The effects of predetermined payment rates for Medicare home healthcare

Health Services Research, Oct, 1997 by Randall Brown, Barbara Phillips, Christine Bishop, Craig Thornton, Grant Ritter, Amy Klein, Peter Schochet, Kathleen Skwara

Finally, there is a basic flaw in any payment approach or limit based on visits: visits can vary widely (and arbitrarily) in content and cost (Bishop, Brown, Phillips, et al. 1996). For example, skilled nursing visits for intravenous therapy typically take much longer and may require more experienced staff than visits to administer insulin or draw blood. Similarly, home health aide visits can be very short (e.g., to turn a bed-bound patient) or can last several hours (to help patients with bathing, toileting, and eating). Furthermore, whether agencies provide a given type of care in multiple short visits or in a single longer visit is often at their discretion. We saw no evidence that agencies exploited this opportunity to increase revenues (perhaps because of the volume adjuster), but the possibility clearly exists. Variation in visit length and content could explain some of the current variation across agencies in per-visit costs.

The recent explosion in the number of visits per episode is the primary source of the large increases in Medicare home health costs, and setting a fixed reimbursement rate prospectively for home health visits would not reverse this trend, even if it did lower unit costs. The demonstration results suggest that the effects of prospective rate setting on costs per visit were limited at best because agencies had to compete for staff and referrals, and this competition--plus regulations and other factors - constrained their ability to make major reductions in costs. Furthermore, agencies already had the incentive to hold their costs below the cost limits.

Nonetheless, there are important lessons to be learned from this demonstration. Agencies can make some changes to slow the rate of increase in their per-visit costs. These actions will be important under any system of payment, because even small decreases in average costs could save millions of dollars for Medicare. The demonstration suggests that when paid a prospectively set rate, most Medicare-focused agencies found a way to hold average cost increases at least slightly below the inflation rate used by HCFA, and they were nearly twice as likely to do so as agencies under cost reimbursement.

Moreover, policies directed at reducing the number of home health visits may turn out to be less desirable than ones directed at reducing per-visit costs, because of the greater potential for adverse effects on quality of care and access. Shaughnessy, Schlenker, and Hittle (1994) found that HMO patients had much shorter home health episodes but were more impaired at discharge than home health patients with traditional fee-for-service Medicare coverage. Although setting per-visit rates prospectively would not be expected to reduce costs substantially, it might be a safe interim strategy for encouraging agencies to moderate their cost increases and for getting high-cost agencies to bring their costs more in line with those of other agencies. This approach would allow time to assess quality assurance procedures for ensuring that agencies paid a per-episode or per-beneficiary capitation do not underserve patients, and to develop an adequate case-mix adjuster so that agencies are paid fairly for the types of patients they serve. If prospective rate setting were adopted, however, it might be necessary to have some provision, such as the volume adjustments to payment rates used during the demonstration, to encourage agencies to limit increases in visits per episode. It would also be important to develop a usable definition of what constitutes a home health visit.