Social Risk Adjustment In the Hospital Readmissions Reduction Program
I am a semi-retired medical writer and consultant with a long-standing desire to apply the systematic review skills
I developed conducting health technology assessments to health delivery and health policy questions.
The systematic review that colleagues and I published in the September issue of Health Affairs shows the value of this form of secondary research in health policy.
I am very grateful for the encouragement and initial guidance provided by Rajender Agarwal (see prior examples of systematic reviews led by Agarwal and published by Health Affairs: Medicare Advantage versus Traditional Medicare and the impact of bundled payments) and to the four colleagues
who joined me in this unfunded but labor-intensive project.
Our team wanted to explore social risk adjustment in the context of value-based payment.
Policy experts increasingly acknowledge the research showing that factors such as low income, food insecurity and inadequate housing are associated with poorer health outcomes, independent of conventional risk adjustment measures.
But there is considerable controversy over how and whether provider payment methods should take this association into account.
The evidence we reviewed suggests that social risk
adjustment achieves more equitable provider reimbursement without unintended consequences.
We focused on the CMS Hospital Readmissions Reduction Program (HRRP).
Under the HRRP, a hospital receives an annual penalty (reduced reimbursement) if its adjusted thirty-day readmission rates for any of a small set of admission conditions are below the average for hospitals with a similar patient mix.
Current HRRP methodology compares hospitals against only those hospitals in the same quintile defined by proportion of patients dually eligible for Medicaid and Medicare. This peer grouping methodology retains the original CMS risk adjustment model.
Our objective was to synthesize the evidence concerning how HRRP peer grouping and other approaches to social risk adjustment affect readmission performance measures and financial penalties for hospitals subject to the HRRP.
The controversy over social risk adjustment stems from two competing goals.
One goal is fairness in provider reimbursement. It stands to reason that hospitals with a relatively large proportion of socially vulnerable patients (safety net hospitals) would find it more difficult to achieve good health outcomes for their patient populations.
These hospitals are also usually characterized by fewer financial resources and thus are limited in their ability to compensate for social risks through enhanced care delivery.
Data collected soon after the HRRP went into effect, and prior to the peer grouping approach, showed that safety-net hospitals received disproportionately larger penalties.
The other goal contributing to the controversy is continuous improvement in patient care. It is not clear whether the poorer outcomes achieved by safety net hospitals are primarily due to the characteristics of the patients served or to a pattern of lower-quality care delivery in safety net hospitals.
Social risk adjustment could have the unintended effect of “adjusting away” poor quality care. So, in our review we also looked for evidence that might speak to this question.
Our team identified fourteen studies that served our objective. The selected studies investigated one or more of three general approaches: addition of social risk variables to the CMS risk adjustment model, which we referred to as model augmentation; peer grouping
(the current approach used by HRRP); or a combination of model augmentation and peer grouping.
In ten of the fourteen selected studies social risk adjustment reduced adjusted readmission rates and/or penalties for safety net hospitals and increased those measures or made no difference for hospitals serving populations with the smallest proportion of socially vulnerable patients.
Although study heterogeneity clinical populations and measurement methods prevented a quantitative summary of outcomes across studies, the general consistency in the direction of findings despite these methodological inconsistencies strengthens the general conclusion of our review.
An especially interesting discovery was evidence that the differences in readmissions performance between safety-net hospitals and other hospitals was primarily attributable to
differences in patient characteristics rather than differences in the quality of care or other hospital characteristics.
Three of four studies that addressed this question came to this conclusion.
Two studies provided particularly compelling evidence by using fixed effect hospital variables, which allowed any potential systematic differences in quality of care to influence adjusted readmission rates while still adjusting for patient-level social risk and other characteristics.
The evidence regarding patient versus hospital effects speaks to the skepticism often voiced about the feasibility of expecting providers to satisfy health-related social needs.
The evidence we
reviewed suggests that the full answer to healthcare inequities goes beyond provider responsibilities.
Some have argued that even though adjustments to payment models or subsidies may be warranted for providers who disproportionately serve more socially at-risk patients, public reporting should remain transparent—that payers and patients need to know which hospitals actually have higher readmission rates after adjusting only for conventional risk factors.
However, a small body of research suggests that differences in outcomes are influenced primarily by patient characteristics (clinical and social) rather than differences in quality of care, at least in the HRRP.
If this is generally the case, are performance measures that remain unadjusted for a key set of patient characteristics fully transparent?
Would a failure to adjust for social risk lead to the false conclusion that better health outcomes for socially at-risk patients could be
achieved if the patients just went to a different provider?
We trust our work underscores the need to consider these questions as policy makers seek to foster efforts to address social determinants of health.
If you enjoyed reading today's newsletter, you might also like our Social Determinants of Health newsletter, which is available to Health Affairs Insiders. Use discount code HAInsider10 to get $10 off if you sign up before October 5.
Katie Keith discusses the recent ruling in the Braidwood Management v. Becerra lawsuit in which Judge Reed O’Connor held that a key part of the preventive services mandate under the
ACA violates the Appointments Clause and is thus unconstitutional.
Michael Barr and coauthors discuss comprehensive medication management, a systematic approach to ensure that medications are individually assessed for appropriateness, safety, efficacy, and correctness.
David
Introcaso discusses how CMS could use reimbursement to reduce greenhouse gas emissions, measurably lowering Medicare and Medicaid beneficiaries' burden of disease, improving their care outcomes and safety, advancing health equity, addressing the industry’s financial sustainability, and helping heal the planet.
State Of The Health Care Workforce
Listen to Health Affairs' Ellen Bayer and Rob Lott review the state of the health care workforce in this Labor Day themed episode of Health Affairs This Week.
Health Affairs is the leading peer-reviewedjournalat the intersection of health, health care, and policy. Published monthly by Project HOPE, the journal is available in print and online. Late-breaking
content is also found through healthaffairs.org, Health Affairs Today, and Health Affairs Sunday Update.
Project HOPE is a global health and humanitarian relief organization that places power in the hands of local health care workers to save lives across the globe. Project HOPE has published Health Affairs since 1981.
