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

The voluntary nature of the demonstration may also have biased the findings. Agencies that declined to participate typically cited as reasons their fear of losing money under the demonstration, the limited opportunities they perceived for reducing cost per visit (e.g., because of competition for staff and referrals), concerns about the costs of participation, and an aversion to switching fiscal intermediaries for three years. If the agencies that entered were those most able to prosper under the demonstration rules, the estimates could overstate the effects of a mandatory program of prospective rate setting. On the other hand, if most agencies that entered expected their costs to fall for other reasons (such as computerization of records or changes in staffing patterns), the effects of national prospective rate setting could be underestimated here. Under this scenario, control group agencies' costs would grow more slowly than those of the typical agency. Given that we find no major effects on costs, it is clear that the change in payment structure was not an effective way to reduce costs for the group of participating agencies. If nonparticipating agencies were not as motivated to control their costs, prospective rate setting might provide the incentive they need to adopt more effective practices.

A third difficulty with the study is that agencies' behavior under the limited duration of the demonstration might have differed from what behavior would be under permanent prospective rate setting. On the one hand, agencies might have been reluctant to make major cost-reducing changes (such as eliminating staff) that they would not want to sustain when returning to cost-based reimbursement after the three-year demonstration. Conversely, agencies might have postponed until after the demonstration changes that they needed to make in the long run, in order to earn profits during the limited demonstration period. For example, agencies that grew rapidly during this period (as many did) might have delayed adding additional administrators and supervisors until returning to cost-based reimbursement.

Basing an agency's prospective rates on its own past experience rather than on national or regional averages also might have distorted inferences about what would occur under national prospective rate setting. If agencies were forced to lower costs relative to some minimally efficient standard (as would be likely under a national system), some might become more aggressive than they were under the demonstration. Conversely, those with already low costs could profit without changing their behavior at all. Paying all agencies some percentage of the current national average would generate national savings (provided that this process did not prompt increases in volume). The demonstration offered no guidance, however, about how many agencies would be able to survive under such a system, the types of agencies that would have the most trouble adapting, or the problems that could arise with quality and access if agencies' costs were not fully covered by the rate.