did not advocate for or against opioid treatment after surgery. Patients were reassured of the option to call in for a rescue opioid prescription.

Quiz Ref IDPatients in the control group received no counseling. All patients were administered the Patient Health Questionnaire-2 (PHQ-2), which has high sensitivity for the diagnosis of major depression, given that symptoms of depression have been associated with increased postoperative pain and opioid use.^15-17

Intraoperatively, all patients received a bilateral superficial cervical plexus block with 10 mL of bupivacaine, 0.25%, bilaterally and a 5- to 10-mL subcutaneous injection of bupivacaine, 0.25%, with epinephrine 1:200 000 at the incision site before incision. Inpatient perioperative pain treatment was left to the discretion of the anesthesia and surgical teams. All opioids administered perioperatively were converted to oral morphine equivalents (OMEs) to adjust for differences in potency. Patients were observed in the postanesthesia recovery unit for at least 4 hours before discharge or admitted for overnight observation based on patient factors such as age, comorbidities, and distance from hospital to home. Patient pain levels were assessed every 1 to 2 hours in the postanesthesia recovery unit and every 4 hours on the ward.

Before discharge, patients randomized to the control group were routinely prescribed narcotics, in keeping with usual care. Nursing staff directed POINT patients to view the 2-minute pain counseling video and sign a form attesting to whether they chose to receive a narcotic prescription. In both groups, 10 tablets of hydrocodone, 5 mg, with acetaminophen, 325 mg (50 OMEs), was the standard prescription dispensed. Patients who had a history of poor tolerance for hydrocodone with acetaminophen were prescribed 10 tablets of codeine, 30 mg, with acetaminophen, 300 mg (45 OMEs), as a first-line alternative or 10 tablets of tramadol, 50 mg (50 OMEs), as a second-line alternative.

On postoperative days 1 through 7, all participants were administered daily surveys (eAppendix 3 in Supplement 2) to assess pain levels through PRIME for Patients (2020 UCLA Health), an iOS operating system–compatible, Health Insurance Portability and Accountability Act–compliant mobile application. Patients without iOS devices were administered surveys via an online password-protected data management and survey platform (REDCap, version 10.0.33; Vanderbilt University) or by telephone. An electronic gift card was offered to patients at the end of the study to encourage survey completion. Patients were asked to rate their peak pain score using the 11-point numerical rating scale over the previous 24 hours, identify sites of pain, and list the type, dose, and number of tablets of pain medication (both prescription and over-the-counter agents) used. Logged pain scores greater than or equal to 8 triggered an automated alert, which prompted a physician-initiated telephone call to the patient. On postoperative day 7, patients were administered an assessment of HRQOL over the first postoperative week.

The primary outcomes studied were patient-reported pain scores in the first postoperative week. Secondary outcomes studied were the quantity of opioid tablets prescribed and consumed in each group and the HRQOL assessment, which consisted of both the Patient-Reported Outcomes Measurement Information System 29-item profile (PROMIS-29 v2.0) profile, and the Secure Flourish Index. The PROMIS-29 v2.0 profile consists of 7 domains (physical function, anxiety, depression, fatigue, sleep disturbance, social function, and pain interference) standardized using a T-score metric with the mean (SD) of the US general population equal to 50 (10).¹⁸ A higher score in each domain may represent either a positive or negative health-related outcome (eg, higher physical function, higher pain interference). The Secure Flourish Index is a brief, 12-item holistic assessment of psychological well-being.^19,20 Each question is scored on a scale of 0 to 10, with higher scores indicating more favorable response, creating a composite index score of 0 to 120.

We also performed subgroup analyses to determine patient and perioperative risk factors associated with a greater likelihood of requesting and/or consuming opioids. We analyzed inpatient pain scores reported as the maximum score during hospitalization, 4 hours postoperatively, and the last recorded score before discharge (reflective of patients’ pain level when deciding whether to receive an opioid prescription in the POINT group).²¹

Patient demographic and baseline characteristics were compared between the 2 groups. The sample size was calculated based on a noninferiority design, and we defined a noninferiority margin as a difference of 2 points on the 11-point numerical rating scale for pain. To achieve a statistical power of 90% at a 1-sided α level of .05 and an SD of 2 points, 18 patients were needed in each group. In our interim analysis, we observed that half of the POINT patients opted in for a prescription. We expanded our enrollment with the aim to enroll 50 patients in each study arm, such that a projected 25 patients would each opt in and opt out of opioids in the POINT group to sufficiently power analyses of differences in pain scores observed between subgroups, while allowing for a 15% dropout margin. Analysis was conducted on the evaluable population.

We used the Wilcoxon rank sum test or 2-sided, unpaired t test for comparison of continuous variables and Fisher exact test for categorical variables, with statistical significance defined as P &±ô³Ù; .05. Mean differences in pain scores between groups were calculated with 95% CIs, and regression models were used to adjust for any covariates. Noninferiority was established if the upper bound of the 95% CI was less than 2 points on the numerical rating scale of 0 to 10. Mean differences in PROMIS-29 T scores and the Secure Flourish Index level were also calculated with 95% CIs, adjusted for covariates. Because minimal clinically important difference for PROMIS-29 and Secure Flourish have not been validated in patients undergoing endocrine surgery, we used the Norman and colleagues’²² definition of minimal clinically important difference as 0.5 SD, using the SD of the control group. Noninferiority was established in the Secure Flourish Index and the PROMIS-29 physical and social function domains if the lower bound of the 95% CI was within the minimal clinically important difference. Noninferiority was established in the remaining PROMIS-29 domains if the upper bound of the 95% CI was within the minimal clinically important difference. A superiority analysis of the POINT patients who opted in or out of opioid prescriptions was performed using the Wilcoxon rank sum test to determine whether there were differences in pain scores or HRQOL. Multivariate logistic regression was used to analyze independent factors for opting in for a prescription and opioid consumption after identifying factors that were statistically significant in univariate analysis. All analyses were performed in R, version 1.4.1103 (R Foundation for Statistical Computing).

A total of 115 patients were enrolled, with 58 patients (50.4%) randomized to receive routine narcotic prescription (control), and 57 patients (49.5%) randomized to POINT (Figure 1). After exclusions, 102 patients remained in the study, with 54 (52.9%) in the control arm and 48 (47.1%) in the intervention arm. Participants included 79 women (77.5%) and 23 men (22.5%); median age was 52 (IQR, 43-62) years. Baseline characteristics were similar between groups with the exception of median (interquartile range [IQR]) age (control, 55 [IQR, 48-65] vs POINT, 46 [IQR, 39-53] years), distribution of procedure type (parathyroidectomy: control, 29 [53.7%] vs POINT, 16 [33.3%]), and operative time (median, control, 70 [IQR, 57.3-88] vs POINT, 79 [IQR, 64.5-108] minutes) (Table 1). Survey completion of 5 or more daily surveys was 97.1% (99 of all 102 patients). Outcomes data from 3 patients (2 control patients and 1 POINT patient who opted out of opioids) who completed fewer than 5 surveys were excluded from analysis.

Patients in both groups received similar perioperative quantities of acetaminophen and opioid analgesics (Table 2). Few patients received additional nonopioid analgesics, such as gabapentin (n = 6) and ketorolac (n = 1). Postoperative pain scores before discharge were similar between groups; for example, the highest inpatient median score was 6 (IQR, 4-7) in the control group vs 6 (IQR, 3-7) in the POINT group.

Quiz Ref IDThe highest patient-reported pain score in the first postoperative week was a median of 6 (IQR, 4-8) for the control group and 6 (IQR, 5-7) for the POINT group (P = .71). The upper boundary of the 95% CI was within the noninferiority margin of 2 points on the numerical rating scale for each day of the first postoperative week (eFigure in Supplement 2). The most common primary sources of pain were incisional pain (36.5%) and sore throat (35.8%), followed by shoulder pain (8.1%), headache (6.9%), and back pain (3.9%). The proportion of patients who triggered an automated alert by reporting a pain score of 8 or higher was also similar in the control vs POINT groups (32.7% vs 23.4%; P = .27) (Table 2). A total of 28 physician-initiated telephone calls resulted from the alerts in the control group (mean [SD], 0.5 [0.9] calls per person) and 14 in the POINT group (mean, 0.3 [0.6] calls per person; P = .13).

In a subgroup analysis of POINT patients, those who opted in for opioid medication on discharge (POINT-In) experienced significantly greater pain than those who opted out (POINT-Out) through postoperative day 3 (Figure 2; eTable 2 in Supplement 2). The highest reported pain score was a median of 7 (IQR, 6-8) in the POINT-In group vs a median of 5 (IQR, 4-7) in the POINT-Out group (P = .02).

Quiz Ref IDThe control group was prescribed 550 opioid tablets, amounting to a total of 2690 OMEs. Of patients completing 5 or more daily surveys, 27 control patients (50.0%) reported using opioids, collectively consuming a total of 135.5 tablets (24.6% of tablets dispensed). The POINT group was prescribed a total 230 tablets to 23 patients (47.9%) who opted in for a prescription. Of those, 17 POINT patients (73.9% of POINT-In, 35.4% of POINT group) reported consuming a total of 66 tablets (28.7% of tablets dispensed). Only 10% of all patients required more than 6 tablets. Of patients who used opioids, the median OMEs consumed over the week was similar between groups (control, 20 [IQR, 10-34] vs POINT, 15 [IQR, 10-30]; P = .49). The groups did not differ significantly in the use of over-the-counter pain medication. No POINT-Out patients required a rescue prescription.

The HRQOL scores at postoperative day 7 in the POINT group were noninferior to scores in the control group in the Secure Flourish Index and all PROMIS-29 domains, with the exception of the anxiety domain (eTable 1 in Supplement 2); however, anxiety T scores were not significantly different between groups (median T score, 51.4 [IQR, 40.3-57.7] vs 53.8 [IQR, 44.2-58.6]; P = .26). Compared with POINT-Out patients, POINT-In patients had significantly more anxiety (median T score, 56.2 [IQR, 54-63.2] vs 49.8 [IQR, 40.3-53.8]; P = .001) and pain interference (median T score of 55.7 [IQR, 55.7-61.5] vs 54.6 [IQR, 41.6-57]; P = .01), but the latter did not exceed the threshold for minimal clinically important difference.

We performed an analysis of all study patients to determine factors associated with opioid consumption (Table 3). In univariate analysis of opioids consumed vs not consumed, body mass index (median [IQR], 28.8 [24.6-31.5] vs 25.0 [21.7-29.5]; P = .003) (calculated as weight in kilograms divided by height in meters squared), history of narcotic use (38 [86.4%] vs 30 [54.5%]; P = .001), surgical procedure (eg, total thyroidectomy: 15 [34.1%] vs 9 [16.4%]; P = .04), incision length (median, 4 [range, 3-4] vs 3 [range, 2.5-5] cm; P = .02), and higher inpatient postoperative pain scores (median [IQR], 7 [6-8] vs 4 [2-6]; P < .001) were factors associated with opioid consumption. We included only peak pain scores in the multivariate regression model, as opposed to median or last recorded pain score, because the peak scores had the strongest association with opioid use. Quiz Ref IDIn multivariate analysis, higher body mass index (odds ratio, 1.11; 95% CI, 1.01-1.23; P = .03), highest postoperative pain score (odds ratio, 1.24; 95% CI, 1.04-1.52; P = .02), and history of narcotic use (odds ratio, 4.80; 95% CI, 1.51-18.45; P = .01) were identified as independent factors associated with consuming opioids.

We also performed a subgroup analysis of POINT patients to identify factors associated with opting in for postoperative opioids (eTable 3 in Supplement 2). In multivariate analysis, every 1-point increase in last recorded pain score was associated with a 1.7-fold increase in the likelihood of requesting an opioid prescription (95% CI, 1.15-2.68; P = .01). Patients who previously used opioids were 7.5 times more likely than opioid-naive patients to opt in compared with opioid-naive patients (95% CI, 1.61-50.11; P = .02).

We found that patients randomized to an opt-in strategy for postoperative narcotic treatment following cervical endocrine surgery were prescribed fewer narcotics than those randomized to receive routine narcotic prescription without experiencing increased pain levels. Randomization to the POINT program also did not have a negative effect on HRQOL. The cumulative amount of opioids consumed by the control group was twice that consumed by the POINT group, which is consistent with other studies that found an association between prescription availability and higher opioid consumption.^23,24 No patients required a rescue opioid prescription if they opted out at the time of discharge, which indicates that the patients were able to predict their own postoperative opioid requirements. Despite having access to opioids, POINT-In patients had higher postoperative pain levels and worse HRQOL in the anxiety and pain assessments than POINT-Out patients. These findings suggest the subjective perception of pain may be more influenced by patient- and surgery-specific factors; a certain level of postoperative discomfort appears unavoidable and is unaffected by additional opioids.

Other contemporary studies have reported that many patients do not require opioids following cervical endocrine operations; when prescribed opioids after endocrine operations, 34% to 52% of patients do not take opioid medication.^4,25 In our study, 53% of POINT patients chose not to receive a prescription for opioids, and there were no calls for rescue prescriptions. We routinely perform preincision bilateral cervical plexus block, which may lower postoperative opioid requirement.^26,27 Our findings support the opt-in strategy as a means of reducing overprescription of opioids and also indicate that patients who choose to forgo opioids are not experiencing increased pain at home but not reporting it.

The proportion of POINT patients who chose to receive an opioid prescription was 5- to 10-fold greater than in prior studies of an opt-in strategy for opioid prescriptions.^14,28 The larger number of opioid requests may reflect the equipoise we purposefully maintained during patient counseling. Through a patient-centered approach to decision-making, we encouraged participants to carefully consider the risks and benefits of using opioids but did not discourage prescription requests. In addition, patients randomized to POINT understood that those in the control group were receiving opioids as part of our previous standard practice, in contrast to Ruffolo et al¹⁴ and Kwan et al,²⁸ who presented an opt-in strategy as the new standard of care. Most patients in both the control and POINT groups had used opioids after surgery previously, which normalized narcotic treatment as part of routine postoperative care.

Previous narcotic use or substance abuse, depression or anxiety, and postoperative pain scores have been identified as factors associated with opioid consumption.^14,29-31 We screened patients with the PHQ-2, which has been validated with similar or higher sensitivity compared with the PHQ-9 to identify patients with major depression,¹⁵ but PHQ-2 positivity was not predictive of either opting in or opioid consumption. Our factors for opioid consumption were higher body mass index, previous narcotic use, and higher inpatient postoperative pain scores. Body mass index as an independent predictor of opioid consumption may reflect greater positioning-related discomfort and/or pain from more forceful retraction during surgery. Previous narcotic use was the strongest predictor for both opting in and consuming opioids. Most prior opioid experience was in the context of postprocedural treatment and not chronic use. These findings have 2 main implications: every effort should be made to avoid exposing opioid-naive patients to opioids, because opioid naivety predicts against future use, and patients with higher immediate postoperative pain levels and previous opioid exposure may benefit from additional nonopioid adjuncts for postoperative pain control to minimize opioid consumption. Patients reported postoperative pain as frequently from sore throat as incisional pain, suggesting that efforts to minimize postintubation discomfort may lessen opioid consumption. Regardless of individual risk factors, patients can reliably determine their own need for opioids after surgery, demonstrated by the 76% of POINT patients who opted in to receive opioids and subsequently used the medication. Despite this level, only 29% of total dispensed tablets in the POINT-In group were consumed, which suggests reducing the standard prescription quantity would adequately treat most patients requesting a narcotic prescription after cervical endocrine surgery.

Our study has limitations. First, this trial was performed at a single institution, which limits the generalizability of our findings to dissimilar practice settings, operative techniques, and perioperative pain management practices. Second, we used self-reported data on opioids consumed, which are less reliable than pill counts. Third, we did not administer a preoperative HRQOL assessment and assumed patients in both groups shared a similar baseline quality of life.

In this randomized clinical trial, preoperative counseling combined with a patient-choice strategy for prescribing opioid analgesics reduced the number of opioids prescribed without worsening postoperative pain or HRQOL. Patients should be able to choose whether to include opioids in their pain management regimen after cervical endocrine surgery to prevent undertreating those with a postoperative opioid requirement but reduce the risk of opioid exposure in those who require minimal analgesics. Most patients who request postoperative opioids do not require more than 30 OMEs.

Accepted for Publication: June 28, 2021.

Published Online: September 8, 2021. doi:10.1001/jamasurg.2021.4287

Correction: This article was corrected on November 10, 2021, to fix a funding/support omission in the end matter.

Corresponding Author: Catherine Y. Zhu, MD, Department of General Surgery, UCLA David Geffen School of Medicine, 10833 Le Conte Ave, 72-228 CHS, Los Angeles, CA 90095 (catherinezhu@mednet.ucla.edu).

Author Contributions: Drs Zhu and Wu had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Zhu, Schumm, Lin, Yeh, Wu.

Acquisition, analysis, or interpretation of data: Zhu, Schumm, Hu, Nguyen, Kim, Tseng, Yeh, Livhits, Wu.

Drafting of the manuscript: Zhu, Schumm, Hu, Nguyen, Wu.

Critical revision of the manuscript for important intellectual content: Zhu, Schumm, Hu, Kim, Tseng, Lin, Yeh, Livhits, Wu.

Statistical analysis: Zhu, Nguyen, Kim, Tseng, Wu.

Obtained funding: Yeh.

Administrative, technical, or material support: Zhu, Schumm, Nguyen, Lin, Yeh, Wu.

Supervision: Zhu, Yeh, Livhits, Wu.

Conflict of Interest Disclosures: None reported.

Funding/Support: Dr Schumm is supported by a grant from the H&H Lee Research Program.

Role of the Funder/Sponsor: The H&H Lee Research Program had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Data Sharing Statement: See Supplement 3.