Positive surgical margins at radical prostatectomy in the United States: Institutional variations and predictive factors

Positive surgical margins (PSMs) are associated with treatment failure after radical prostatectomy (RP) for patients with prostate cancer (CaP). We investigated institutional variations in PSM after RP, as well as clinical and demographic factors predicting PSM. Patients undergoing RP for clinically localized CaP were identified in the National Cancer Database in 2010 to 2013 and clinicodemographics were recorded. Treating institution was defined as academic (AMC) or nonacademic medical centers (nAMC). The primary outcome was the PSM rate. Multivariable logistic regression and propensity matching with inverse probability treatment weighing were used to both compare outcomes between AMC and nAMC and to identify predictors of PSM following RP. A total of 167,260 patients met our inclusion criteria. PSM rate was significantly lower in patients treated at AMC (13,435, 18.9%) compared with 22,145 (23.0%) in those treated at nAMC (P < 0.01). The difference between PSM rate in AMC and nAMC was more pronounced in lower volume centers while it was not significant in higher volume centers. On multivariable analysis, age, race, prostate-specific antigen (PSA), biopsy Gleason score, comorbidity profile, insurance type, income, and treatment facility were significantly associated with PSM rate. PSM rates appear to be lower at AMC and higher volume facilities, which can potentially reflect institutional differences in surgical quality. In addition, we identified several socioeconomic and demographic factors that contribute to the likelihood of PSM following RP for localized CaP, suggesting potential systematic variation in the quality of surgical care. The cause of this variation warrants further investigation and evaluation.