Population based analysis of incidence and predictors of open conversion during minimally invasive radical prostatectomy.

ABSTRACT

PURPOSE: We used population based data to measure the rates and risk factors of open conversion during minimally invasive radical prostatectomy in the United States. MATERIALS AND METHODS: We retrospectively analyzed the records of 87,415 patients in the NCDB who underwent minimally invasive radical prostatectomy between 2010 and 2011. We compared surgical outcomes and treatment facility characteristics between converted and nonconverted cases. Multivariable analysis was done to evaluate conversion risk factors. RESULTS: There were 82,338 robot-assisted (94%) and 5,077 laparoscopic (6%) radical prostatectomies, and 1,080 conversions (1.2%). Fewer robot-assisted cases were converted than laparoscopic cases (0.9% vs 6.5%, p <0.001). The median yearly treatment facility volume of minimally invasive radical prostatectomy was 32 (IQR 10-72). Patients who underwent conversion were more likely to be rehospitalized within 30 days (4.4% vs 2.7%, p = 0.002) and have a postoperative hospital stay of greater than 2 days (40.4% vs 15.1%, p <0.001) than those without conversion. Facilities in the lowest quartile of the yearly volume of the minimally invasive procedure represented 3.8% of minimally invasive radical prostatectomies but accounted for 22.9% of conversions. The second, third and fourth quartiles of yearly treatment facility minimally invasive volume predicted a lower likelihood of conversion compared to the first quartile (each p <0.001). Facility type (eg academic or community) did not predict conversion. Black race (vs white OR 1.52, 95% CI 1.24-1.86, p <0.001) and laparoscopic radical prostatectomy (OR 4.68, 95% CI 3.79-5.78, p <0.001) predicted higher odds of conversion. CONCLUSIONS: Open conversion during minimally invasive radical prostatectomy is a rare event. However, it is significantly more likely for pure laparoscopic surgery, in black men and at low volume facilities. Facility type did not affect conversion rates.