Feasibility of Prospectively Comparing Opioid Analgesia With Opioid-Free Analgesia After Outpatient General Surgery

This pilot randomized clinical trial (RCT) assessed the feasibility of a full-scale RCT to compare the outcomes of opioid analgesia vs opioid-free analgesia prescribed to adult patients after undergoing outpatient general surgery.

undergoing outpatient surgery (day surgery) and interventions will be delivered with 89 flexibility in medication selection. An embedded qualitative study will be conducted to 90 help optimize trial design based on clinicians' and patients' perspective 22 . The study 91 protocol will be reviewed by the McGill University Health Centre (MUHC) Research 92 Ethics Board and patient recruitment will start after ethics approval. All participants will 93 sign a written consent form and a paper copy of the form will be attached to the patient 94 medical chart. Trial registration and protocol information will be made available at the  or a delegated clinician member of his/her team, will be responsible for signing a pre-148 written analgesia discharge prescription in accordance with the treatment that patients 149 have been allocated to. 150 Participants and treating clinicians (i.e., surgeons, anesthetists, and nurses) will not be 151 blinded to treatment allocation due to the complexity of the medication prescribing 152 strategies. To reduce potential risk of detection bias (systematic differences between 153 groups in how outcomes are determined), outcome assessors will be blinded to 154 treatment allocation. Patient-reported outcomes and treatment adherence data will be 155 collected via self-administered electronic questionnaires distributed using REDCap 156 (http://project-redcap.org/) and completed by patients via smartphone, tablet or personal electronic assessment will complete the questionnaires via telephone interviews with a blinded assessor; in this case, data will be recorded in paper forms and subsequently 162 transferred to the REDCap database. Prior to every telephone interview, patients will be 163 reminded not to disclose their allocation status or information about pain medications. 164 To prevent unblinding, telephone follow-ups to monitor treatment adherence will be 165 done by a team member not involved in outcome assessment. 166 Outcome data that are not patient-reported (e.g., postoperative complications, 167 unplanned healthcare utilization, chronic opioid use) will be obtained from medical 168 records by a blinded assessor. Any inadvertent unblinding will be reported. 169 Effectiveness of blinding will be estimated by asking assessors to guess patients' group 170 allocation at one month after surgery (after the last patient questionnaire is responded).

171
Statistical analysis will also be blinded with information regarding allocation protected by 172 codes that will be revealed only after all analyses are completed.

173
Interventions 174 Opioid analgesia (OA) group 175 Patients randomized to the OA group will receive the current standard of care in the 176 participating centers, which includes the prescription of around-the-clock non-opioid 177 analgesics (acetaminophen and/or NSAIDs/COX-2) and a supply of opioids to be used 178 as a rescue in case of breakthrough pain (i.e., pain that erupts while a patient is already 179 medicated with painkillers). Prior to hospital discharge, patients will undergo a 180 medication education session with the PACU nurse and be advised to fill their 181 prescription at a pharmacy of their preference. Medication education sessions with a 182 nurse prior to discharge are part of standard care at MUHC. In light of the pragmatic 183 nature of this trial, the specific round-the-clock analgesia and rescue opioid regimens  Examples are included in eFigure1.
To confirm if patients randomized to this group are treated according to current standards of care, we will conduct a retrospective chart review of post-discharge 191 analgesics prescribed to patients who underwent the eligible surgeries between 192 September 01 to October 31, 2019. We estimate that, within this 2-month period, the 193 electronic medical charts of approximately 100 patients will be reviewed. Only data 194 regarding the surgical procedure conducted and analgesia regimen prescribed (pain 195 medication received, dosage, frequency of administration, treatment duration) will be 196 collected by the research team. Opioid-free (OF) analgesia group 198 Patients randomized to the OFA group will receive a prescription of around-the-clock 199 non-opioid analgesics (Acetaminophen alone or combined with NSAIDs/COX-2). In As opioid-free analgesia is new to our setting, specific strategies will be implemented to 215 ensure that patients are receiving adequate pain management during the pilot trial. A 216 'hotline' (dedicated mobile phone that will be kept with study staff in shifts) will be 217 available 24/7 in case patients experience persistent pain despite the use of rescue 218 analgesia. When this line is called, study staff will inform patients about the management options available according to their treatment allocation. An information 220 sheet containing the 'hotline' contact details will be provided to patients prior to PACU 221 discharge (see Discharge Information Sheet -Opioid-free Group). 222 Patients in the opioid-free group will have a back-up prescription of opioids (regimen 223 decided by the primary surgeon) faxed to the 24h pharmacy closest to their residence.

224
This prescription will be faxed upon patient discharge from the hospital, with a brief 225 letter informing the study and ethics approval (see Information Sheet for Pharmacy). 226 When a patient calls the study staff reporting persistent pain, they will be informed about 227 the availability of the prescription and the pharmacy address. To prevent patients to fill 228 their opioid prescription 'just in case', they will not be informed about the availability of 229 the prescription unless they report persistent pain. When the prescription is filled, 230 education about the use of opioids will be given by the pharmacist as per routine 231 pharmacy services. If pain persists despite the use of opioids, patients will be advised to 232 proceed according to the management of persistent pain in the opioid group, as 233 described below.

234
As per the institutions' current practice, patients in the opioid group who experience 235 persistent pain will be advised to call their primary surgeon's office/clinic during working 236 hours (weekdays, 8AM to 4PM) or visit a hospital emergency room (ER) for further 237 evaluation (after-hours and weekends). If an ER visit is required, patients will be asked 238 to give preference to visiting the ER of the hospital where his/her surgery had been 239 performed. An information sheet containing specific instruction will be provided to 240 patients prior to PACU discharge (see Discharge Information Sheet -Opioid Group). 241 Changes of initial prescription will be entirely up to the patients' surgical team and/or ER 242 physician.

243
Adherence and study discontinuation 244 Treatment adherence (i.e., patients in each group taking their pain medications as 245 prescribed) will be monitored via self-administered electronic questionnaires distributed 246 using REDCap (http://project-redcap.org/) and completed by patients via smartphone, 247 tablet or personal computer from postoperative day (POD) 1 to POD 7 and at 2, 3 and 4 248 weeks after surgery. Electronic adherence data will be transmitted directly to the REDCap database and verified by unblinded study staff. Patients will also be offered the 250 option to respond to adherence questionnaires via telephone; in this case, data will be 251 recorded in paper forms by unblinded staff and subsequently transferred to the REDCap 252 database. Patients will be instructed to take medications for postoperative pain only in 253 accordance with the initial discharge prescription or based on prescriptions given by 254 healthcare providers after hospital discharge. If patients desire discontinuation of any of 255 the study medications, they will be advised to discuss other medication options with the   As a pilot RCT, this study will primarily focus on feasibility outcomes. Clinical outcomes 275 will be assessed secondarily to inform the measurement strategy and sample size 276 requirements for a future full-scale RCT.

Assessment of feasibility outcomes (primary)
A full-scale RCT be deemed feasible if, during the pilot study period (4 months):

279
• At least 70% of patient undergoing the outpatient general surgery procedures of 280 interest are eligible to be randomized.

281
• At least 90% of the surgeons who agreed to have their patients randomized will 282 comply with the agreement, i.e., not change their minds (see section 'pilot study 283 sample size and feasibility' below).

284
• At least 50% of eligible patients agree to participate in the study and are 285 randomized.

286
• At least 80% of the randomized patients comply with their allocated treatment (i.e.

287
will take their pain medications as prescribed).

288
• At least 80% of the patients randomized complete outcome assessment at 30-days 289 after surgery.

290
• Among patients who complete outcome assessments, the proportion of missing 291 data is less than 10% (i.e., non-response to questionnaires or specific questionnaire 292 items).

293
To determine recruitment rates, study staff will keep a screening log of patients 294 approached, patients who fulfill eligibility criteria and those who do not fulfill eligibility 295 criteria. Reasons for ineligibility will be recorded. This log will also record information 296 about eligible patients who were successfully recruited, and those who were not 297 recruited despite being eligible. In the event of surgeons opting for not recruiting 298 patients despite eligibility, rates and reasons will be recorded. Adherence to treatment 299 will be assessed by comparing patients' analgesia prescription at discharge to self-300 reported analgesic intake at each time-point of assessment. Follow-up completion rates 301 and missing outcome data will be computed based on REDCap entries (date-and time-302 stamped). Patients will be considered to have withdrawn from the trial if they miss three 303 consecutive assessments and then permanently stop responding the questionnaires.

304
Reasons for patients not consenting participation, not completing follow-ups or 305 withdrawing from the trial will be recorded whenever possible.

Assessment of clinical outcomes (secondary)
Our clinical outcome measurement strategy was informed by the World Health text message or email (according to the patient's preference) in the morning, with up to 339 3 reminders sent in case of no response. Participants will be asked to, preferably, 340 complete the questionnaires in the morning to prevent bias associated to 341 chronobiological variations in pain 47 . Patients who opt for non-electronic assessment will 342 complete the questionnaires via telephone interviews, preferably conducted before 343 12PM. Information regarding postoperative complications and unplanned healthcare 344 utilization will be obtained via patient self-report (week 4) and verified using electronic 345 medical records. Information regarding opioid prescription dispensing will be obtained   Data generated from the pilot study will help inform a full-scale RCT by testing the study 372 procedures; therefore, no inferential statistical analyses will be performed to compare 373 groups. Continuous variables will be summarised using means, standard deviations 374 (SDs), medians, lower and upper quartiles, minimum, maximum and number of 375 observations. Categorical variables will be summarised using frequencies and 376 percentages. To address feasibility, descriptive statistics of patients approached, 377 screened, eligible, consented and randomised, treatment adherence and follow-up 378 completion rates will be computed. Completeness of follow-up will be compared 379 between trial arms. Reasons for non-consent, exclusion and trial withdraw will be 380 recorded and reported. Baseline data will be summarized descriptively to assess 381 comparability between treatment arms and to highlight any differences between patients 382 who were randomized, who withheld consent and who did not meet eligibility criteria.

383
Analyses of postoperative outcomes will be exploratory, descriptive and follow the 384 intention-to-treat principle, with all patients analyzed in their assigned treatment group.

385
The primary outcome measure to be addressed in the full-scale RCT will be informed by 386 data from this pilot trial. Decision will be based on acceptability and relevance to 387 patients and clinicians (qualitative study described below), completion rates, evidence of measurement properties according to previous literature, effect sizes and sample size requirements. There are no planned interim data analyses; however, if the TMT 390 identifies that recruitment, randomization and data collection are below target, strategies 391 will be implemented to improve progress. Any changes to methods after trial 392 commencement will be documented and reported. Any future revisions to protocol and 393 consent forms will be implemented only after IRB approval.

Study objective:
The objective of this study is to inform, via qualitative research methods, optimal study 417 design of a full-scale RCT by assessing patient and clinician perspectives on trial 418 conduct, participation, interventions and measurement strategy. will be invited to participate in one-on-one qualitative interviews. Patients who do not 437 consent to randomization in the trial will also be invited to participate in the interviews as 438 they may provide relevant insights regarding the consent process and study 439 acceptability. In order to capture the heterogeneity of outpatient general surgery 440 procedures and improve sample representativeness, we will use a quota sampling 441 method 57 targeting patients representing a broad spectrum of demographic, clinical and 442 surgical characteristics (Table 1). Patients will be offered the opportunity to be 443 interviewed face-to-face or by telephone. Patients will be informed about the qualitative 444 interviews during preoperative recruitment and those who are interested will be contacted after their involvement with the trial. A consent form specific to the qualitative 446 study will be signed prior to the interviews. To ensure accurate recall, patients will be 447 interviewed no later than 6 weeks after their surgery. Interviews will focus on (1) 448 acceptability of the study, (2) personal experience with the process of recruitment and

Patient recruitment process
Subsequent contact for participation in the qualitative study will be made upon patient 457 authorization. Patients will be approached as follows, depending on whether they 458 agreed or not to participate in the pilot RCT:

459
(1) Patients who agreed to participate in the pilot RCT and signed the informed consent one-on-one qualitative interviews. Interviews will be conducted face-to-face or by 477 telephone after informed consent is obtained. In order to improve sample 478 representativeness, we will use a quota sampling method 57 targeting clinicians 479 representing a broad spectrum of demographic and professional characteristics (Table   480 2). Interviews will be conducted within the period of patient recruitment to ensure 481 accurate recall. Interviews with clinicians will focus on (1) acceptability of the study, (2) 482 experience operationalizing the study in practice (i.e., recruiting patients and providing  of interview data will be conducted via inductive thematic analysis informed by Braun 503 and Clarke (2006) 55 . Thematic analysis is a method used to identify, analyze, and report 504 themes and subthemes within the interviews to provide a rich description of the 505 qualitative data. The inductive approach to thematic analysis is data-driven, where the 506 themes will be derived from within the data themselves and no pre-existing coding 507 framework will be applied during analysis. Based on data obtained from the first 508 interviews, two independent researchers (coders) will code each interview transcription 509 and search for recurring themes. The coding process will be conducted using the

517
The findings from this qualitative study will be regularly fed back to the trial steering 518 committee so that aspects of the pilot study conduct can be reviewed iteratively where 519 appropriate. Themes for which saturation is reached will be classified as meaningful 520 issues to inform the optimal design of the full-scale RCT.

PART II. Embedded qualitative study
20 participants (estimate) -A minimal of 10 participants (5 patients, 5 clinicians) will be recruited but the total sample may vary according to data saturation.

Total sample size
100 participants (estimate). who are committed to encouraging recruitment across both study sites.

571
As certain surgeons may heavily rely on opioids to treat postoperative pain, we 572 anticipate that some may refuse to recruit selected patients or refuse to recruit patients 573 altogether. Similarly, some patients may be doubtful about the efficacy of pain treatment 574 without opioids and refuse randomization. This issue will be addressed by comparing 575 demographic and surgical data of randomized patients versus non-randomized patients.

576
Differences may suggest that our results are not generalizable to certain surgical 577 populations, indicating venues to improve our patient selection criteria and/or 578 recruitment process. Our integrated qualitative study including interviews with patients who refused randomization and surgeons with low recruitment rates will provide 580 fundamental insights into the strategies to mitigate these potential issues. The 581 qualitative study will also provide relevant information to optimize our measurement 582 strategy, which currently includes daily follow-up in the first 7 days after surgery. The 583 use of daily outpatient follow-up assessment has been successful in a recent RCT on 584 postoperative analgesia 37 but, if proven unfeasible in our setting, strategies will be 585 implemented to reduce patient burden (e.g., reducing follow-up frequency).

597
Retrospective chart review: All the information collected during our preliminary chart 598 review will remain confidential to the extent required and provided by law. A study ID 599 number will be assigned to each patient's chart. No code linking patient identifiers to 600 patient data will be kept and it will not be possible to identify patients.

601
Pilot Trial: All data collected in our pilot trial will be entered and stored in a password- statistical program for analysis. A study ID number will be assigned to each participant.

605
Information collected in paper-based forms will be kept in locked cabinets within a 606 locked office (R2-111). Participants will be identified by a code to protect their identity. A 607 document linking the codes to the participants' identity will be kept separately in a 608 password protected file, which can only be accessed by the study staff.
All data will be kept under safe storage for 7 years and then deleted, shredded or incinerated. Only investigators will have access to the data. Furthermore, the results 611 and the project may be published, but patients' identity will not be revealed.

613
Results from this pilot trial will inform the planning and commissioning of a future full-  • After discussion, the team decided that the two components of this pilot study (quantitative and qualitative) would be reported in separate manuscripts. June 2021. After patient recruitment Outcome measure/data analysis Data on overall impression of treatment effectiveness at each postoperative timepoint.
• We noticed that this author-generated question was accidently excluded from the final version of the Redcap questionnaire distributed to patients. Therefore, these data were not analyzed or reported in the manuscript. Impressions about treatment effectiveness were detected via other patient-reported questionnaires. June 2021. After patient recruitment and data analysis Outcome measure/data analysis Data regarding satisfaction with pain management at postoperative week 4 • After data analyses, the team realized that findings regarding satisfaction with pain management at postoperative week 4 were redundant (did not add relevant information in comparison to the data reported by patients on week 1). For this reason, this information was not reported in the manuscript. This data would not be useful as it is subject to recall bias given that most patients do not use pain medications beyond week 1.