Shorter hospital stays have led to an expanded role for skilled nursing facilities (SNFs). The goal at SNFs is to restore recently hospitalized people*#8212;commonly those who are recovering from hip fracture, stroke, pneumonia, or heart failure—to their prior level of functioning. Under the old Medicare Prospective Payment System (PPS), SNFs were reimbursed for the cost of providing therapy without regard to the minutes per week provided to each resident.
Under the revised (1998) PPS, SNFs receive fairly generous payments for rehabilitation therapy. However, beyond a certain point (12 hours a week), additional therapy generates no additional payments. As a result, the percentage of residents of freestanding SNFs receiving extremely high levels of rehabilitation therapy dropped significantly, and the percentage receiving moderate levels increased.
Payment for rehabilitation therapy should be tied not just to the amount of therapy provided but also to clinical appropriateness, according to Chapin White, of the National Bureau of Economic Research in Cambridge. In a study supported by the Agency for Healthcare Research and Quality (National Research Service Award training grant T32 HS00020), he used Medicare administrative data to determine average SNF charges (for physical, occupational, and speech therapy) per hospital stay.
The average SNF rehabilitation charges per hospital stay dropped by 45 percent between 1997 and 2000, declining from $421 to $233. The most striking change occurred at for-profit freestanding SNFs. In 1997, 19 percent of residents at such facilities were receiving more than $200 per day in rehabilitation therapy; by 2000, this group had dropped to 1.6 percent. The timing of the drop corresponded precisely with the phasing in of the new PPS in 1998; the drop was not explained by a change in SNF length of stay, which was relatively stable during this period.
See "Rehabilitation therapy in skilled nursing facilities: Effects of Medicare's new prospective payment system," by Mr. White, in the May/June 2003 Health Affairs 22(3), pp. 214-223.