1 Introduction

The opioid epidemic remains a persistent driver of deaths for Americans, and current estimates suggest that opioid overdose is among the top five causes of death.¹ In 2019 alone, over 70% of overdose deaths involved an opioid, with the Center for Disease Control estimating that around 136 people die per day from an opioid overdose.² Despite increased recognition of the danger of narcotics, opioid analgesia continues to be prescribed at a high rate.³ Musculoskeletal injuries are among the most common reasons patients will receive an opioid prescription,⁴ and orthopaedic surgeons prescribe opioids more frequently, with more refills than any other surgical specialty.⁵ Following orthopaedic procedures, 9% of patients using postoperative opioids will become opioid dependent.⁶ Continued opioid use following orthopaedic procedures has demonstrated an association of increased risk of Deep Vein Thrombosis (DVT)/Pulmonary Embolism (PE), postoperative pulmonary complications, wound infection, wound dehiscence, and prosthetic joint infection.⁷ Preoperative and prolonged postoperative opioid use has been correlated with worse patient reported outcomes, increased health care utilization, and early revision surgery across a variety of orthopaedic sports procedures.^8–11

Opioid stewardship has become imperative to orthopaedic clinical practice. Orthopaedic surgeons across subspecialties have successfully pursued opioid-free postoperative multimodal pain management regimens. Patients undergoing arthroscopic procedures, total shoulder arthroplasty, total hip arthroplasty, total knee arthroplasty, and various trauma procedures have shown satisfactory pain control with the use of multimodal pain medications.^12–18 In a study prospectively evaluating breakthrough opioid use following anterior cruciate ligament (ACL) reconstruction, rotator cuff repair (RCR), partial meniscectomy, and shoulder labrum repair, Moutzouros et al. demonstrated that approximately 50% of patients tolerated an opioid-free multimodal pain regimen.¹⁹ Further prospective randomized control trials showed that a multimodal non-opioid pain regimen provided at least equivalent pain control and patient satisfaction compared with a standard opioid-based regimen following arthroscopic shoulder labral repair, RCR, meniscus surgery, and ACL reconstruction.^20–23 Despite these promising results, patients continue to request opioid analgesia for postoperative pain control.

Multiple studies have evaluated risk factors associated with increased opioid use in the postoperative period. These studies have demonstrated an association between persistent use and a history of psychiatric diagnosis,^24–29 preoperative opioid use,^29–33 and concomitant procedures.^34,35 However, the majority of these studies were retrospective evaluations in which patient preferences were not evaluated. These studies also did not assess patient experiences or perceptions about opioid analgesia. The purpose of the present investigation was to characterize patient preferences for postoperative analgesia between opioid versus nonopioid pain management. We hypothesized that patients reporting higher levels of preoperative pain, history of a psychiatric disorder, and preoperative opioid use would have a higher propensity for choosing an opioid-based pain regimen.

2 Methods

The present study received IRB approval prior to data collection and initiation of the study. All patients aged 15 years or older presenting to a single fellowship-trained sports surgeon from August 2020 to May 2021 at a single institution who met operative indications were prospectively recruited for the study. A wide range of ages were included to capture the whole spectrum of patients who typically present to sports medicine clinics and undergo surgery. Prospective recruitment of patients was completed after 100 patients were enrolled. Our primary outcome of interest was pain management protocol preference. Eligible surgeries included arthroscopic RCR, ACL reconstruction, meniscal surgery (meniscus repair or partial meniscectomy), arthroscopic shoulder labral repair or subacromial decompression, distal clavicle excision, elbow ulnar collateral ligament (UCL) reconstruction, elbow bursectomy, quadricep tendon repair, and shoulder hemiarthroplasty. Patients were excluded from the study if undergoing revision surgery, had concomitant injuries, had used opioids for more than three months preoperatively as documented via chart review, or were unable to read English.

The primary surgeon was blinded to postoperative analgesic protocols. All patients received counseling about the risks and benefits of taking opioids and multimodal drug protocols. At a preoperative visit, patients were then presented with the choice between an opioid and non-opioid pain regimen for postoperative pain control. Informed consent was obtained. Study participants completed two separate, one-page surveys. The first survey was labeled “Postoperative Pain Control” (Appendix 1) and was designed to determine patient's preference of opioid versus non-opioid pain control. Each question contained 4 multiple choice questions with 2–6 potential responses for each question. The second survey was a validated Opioid Risk Tool (ORT) (Appendix 2),³⁶ which classifies patients as low, moderate, or high risk for opioid abuse. Patients were then asked to rank their pain on a Visual Analog Scale (VAS) preoperatively. Patient-Reported Outcomes Measurement Information System (PROMIS) surveys were administered on an iPad during each clinic visit. PROMIS surveys provide a reliable, precise measure of patient-reported health status for physical, mental, and social wellbeing; are publicly available from the NIH; and span approximately 70 domains.³⁷

Collected demographic information included the following: age, race, sex, and body mass index (BMI); past medical history of anxiolytic (diazepam) use, alcohol abuse, illegal drug abuse, sexual abuse, attention deficit disorder (ADD), obsessive compulsive disorder (OCD), bipolar disorder, schizophrenia, and depression; family history of alcohol abuse, illegal drug abuse, and prescription drug abuse; history of previous surgery and type of pain control with previous surgery; and worker's compensation status. The PROMIS scores collected included PROMIS Upper Extremity (PROMIS-UE), PROMIS Physical Function (PROMIS-PF), PROMIS Pain Interference (PROMIS-PI), and PROMIS Depression (PROMIS-D). VAS score ranged from 1 to 10, and risk of opioid abuse was based on the aforementioned ORT. Patient preference for opioid versus nonopioid analgesia was evaluated by using the following concerns: postoperative pain, addiction, cost, previous experience with opioids, and side effects of the medications.

3 Results

One hundred patients met inclusion criteria with 35% (35/100) of participants undergoing a meniscus repair or debridement, 28% (28/100) ACL reconstruction (18 bone-patellar tendon-bone (BPTB) autografts vs. 10 with hamstring autografts), 11% (11/100) shoulder labral repair and/or subacromial decompression, 10% (10/100) rotator cuff repair (RCR), and 16% (16/100) underwent a procedure classified as other, which included elbow UCL reconstruction, elbow bursectomy, quadricep tendon repair, and shoulder hemiarthroplasty. There were 58 patients who preferred nonopioid multimodal analgesia. Patient demographics by group (nonopioid versus opioid) are listed in Table 1. Of the total cohort, prior surgical history was available for 97 of the 100 patients in the cohort, 78.4% (76/97) of which had undergone previous surgery. Prior pain control after surgery was available for 96 of the 100 patients in the cohort and 68.0% (66/96) of those individuals had used opioids for pain control for their initial surgeries. No significant demographic differences were found between patients who chose opioid and patients who chose non-opioid pain regimens for postoperative pain control (Table 1).

Preoperative PROMIS and ORT scores did not differ significantly between those preferring non-opioid pain control versus opioids for pain management (Table 2). Those who preferred opioids did, however, have significantly higher preoperative VAS scores (5.19 vs. 3.14 respectively, p < 0.001) and were more likely to have undergone previous surgery (p = 0.023) with 90.2% of patients who chose opioid pain control postoperatively had previous surgery compared to 69.6% of patients who chose non-opioid postoperative pain control. Those who chose opioids were also more likely to have received opioids for pain control after prior procedures (p = 0.003). No association was found between patient preference for opioids and workers compensation, psychiatric history, a personal or family history of substance use, or a history of sexual abuse (Table 3). There was a 97% response rate from the patient cohort when asked about reasons for their pain control preferences, depicted in Table 4. Patients preferring opioids reported more concerns about postoperative pain (p < 0.001) and were less concerned about the risk of addiction (p < 0.001), side effects (p < 0.001), or the pill burden associated with taking pain medications (p < 0.001). Of those preferring non-opioid pain control, 25.5% (14/55) of patients cited prior experience with opiates as a reason they chose non-opioid management, which was significantly more than patients who chose opioid pain control postoperatively (p < 0.001).

Multivariate regression analysis revealed that the odds of preferring an opioid medication were 29.6 times higher for patients who were concerned about postoperative pain (OR = 29.60, p < 0.001, 95% CI [9.43–92.92]) and 10.6 times higher for concerns about pill burden (OR 10.60, p = 0.003, 95% CI [2.22–50.54]). African Americans were 2.99 times more likely to prefer opioids than Caucasians (OR 2.99, p = 0.033, 95% CI [1.10–8.16]). Patients were less likely to prefer opioids when concerned about addiction (OR 0.06, p < 0.001, 95% CI [0.01–0.27]) and side effects of the medications (OR 0.12, p < 0.001, 95% CI [0.04–0.35]). Patients were more likely to choose a nonopioid regimen after already undergoing surgery in which nonopioid analgesia was utilized postoperatively (OR 0.11, p = 0.043, 95% CI [0.01–0.95]), and were more likely to choose nonopioid analgesia if they had never had surgery before (OR 0.21, p = 0.010, 95% CI [0.06–0.69]). Additionally, for every unit increase in VAS score, the odds of preferring an opioid pain medication increased 1.41 times (OR 1.41, p < 0.001, 95% CI [1.15–1.72]). No association was found between opioid preference and ORT or PROMIS scores. Odds ratios for each covariate analyzed can be referenced in Table 5.

4 Discussion

The present study demonstrated several important patient characteristics and concerns associated with a preference for opioids after common orthopaedic sports procedures. Most notably, patients with a prior surgical history and previous opioid consumption for postoperative pain control were more likely to request opioid analgesia. Patients with concern for postoperative pain and those with higher VAS scores preoperatively preferred postoperative opioid analgesia. We also found that nonopioid analgesia was preferred by the cohort of patients who had no past surgical history, had successfully undergone previous surgery without the use of opioids, and who had reported concerns regarding addiction and side effects of the medications. Given the known risks associated with the use of opioid medications, providers have an important opportunity to practice opioid stewardship and educate patients about the risks of opioids and available alternatives. The findings of this study provide additional insight into patient factors and patient preferences about analgesia choice to help guide orthopaedic surgeon's preoperative decision making with regards to pain control.

Preoperative opioid use has been strongly correlated with continued opioid consumption after a variety of orthopaedic procedures. ^{11,31,33,34,38–42} Westermann et al. retrospectively reviewed 35,155 patients after rotator cuff repair and found that patients filling opioid prescriptions at one to three months preoperatively were 7.45 times more likely to still be filling prescriptions for opioids three months after surgery (95% CI, 6.95–7.98).²⁸ Patients filling opioid prescriptions at any time point before surgery were more likely to continue filling these medications for a prolonged period of time compared to those who were opioid naïve. Jildeh et al. similarly found that preoperative opioid use significantly increased the risk for prolonged refills up to 12 months after arthroscopic meniscal surgery in a patient population excluding those with opioid dependency.³² The chronicity of opioid use was significantly correlated with the number of refills, with 6.6±6.8 refills in chronic users, 3.8±4.9 in acute users, and 1.7±1.9 in nonusers. In a similar study by the same study authors, both acute and chronic preoperative opioid use was correlated with continued postoperative opioid use at greater than two months after arthroscopic shoulder labral repair.³³ While our study did not quantitatively analyze patient opioid consumption preoperatively, we did find that patients who have had previous surgeries with opioids were more likely to request an opioid pain protocol for a future surgery. This phenomenon may be explained by patients’ previous success with narcotic pain control and hesitancy to try a new protocol. These patients harbor a higher risk for prolonged narcotic consumption after surgery, and thus would likely benefit from additional education about weaning opioid pills postoperatively and preoperative discussion of alternative multimodal pain regimen options.

The evidence supporting multimodal and narcotic-free postoperative pain management continues to grow. Kamdar et al. followed 100 patients in the postoperative period following knee arthroscopy who received only 5 tablets of 5 mg oxycodone to supplement non-opioid pain medications.⁴³ They found that 90% of patients consumed fewer than 5 opioid tablets following the surgery and over half consumed no opioid pills whatsoever. However, despite this success, certain patients still exist who subjectively experience more pain than others and require a higher narcotic burden for equivalent pain control. Higher patient-reported pain has been shown to correlate with increased opioid use.¹⁹ Theosomy et al. found that patients who required oxycodone after arthroscopic RCR reported higher overall pain scores compared to those who utilized opioid-free management.⁴⁴ Our study revealed similar findings in the preoperative period with regards to patient preference for opioids. For every 1-point increase in preoperative VAS, patients were 1.41 times more likely to prefer an opioid pain protocol (OR 1.41, p < 0.001, 95% CI [1.15–1.72]). These results suggest that those patients experiencing higher preoperative pain are already more inclined to choose a narcotic pain management protocol, even before the surgery has occurred. Pain perception varies considerably between patients with similar pathology. In addition, patients preferring opioid regimens in our study were more often concerned about pill burden associated with the nonopioid regimen. Patients with concerns of taking too many pills will be an important demographic to identify preoperatively in order to provide additional education about the benefits and safety of multimodal analgesia. Surgeons must be mindful that the preoperative visit is an important timepoint to provide education and counseling to discuss patient expectations, educate about the value of multimodal pain management, and address concerns about pain throughout the postoperative course. Future prospective studies would be helpful to provide granularity in terms of specific procedures and how patient perceptions of the surgery and their preoperative level of pain impact their preference for opioid prescriptions.

The association between continued opioid use and a history of psychiatric disorders and substance use has been demonstrated among a variety of patients following surgery for orthopaedic trauma injuries, ACL reconstruction, RCR, meniscectomy or repair, and shoulder labral repair. ^{19,25,26,28,31,35,40–42,45} Sun et al. identified preoperative history of drug abuse, alcohol abuse, depression, benzodiazepine use, and antidepressant use to be associated with chronic opioid use in previously-naïve patients undergoing a variety of orthopaedic and non-orthopaedic procedures.²⁷ Similarly, in an analysis of 14,830 patients who underwent hip arthroscopy, a significantly larger proportion of patients (5,208/14,830, 35.1%) with a preoperative diagnosis of depression or anxiety filled opioid prescriptions both in the six months before surgery and >60 days and >90 days following surgery.⁴⁶ The present study failed to show an association between psychiatric diagnoses or a personal history of substance use and a preference for opiate pain management. It is possible that there is a discrepancy between patient preferences before surgery and actual analgesic consumption postoperatively in this cohort of patients. Although no association between psychiatric history or past history of addiction and a preference for opioids, patients who demonstrated a concern for addiction were significantly less likely to prefer opioids for pain control (p < 0.001). We cannot comment whether those patients eventually chose a nonopioid regimen or not. While patients may have good intentions to utilize only a multimodal regimen, those with psychiatric diagnoses are at inherently higher risk of postoperative consumption, and it is reasonable to believe that patients’ preoperative preference may not have always correlated with their postoperative consumption.

5 Conclusions

Patients with a history of prior surgery utilizing opioid pain control, higher VAS scores preoperatively, and concern for inadequately managed postoperative pain were more likely to prefer opioid pain control following common orthopaedic sports procedures. Patients may benefit from increased preoperative education about opioid risks and the role of multimodal pain management regimens in reducing opioid consumption after surgery. In addition, our study shows patient VAS scores to be the superior predictor to ORT or PROMIS scores and should be surgeons preferred method for pre-operative assessment of opioid preference and use.

Funding/sponsorship

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Informed consent

Not applicable to this study (not a case report, no clinical images).

Institutional ethical committee approval

The study received IRB approval prior to data collection and initiation of the study.

Authors contribution

Joshua P. Castle, MD: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Project administration; Validation; Visualization; Roles/Writing – original draft; Writing – review & editing Toufic R. Jildeh, MD: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Project administration; Validation; Visualization; Roles/Writing – original draft; Writing – review & editing Muhammad J. Abbas, MD: Formal analysis; Investigation; Methodology; Validation; Visualization; Roles/Writing – original draft; Writing – review & editing Mary E. Hennekes, MD: Data curation; Formal analysis; Investigation; Methodology; Validation; Visualization; Roles/Writing – original draft; Writing – review & editing Patrick J. Buckley, MD: Formal analysis; Investigation; Methodology; Validation; Visualization; Roles/Writing – original draft; Writing – review & editing Christina L. Shabet, BSE: Formal analysis; Visualization; Roles/Writing – original draft; Writing – review & editing Daniel Cotter, MD: Formal analysis; Investigation; Visualization; Roles/Writing – original draft; Writing – review & editing Vasilios Moutzouros, MD: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Project administration; Resources; Supervision; Validation; Roles/Writing – original draft; Writing – review & editing.