Preoperative Patient-Reported Data Indicate the Risk of Prolonged Opioid Use After Hand and Upper Extremity Surgeries

Abstract

Opioids play an important role in pain management after surgery but also increase the risk of prolonged opioid use in patients. The identification of patients who are more likely to use opioids after intended short-term treatment is critical for employing alternative management approaches or targeted interventions for the prevention of opioid-related problems. We used patient-reported data (PRD) and electronic health record information to identify factors predictive of prolonged opioid use after surgery. We used our institutional registry containing data on all patients who underwent elective upper extremity surgeries. We evaluated factors associated with prolonged opioid use in the cohort from the year 2018 to 2019. We then validated our results using the 2020 cohort. The predictive variables included preoperative PRD and electronic health record data. Opioid use was determined based on patient reports and/or filled opioid prescriptions 3 months after surgery. We conducted bivariate regression, followed by multivariable regression analyses, and model validation using area under the receiver operating curve. We included 2,114 patients. In our final model on the 2018-2019 electronic health records and PRD data (n= 1,589), including numerous patient-reported outcome questionnaire scores, patients who were underweight and had undergone trauma-related surgery had higher odds of being on opioids at 3 months. Additionally, each 5-unit decrease in the preoperative Patient-Reported Outcomes Measurement Information System Global Physical Health score was associated with a 30% increased odds of being on opioids at 3 months. The area under the receiver operating curve of our model was 70.4%. On validation using data from the 2020 cohort, the area under the receiver operating curve was 60.3%. The Hosmer-Lemeshow test indicated a good fit. We found that preoperative questionnaire scores were associated with prolonged postoperative opioid use, independent of other variables. Furthermore, PRD may provide unique patient-level insights, alongside other factors, to improve our understanding of postsurgical pain management. Prognostic II.