Impact of alternative payment models on specialist care for patients with new-onset end-stage renal disease - Project Summary
Background: End-stage renal disease (ESRD), which affected >746,000 Americans in 2017, is the final stage
of chronic kidney disease, leading to death unless renal replacement therapy is initiated. Of the treatment
options available (in-center hemodialysis, home dialysis, kidney transplantation), transplantation is the
preferred modality, offering the best survival time, quality of life outcomes, and lowest average cost. Home
dialysis modalities are also less costly than in-center dialysis and associated with improved patient satisfaction.
Despite these advantages, <15% of incident ESRD patients received home dialysis or a living donor kidney
transplant in 2016, and only 14% of patients joined a transplant waiting list or received a transplant within one
year of initiating treatment. The Advancing American Kidney Health initiative (July 2019) proposes to enroll
50% of ESRD care providers in a shared-risk alternative payment model (APM) that would tie providers'
payment to cost savings on health service use, with the goal of increasing use of cost-effective treatments
(home dialysis and transplantation). Currently, no evidence exists about nephrologist participation in APMs, or
their impact on treatment outcomes in specialty care. This study will examine nephrologist participation in two
shared-risk APMs (accountable care organizations [ACOs] and ESRD seamless care organizations [ESCOs]),
to characterize the population of APM-affiliated nephrologists and estimate the impact of nephrologist APM
affiliation on treatment use and clinical outcomes for patients with new-onset ESRD. Data: This project will link
11 years (2006-2017) of United States Renal Disease System data—rich clinical and administrative data for
>1,250,000 incident ESRD patients and >7,500 dialysis facilities—to provider-level data on ACO (Medicare
Shared Savings Program Provider-level files) and ESCO participation (provider-level public data). Several
additional public data sets (e.g., US Census Bureau, CMS Hospital Compare) will supplement the provider-
level dataset with facility and community characteristics. Analysis: Aim 1 will test associations between APM
affiliation and nephrologists' key geographic, sociodemographic, and facility characteristics using logistic
regression models. Aim 2 will estimate the impact of nephrologist APM participation on treatment use and
ESRD patient health outcomes using rigorous quasi-experimental methods (difference-in-difference
regression), comparing nephrologists' treatment use and patients' outcomes before and after APM affiliation,
relative to contemporary changes in these outcomes among never-affiliated nephrologists during an 11-year
period of widespread APM adoption. Implications: APM participation will be mandatory for half of ESRD care
providers in 2020. This research will provide foundational information about nephrologists that have been
affiliated with APMs to date and about the impact of APM affiliation on treatment use and health outcomes for
patients with new-onset ESRD. Importantly, this study's findings will inform broader policy and health systems
deliberations about the design and implementation of APMs in specialty care settings.
