Initial Opioid Prescription Is Associated With and Predictive of Prolonged Postoperative Opioid Use in Hand Surgery Patients.

Purpose: Our purpose was to determine whether the quantity of initial opioid prescriptions combined with routinely collected clinical factors and patient-reported data (PRD) can be used to predict prolonged opioid use after hand surgery and to generate a presurgical prediction model that can be tested for use every day. Methods: We performed a retrospective analysis of 12,117 adults who underwent hand surgery at a single, large academic hand center from 2018 to 2022. Opioid prescription data were obtained from electronic medical records, and patients were categorized into high/low initial opioid prescription groups based on quantile regression-adjusted total morphine milligram equivalents (MMEs). Multivariable logistic regression was performed to predict postoperative opioid use at 3 months, incorporating demographic, clinical, and PRD variables, including high versus low initial prescription status from the quantile model. Stepwise logistic regression across 15 imputed data sets generated pooled odds ratios and 95% confidence intervals. Model performance was assessed using receiver operating characteristic and precision-recall curves. Results: Patients receiving adjusted high initial opioid doses had significantly greater odds of continued opioid use 3 months after surgery. Additional predictors included higher Charlson Comorbidity Index, greater preoperative pain, lower preoperative Patient-Reported Outcomes Measurement Information System Global Physical Health scores, opioid/marijuana/medication history, postoperative antibiotic use, minority racial background, and Medicare/Medicaid insurance. Predictive model performance was moderate, with a 3-month precision-recall curve area under the curve of 0.135 in the training set and 0.157 in the test set. Conclusions: Combining adjusted postoperative prescription amount and PRD with routinely captured electronic health record variables yields a predictive algorithm that accurately flags hand surgery patients at risk for prolonged postoperative use. If prospectively validated, embedding this tool into clinical workflows may enable targeted counseling, opioid prescribing guidance, proactive multimodal analgesia, and overall safer, data-driven opioid stewardship. Type of study/level of evidence: Prognostic IIb.