Institutional Variation in MRI Utilization After Prostatectomy for Prostate Cancer: A Multicenter Study

Background/Objectives: Healthcare utilization may vary because of institutional practice patterns and policy context. This study aimed to evaluate institutional variation and clinical factors associated with repeated MRI examinations after prostatectomy and to explore whether cumulative MRI use was associated with overall survival. Methods: This retrospective multicenter study included 3754 patients who underwent prostatectomy for prostate cancer at seven hospitals. Repeated MRI examinations were analyzed using an Andersen-Gill counting-process Cox model with patient-clustered robust variance, and a complementary Poisson model was used to evaluate postoperative temporal patterns and the robustness of the institutional associations. Biochemical recurrence (BCR; postoperative prostate-specific antigen ≥ 0.2 ng/mL) was treated as a time-updated covariate. A separate exploratory Cox model evaluated overall survival. Results: During a median follow-up of 936 days (interquartile range, 432-1687 days), 2027 postoperative MRI examinations were recorded. Repeated postoperative MRI examinations were more frequent at Hospitals 2-7 than at Hospital 1 (average hazard ratios [HRs] over follow-up, 4.083-23.136; all p < 0.001), although the magnitude of these institutional associations varied over postoperative time (joint hospital-by-time interaction, p < 0.001). MRI examinations were also more frequent after BCR (HR, 1.565; 95% confidence interval [CI], 1.371-1.786; p < 0.001). Grade group and pathological T stage were not significantly associated with repeated MRI use. The annual MRI rate decreased from 18.08 per 100 person-years in postoperative year 5 to 5.94 in year 6. In the exploratory survival analysis, cumulative MRI use was not significantly associated with overall survival (HR, 0.782; 95% CI, 0.564-1.086; p = 0.142). Conclusions: MRI utilization after prostatectomy varied substantially across institutions and over postoperative time, with a marked decline after postoperative year 5. Although MRI examinations were more frequent after BCR, the large institutional differences were not fully explained by the measured clinical factors. These findings highlight the need for more standardized and evidence-based criteria for MRI use after prostatectomy.