The Effect of Medicare’s Skilled Nursing Facility Benefit Design

Post-Acute Care (PAC) delivered in skilled nursing facilities (SNFs) is thought to be an important source of waste within the health system. As a result, a number of policy approaches are currently being implemented to create incentives to reduce SNF use. Little is known, however, about how much SNF care on the margin is wasteful or how patients and providers affect discharge decisions. This study exploits a discontinuity in the standard Medicare SNF benefit and variation in beneficiaries’ exposure to this coverage shift according to Medicaid and supplemental insurance status to assess both supply- and demand-side responses to a change in the price of SNF care. We further examine whether responses vary by a number of patient characteristics, including sociodemographic and health status variables, to assess the extent to which discontinuities in discharge practices are consistent with waste (i.e., longer than necessary SNF stays) or skimping (i.e., premature SNF discharges). Finally, we examine the effect of this benefit design on patient outcomes including mortality and hospitalization rates.

We find sizeable spikes in the probability of being discharged on the last day of full Medicare coverage, both when only the SNF is exposed to the price change and when the patient and the SNF are jointly exposed. Supply-side-only responses are concentrated in relatively healthier patients with no evidence of adverse patient outcomes associated with this spike in discharges, consistent with wasteful marginal SNF care use. In instances where the patient is responsible for paying the Medicare cost sharing amount, we do not find evidence that price-induced discharges are concentrated among healthier individuals. Instead, these discharges are disproportionately among potentially vulnerable populations, including racial/ethnic minorities and lower-income individuals, and are associated with a significant increase in the risk for subsequent hospitalizations, consistent with clinically inappropriate discharge decisions.

These results provide insight into the extent to which current SNF use is wasteful and the effects of current Medicare SNF benefit design. They also provide evidence about the potential effects of Medicare payment reforms aimed at reducing SNF lengths of stay. In particular, they shed light on the ability of PAC patients and providers to identify unnecessary SNF care, which can inform discussions about who should be financially at risk under alternate value-based payment models.