The Effect of Prospective Payment on Admission and Treatment Policy: Evidence from Inpatient Rehabilitation Facilities

Working Paper: NBER ID: w17125

Authors: Neeraj Sood; Peter J. Huckfeldt; David C. Grabowski; Joseph P. Newhouse; Jos J. Escarce

Abstract: We examine provider responses to the Medicare inpatient rehabilitation facility (IRF) prospective payment system (PPS), which simultaneously reduced marginal reimbursement and increased average reimbursement. IRFs could respond to the PPS by changing the total number of patients admitted, admitting different types of patients, or changing the intensity of care for admitted patients. We use Medicare claims data to separately estimate each type of provider response to the PPS. We also examine changes in patient outcomes and spillover effects on other post acute care providers. We find that costs of care initially fell following the PPS implementation, which we attribute to changes in treatment decisions rather than the types of patients admitted to IRFs. However, the probability of admission to IRFs increased after the PPS due to the expanded admission policies of providers. We find modest spillover effects on skilled nursing home costs and no substantive impact on patient health outcomes.

Keywords: Medicare; Inpatient Rehabilitation Facilities; Prospective Payment System; Provider Behavior

JEL Codes: H42; H51; I11; I18

Causal Claims

Cause	Effect
Implementation of the IRF PPS (H55)	Increase in payments to IRFs (I23)
Increase in payments to IRFs (I23)	Changes in admission policies (I24)
Changes in admission policies (I24)	Increase in Medicare spending (H51)
Implementation of the IRF PPS (H55)	Shift in treatment policies (E65)
Shift in treatment policies (E65)	Decrease in intensity of care provided (I11)
Decrease in intensity of care provided (I11)	No significant impact on health outcomes (I14)
Implementation of the IRF PPS (H55)	Reduced resource use (Q20)
Changes in admission policies (I24)	Increased costs in other postacute care settings (I11)