Patient Outcomes and Health Care Use in U.S. Hospices After Private Equity Acquisition

About This Grant

Hospice care is increasingly important for 67 million Medicare beneficiaries in the U.S.—accounting for a quarter of Medicare spending. At the same time, hospices across the U.S. are increasingly acquired by private equity (PE) firms. Over the last 20 years, half of U.S. hospices have turned from non-profit to for-profit status, with PE driving the recent increase. To date, the clinical and economic effects of PE acquisitions remain poorly understood—both on average and across subgroups. Recent evidence on PE from other parts of the delivery system including hospitals, nursing homes, and physician practices show that PE acquisition is associated with changes in patient selection, increased adverse events (hospitals) and mortality (nursing homes), along with increased charges, prices, and spending. To date, evidence of its impact in hospice remains sparse. Rigorous evidence on PE's effects on hospice patient selection, utilization, and outcomes is urgently needed. PE firms have several strategies to maximize revenues from hospice facilities, given that hospice payment by Medicare (the main payer for hospice) is a per diem (day), non-risk-adjusted fee. We hypothesize that PE acquisition leads to selection of lower-risk or healthier Medicare patients, who are more likely to exhibit longer lengths of stay on hospice while using less care (fewer visits and services) on average. We hypothesize that, on average, PE acquisition leads to an increase in the live discharge rate, shifts toward lower-cost modalities of hospice (e.g. home hospice rather than facility-based), and increased Medicare spending. Via a hypothesis- driven approach, this project focuses on these key mechanisms by which PE firms may alter hospice behavior, using a novel dataset of hospice PE acquisitions linked to 100% CMS Medicare data spanning 2016-2018. To better approach causal inference, we use a quasi-experimental design that incorporates more novel statistical innovations in matching and weighting to derive two distinct, but related, counterfactuals of non-PE hospices: hospices that are non-profit and hospices that are for-profit but not PE-owned. Using rigorous methods, we will also examine treatment effect heterogeneity by patient and hospice characteristics. Understanding the mechanisms by which PE affects U.S. hospice care could inform policies that mitigate undesirable consequences for patients, including access to hospice, and help protect the solvency of Medicare going forward. Findings will contribute to a nascent scientific literature on PE's effects on hospice, which is key for the aging U.S. population, and will continue to be policy relevant as federal and state governments consider ways to protect patients and clinicians from the undesirable implications of PE. Given the labor-intensive effort required to collect and verify data on PE acquisitions, we will also make publicly available a novel database of PE-acquired hospices in the U.S. Our research aims support the NIA strategic objectives for research and the aging population, including improving the health and wellbeing of older adults (Goals C, E, F) and using evidence to inform stewardship of public resources (Goals H, I).