Variability in Partial Nephrectomy Outcomes: Does Your Surgeon Matter?

Background: Understanding physician-level discrepancies is increasingly a target of US healthcare reform for the delivery of quality-focused patient care. Objective: To estimate the relative contributions of patient and surgeon characteristics to the variability in key outcomes after partial nephrectomy (PN). Design, setting, and participants: Retrospective review of 1461 patients undergoing PN performed by 19 surgeons between 2011 and 2016 at a tertiary care referral center. Intervention: PN for a renal mass. Outcomes measurements and statistical analysis: Hierarchical linear and logistic regression models were built to determine the percentage variability contributed by fixed patient and surgeon factors on peri- and postoperative outcomes. Residual between- and within-surgeon variability was calculated while adjusting for fixed factors. Results and limitations: On null hierarchical models, there was significant between-surgeon variability in operative time, estimated blood loss (EBL), ischemia time, excisional volume loss, length of stay, positive margins, Clavien complications, and 30-d readmission rate (all p<0.001), but not chronic kidney disease upstaging (p=0.47) or percentage preservation of glomerular filtration rate (p=0.49). Patient factors explained 82% of the variability in excisional volume loss and 0-32% of the variability in the remainder of outcomes. Quantifiable surgeon factors explained modest amounts (10-40%) of variability in intraoperative outcomes, and noteworthy amounts of variability (90-100%) in margin rates and patient morbidity outcomes. Immeasurable surgeon factors explained the residual variability in operative time (27%), EBL (6%), and ischemia time (31%). Conclusions: There is significant between-surgeon variability in outcomes after PN, even after adjusting for patient characteristics. While renal functional outcomes are consistent across surgeons, measured and unmeasured surgeon factors account for 18-100% of variability of the remaining peri- and postoperative variables. With the increasing utilization of value-based medicine, this has important implications for the goal of optimizing patient care.