Magnetic resonance imaging of the prostate is increasingly being performed at academic centers but implementation in community based health systems has lagged and literature regarding clinical impact in this setting is limited. We describe our experience developing a community based prostate magnetic resonance imaging program, including the evolution of interpretation and reporting methods, and the resulting clinical impact during a period of more than 5 years (August 2010 to December 2015).Methods:
Data collected for prostate magnetic resonance imaging included demographic, clinical, scanning, pathology and treatment/management information. Suspicion level on prostate magnetic resonance imaging was correlated with pathology results when available. Outcomes were compared across 3 reporting eras, ie early, mid and Prostate Imaging Reporting and Data System, version 2.Results:
A total of 537 prostate magnetic resonance images were obtained for diagnosed prostate cancer (60%) or screening (37%). During the study period the number of scans and ordering physicians increased. The proportion of patients with suspected extraprostatic extension (17.5%), lymph node metastasis (6.9%) and bone/other metastasis (4.3%) on prostate magnetic resonance imaging remained relatively constant. When stratified by era, there was a significant increase in low suspicion studies (p = 0.0002) and a trend toward a significant increase in cancer detection at biopsy (p = 0.09), reflecting increased specificity in the Prostate Imaging Reporting and Data System, version 2 era.Conclusions:
While staging information with prostate magnetic resonance imaging was accurate early in the implementation of the program, lesion characterization improved with use of Prostate Imaging Reporting and Data System, version 2 criteria and standardized reporting. Regular multidisciplinary participation in community based prostate magnetic resonance imaging programs may maximize clinical impact.