Urologist-Level Variation in Androgen Deprivation Therapy Intensification in Patients with Metastatic Hormone-Sensitive Prostate Cancer: A Retrospective, Population-Based Cohort Study in Ontario, Canada

Background: Guidelines recommend androgen-deprivation therapy (ADT) intensification for individuals with de novo metastatic hormone-sensitive prostate cancer (mHSPC). Urologists frequently initiate ADT and influence early treatment decisions, yet the extent to which variation in intensification reflects physician practice vs patient factors remains unclear. Methods: We identified patients with de novo mHSPC initiated on ADT by urologists in Ontario, Canada (2014-2022). For each urologist, we calculated the annual proportion of patients receiving treatment intensification, attributed to the initiating urologist, whether prescribed directly or following referral. Multilevel logistic regression models clustered at the physician level were used, and the variance partition coefficient (VPC) was used to estimate the proportion of variation attributable to physician effects. Caterpillar plots visualized physician-specific random intercepts. Results: The analytic cohort included 345 urologists treating 4, 191 patients. Overall, physicians intensified therapy for an average of 26.5% of their patients (median 21%, IQR 0-33%). Descriptive yearly analyses showed a steady rise in intensification: mean proportion per urologist increased from 6% in 2014 to 59% in 2022. In the complete cohort, physician-level clustering accounted for 4.6% of the variation in treatment intensification, with most of the variation attributable to patient characteristics. In a contemporary subcohort (2020-2022), physician clustering explained a slightly larger share (VPC 7.8-8.5%). There was wide heterogeneity in physician-specific intercepts, with most urologists clustering around the null effect but a long tail of higher-intensifying physicians. Conclusions: From 2014 to 2022, ADT intensification among patients initiated on therapy by urologists rose substantially. Most of the variation was explained by patient factors, although physician influence became more apparent in later years as early adopters drove uptake. These findings imply that accelerating equitable adoption of intensification may require system-level supports and targeted efforts to align later adopters with evolving standards.