Toward Antiracist Reimbursement Policy in End-Stage Kidney Disease: From Equality to Equity

The year 2020 saw ongoing police violence against Black people; their disproportionate burden of coronavirus disease 2019 morbidity and mortality; and subsequent national protests. In the context of calls for racial justice, the kidney community is grappling with how its established policies and practices perpetuate disparities in kidney disease. Currently, the inclusion of race in kidney function estimation is being reassessed. The decade-old ESKD value-based program (VBP) also deserves new scrutiny. Antiracism, a paradigm characterized by the active identification and dismantling of the root causes of racial disparities, provides a lens for reimbursement policy evaluation.

The Centers for Medicare & Medicaid Services (CMS) VBP for ESKD largely ignores race. CMS holds a facility in Portland to the same quality benchmarks as one in Detroit. Although this parity seems ethical on its face, large studies demonstrate that VBP disproportionately penalizes dialysis facilities in predominantly Black (hereafter, “residentially segregated”) communities and therefore, treats facilities equally but inequitably.^1–3 Antiracist ESKD reimbursement policies can address this inequity in thoughtful ways right now. This perspective describes the scope and implications of inequity in CMS VBP for dialysis facilities and offers solutions that can be readily implemented.

Racial Inequity in VBP for ESKD

In 2012, CMS implemented its first VBP, the ESRD Quality Incentive Program (QIP). Via QIP, CMS can reduce prospective payments to a dialysis facility by up to 2% on the basis of its performance or total performance score (TPS) on quality indicators relative to national benchmarks.⁴ Proponents of VBPs argue that data feedback and financial penalty should motivate dialysis facilities to improve performance. However, one concerning potential consequence of VBPs is that they may penalize dialysis facilities serving patients at greater risk of poor outcomes. CMS currently adjusts for patient-level characteristics (e.g., comorbidities) that might influence performance on standardized measures. However, they do not account for the racial composition of a facility or that of the neighborhood surrounding the dialysis facility, despite established associations between these factors, patient outcomes, and dialysis facility quality scores.

Large observational studies controlling for important covariates have demonstrated associations between the proportion of Black patients in the dialysis facility, the proportion of Black residents in the neighborhood surrounding the dialysis facility, and dialysis facility TPS and/or QIP payment deductions.^1–3 For example, Qi et al.² found that dialysis facilities with >60% Black patients had 64% higher odds of payment deduction through QIP after adjusting for neighborhood poverty, dual eligibility, facility characteristics, and geography. Saunders et al.¹ demonstrated that dialysis facilities in neighborhoods with >25% Black residents had a 25% increased odds of payment deduction through QIP after adjusting for facility characteristics and neighborhood poverty but not facility racial composition.

These findings seem readily interpretable, as Black patients are less likely than White patients to achieve hemoglobin targets, receive dialysis through a fistula, or receive a kidney transplant—disparities relevant to quality indicators in QIP.⁵ However, recent work has begun to disentangle associations between race, place, and poverty by controlling for dialysis facility racial composition, neighborhood racial composition, and poverty simultaneously. It has elucidated the central importance of residential segregation to patient outcomes and subsequently, dialysis facility performance on quality indicators. In a study of nearly 5000 patients and over 150 dialysis facilities, Golestaneh et al.⁶ found that patients receiving hemodialysis in neighborhoods composed of over 14% Black residents were hospitalized at a 32% higher rate than patients receiving dialysis in neighborhoods where <2% of residents were of Black race. This association held after adjusting for patient-level sex, race, age, and comorbidities; clinical indicators; primary payer; facility characteristics; and neighborhood poverty.⁶ That is, even comparable White patients receiving comparable dialysis treatment in comparable facilities had worse outcomes when receiving dialysis in a predominantly Black neighborhood.

These findings indicate that structural inequities rather than inherent differences between racial groups explain the association between residential segregation and QIP scores. Dialysis facilities in segregated neighborhoods may perform well on QIP measures affected by the dialysis treatment itself but are not designed to affect upstream factors (i.e., social determinants of health). Hall et al.⁷ demonstrated that minority-serving dialysis facilities met targets for dialysis adequacy and hemoglobin levels, but a greater proportion of them had worse than expected patient survival compared with facilities with predominantly White patients (19% versus 3%). Residential segregation and subsequent neighborhood disinvestment increase risk for patients living in such neighborhoods to experience food insecurity, housing instability, and limited pre-ESKD nephrology care. DaVita Inc. leaders recently documented linear associations between many of these factors and Dialysis Facility Compare Star Program scores, a composite score like QIP TPS.⁸ Additional research is needed to identify other relevant contextual factors.

Implications of a reimbursement policy disproportionately penalizing dialysis facilities caring for Black patients in Black communities have not been fully investigated. Patients may experience stigma for “bringing down quality scores.” Quality improvement initiatives targeting patient behavior or dialysis facility processes but not addressing structural inequity may be ineffective. Although QIP payment penalty is relatively small, financial losses may require some facilities to close, limiting access for vulnerable patients. Moreover, facilities in these communities may benefit the most from additional financial resources (not less).

Recommendations for an Antiracist ESKD Reimbursement Policy

The National Academies of Science, Engineering and Medicine has offered strategies to account for race and other social risks in Medicare’s VBPs: (1) stratified public reporting, (2) adjustment of performance measurement scores, (3) direct adjustment of payment, and (4) restructuring of payment incentive design.⁹ Strategies 3 and 4 are most promising for making QIP payment reductions more equitable. The utility of stratified public reporting may be limited if racial disparities in quality scores exist across facilities but not within the same facility. Adjusting QIP performance scores for race or other social determinants of health is not ideal because disparities in quality performance data would “disappear,” exposing patient subgroups to persistently poorer-quality care.

CMS can implement antiracist reimbursement policy via direct adjustment of payment in three ways. (1) They can immediately pause QIP payment penalties to dialysis facilities in residentially segregated neighborhoods. CMS has a process for granting Extraordinary Circumstances Exceptions that temporarily lift QIP data-reporting requirements (e.g., during natural disasters). They can create a similar process to adjudicate payment adjustments to dialysis facilities affected by the social “disasters” of residential segregation and neighborhood disinvestment. (2) CMS can stratify dialysis facilities by neighborhood racial composition and use different QIP TPS thresholds (e.g., median TPS in each strata) to guide payment penalties. They currently stratify hospitals by proportion of dually eligible patients in their Hospital Readmission Reduction Program. (3) Alternatively, CMS can increase the prospective payment base rate to dialysis facilities caring for patients with a high burden of social risk factors for care coordination or other services. CMS can provide guidance on standard measures for neighborhood-level social risk factors that are valid and minimize the data-reporting burden on dialysis facility staff. CMS can also consider restructuring the payment incentive design. They can conduct new research to identify QIP metrics most sensitive to social determinants of health and down weight them in penalty calculations.

We are heartened by CMS’s proposed rule to address health equity in the ESRD Treatment Choices Model by considering dual eligibility in scoring methodologies.¹⁰ We urge CMS to consider the role of residential segregation in QIP. Nearly 10 years after QIP implementation, it is time to stop treating dialysis facilities as if the communities they serve are all the same.

Disclosures

D.C. Crews reports consultancy agreements with the Yale New Haven Health Services Corporation Center for Outcomes Research and Evaluation; research funding from Somatus, Inc.; scientific advisor or membership via editorial boards of Clinical Journal of the American Society of Nephrology, Journal of the American Society of Nephrology, and Journal of Renal Nutrition; scientific advisor or membership via associate editor for Kidney360; scientific advisor or membership via cochair of Bayer HealthCare Pharmaceuticals Inc. and a member of the Patient and Physician Advisory Board Steering Committee for the Disparities in Chronic Kidney Disease Project; and other interests/relationships via the Nephrology Board of the American Board of Internal Medicine, a council member of Subspecialist Societies for the American College of Physicians, and the board of directors for the National Kidney Foundation of Maryland/Delaware. K. Taylor was the Corporate Vice President of Quality for Fresenius Kidney Care from September 2016 through December 2017.

Funding

D.C. Crews was supported by National Heart, Lung, and Blood Institute grant K24HL148181. K. Taylor was supported by National Institute of Nursing Research grant F31NR109461.