Explore the vast range of titles available.

Background Opioid overuse is a costly and significant problem in the United States. Medical specialties including surgery are a contributor to opioid prescriptions while having few clear prescribing guidelines. Machine learning predictive tools can assist surgeons with evaluating patients for opioid prescriptions and help prevent prescriptions for patients at-risk for prolonged post-operative opioid use. This project aims to develop accurate and transparent machine learning models to predict prolonged opioid use by individuals undergoing rotator cuff surgery. Methods Six machine learning models were trained on a dataset of 852 individuals undergoing rotator cuff surgery and evaluated for predictive accuracy. Machine learning explainability techniques were used to improve model transparency, including Shapley Additive explanations (SHAP) for global variable importance and local interpretable model-agnostic explanations (LIME) for local variable effects and importance. Results Four of six machine learning models developed had predictive accuracy greater than 0.71, with our top three models having an accuracy of 0.98 (XGBoost), 0.94 (Random Forest), and 0.74 (Decision Tree). SHAP and LIME explanations were created for models allowing insight and interpretation into model outputs. Conclusion We were able to develop predictive machine learning models that displayed predictive accuracy in predicting prolonged post-operative opioid use while improving model explainability and transparency.

Purpose This study explored the performance of the Patient-Reported Outcomes Measurement Information System (PROMIS) Anxiety assessment relative to the Depression assessment in orthopedic patients, the relationship between Anxiety with self-reported Physical Function and Pain Interference, and to determine if Anxiety levels varied according to the location of orthopedic conditions. Methods This cross-sectional evaluation analyzed 14,962 consecutive adult new-patient visits to a tertiary orthopedic practice between 4/1/2016 and 12/31/2016. All patients completed PROMIS Anxiety, Depression, Physical Function, and Pain Interference computer adaptive tests (CATs) as routine clinical intake. Patients were grouped by the orthopedic service providing care and categorized as either affected with Anxiety if scoring > 62 based on linkage to the Generalized Anxiety Disorder-7 survey. Spearman correlations between the PROMIS scores were calculated. Bivariate statistics assessed differences in Anxiety and Depression scores between patients of different orthopedic services. Results 20% of patients scored above the threshold to be considered affected by Anxiety. PROMIS Anxiety scores demonstrated a stronger correlation than Depression scores with Physical Function and Pain Interference scores. Patients with spine conditions reported the highest median Anxiety scores and were more likely to exceed the Anxiety threshold than patients presenting to sports or upper extremity surgeons. Conclusions One in five new orthopedic patients reports Anxiety levels that may warrant intervention. This rate is heightened in patients needing spine care. Patient-reported Physical Function more strongly correlates with PROMIS Anxiety than Depression suggesting that the Anxiety CAT is a valuable addition to assess mental health among orthopedic patients. Level of Evidence Diagnostic level III.