Confounders of mortality and hospitalization rate calculations for profit and nonprofit dialysis facilities: analytic augmentation (BMC Nephrology)

Background: Patient outcomes have been compared on the basis of the profit status of the dialysis provider (for-profit [FP] and not-for-profit [NFP]). In its annual report, United States Renal Data System (USRDS) provides dialysis provider level death and hospitalization rates adjusted by age, race, sex, and dialysis vintage; however, recent analyses have suggested that other variables impact these outcomes. Our current analysis of hospitalization and mortality rates of hemodialysis patients included adjustments for those used by the USRDS plus other potential confounders: facility geography (end-stage renal disease network), length of facility ownership, vascular access at first dialysis session, and pre-dialysis nephrology care. Methods: We performed a provider level, retrospective analysis of 2010 hospitalization and mortality rates among US hemodialysis patients exclusively using USRDS sources. Crude and adjusted incidence rate ratios (IRRs) were calculated using the 4 standard USRDS patient factors plus the 4 potential confounders noted above. Results: The analysis included 366,011 and 34,029 patients treated at FP and NFP facilities, respectively. There were statistical differences between the cohorts in geography, facility length of ownership, vascular access, and pre-dialysis nephrology care (p < 0.001), as well as age (p < 0.01), race (p < 0.001), and vintage (p < 0.001), but not sex (p = 0.12). When using standard USRDS adjustments, hospitalization and mortality rates for FP and NFP facilities were most disparate, favoring the NFP facilities. Rates were most similar between providers when adjustments were made for each of the 8 factors. With the FP IRR as the referent (1.0), the hospitalization IRR for NFP facilities was 1.00 (95% confidence interval [CI] 0.97-1.02; p = 0.69), while the NFP mortality IRR was 1.01 (95% CI 0.97-1.05; p = 0.64). Conclusions: These data suggest there is no difference in mortality and hospitalization rates between FP and NFP dialysis clinics when appropriate statistical adjustments are made.