Can Nursing Homes Predict How Profitable Post-Acute Patients Will Be? Effects on Patients' Choice of Provider and Outcomes

Brown University, Providence RI

Investigators

Abstract

PROJECT SUMMARY Nursing homes (NHs) provide care and residence to an estimated 1.4 million Americans, including some of the country?s most vulnerable populations: the frail elderly, disabled, and cognitively impaired. Despite numerous policy efforts to address quality, NH care remains a two-tiered system differentiated by patients? payer mix. NHs in the high-quality tier provide much Medicare-funded post-acute care (PAC) while lower tier NHs predominantly serve long-term residents financed by Medicaid. The differential between Medicare and Medicaid reimbursement rates creates an incentive to provide PAC. However, even among post-acute patients, individuals who are non-white or dually eligible for Medicaid are often admitted to the lowest quality NHs. The mechanism for why these vulnerable populations receive care in lower quality facilities are not fully known, however, financial incentives produced by PAC payment rules potentially play an important role. For example, Medicaid typically does not pay the cost-sharing requirements of the Medicare Skilled Nursing Facility (SNF) benefit. Additionally, vulnerable populations are characterized by higher levels of clinical and behavioral complexity. Under the existing payment scheme, Medicare SNF reimbursements are lowest for these types of conditions. Thus, the objective of this study is to examine whether current Medicare payment policies, which distinguish patients going to NHs based on the amount of SNF revenue they are expected to generate, create barriers to admission to high quality NHs for post-acute patients. Based on recent qualitative evidence showing that high-quality NHs increasingly solicit referrals for the most lucrative PAC patients, this study uses a novel strategy to emulate the process of screening patients according to the level of anticipated SNF reimbursement. The proposed set of analyses will (1) use an innovative NH choice model to measure the impact of patients? expected SNF revenues per day on the likelihood of admission to high-quality NHs following an acute hospitalization and (2) use an instrumental variables approach to determine whether high NH quality impacts the outcomes of patients who are expected to yield high and low levels of SNF revenues differently. The study hypotheses are that patients expected to yield higher SNF revenues are more likely to be admitted to high-quality NHs while patients expected to yield lower SNF revenues will benefit relatively more from high NH quality while being less likely to access it. Identifying barriers to accessing high-quality care is imperative because millions of older and disabled adults are at risk of poor outcomes when NH quality is low. The study?s results will also provide insight into whether efforts to improve NH quality or access to high quality NH care have the potential to improve patients? outcomes.