Understanding the Performance of Active Surveillance Selection Criteria in Diverse Urology Practices

Abstract

We used data from MUSIC (Michigan Urological Surgery Improvement Collaborative) to evaluate the performance of published selection criteria for active surveillance in diverse urology practice settings. For several active surveillance guidelines we calculated the proportion of men meeting each set of selection criteria who actually entered active surveillance, defined as the sensitivity of the guideline. After identifying the most sensitive guideline for the entire cohort we compared demographic and tumor characteristics between patients who met this guideline and entered active surveillance, and those who received initial definitive therapy. Of 4,882 men with newly diagnosed prostate cancer 18% underwent active surveillance. When applied to the entire cohort, the sensitivity of published guidelines ranged from 49% in Toronto to 62% at Johns Hopkins. At apractice level the sensitivity of Johns Hopkins criteria varied widely from 27% to 84% (p <0.001). Compared with men undergoing active surveillance, those meeting Johns Hopkins criteria who received definitive therapy were younger (p<0.001) and more likely to have a positive family history (p = 0.003), lower prostate specific antigen (p <0.001), a greater number of positive cores (2 vs 1) on biopsy (p <0.001) and a higher cancer volume in positive core(s) (p = 0.002). The sensitivity of published active surveillance selection criteria varies widely across diverse urology practices. Among patients meeting the most stringent criteria those who received initial definitive therapy had characteristics suggesting greater cancer risk, underscoring the nuanced clinical factors that influence treatment decisions.