Eighteen months into HEP I, William D. Ryan, chair of the RAHC board of directors, announced “good news from Rochester, New York,” in testimony before a congressional committee.

HEP was designed to test the proposition that a community could more successfully control the rate of increase in hospital costs, improve the efficiency of hospital services, and maintain or improve the solvency of the participating hospitals on a voluntary basis than under New York State regulations. Ryan's “good news” at that early point in the payment experiment signaled that HEP was a success.

If HEP's purposes and participants—payers as well as hospitals—had remained constant between 1980 and 1990, evaluations of its success over this decade would be relatively straightforward. However, that was not the case. HEP I (1980–84) and HEP II (1985–87) were prospective-payment programs. The major payers (Medicare, Medicaid, and Blue Cross) waived their usual payment rules and contributed an agreed-upon portion of the guaranteed budget for the nine RAHC hospitals. For each hospital, total annual revenue was limited to base-year costs, adjusted for inflation, plus an allowance for changes in costs associated with new equipment and facilities. HEP II added a community constraint on capital spending.

HEP III (1988–90) was markedly different. It was a transition from the prospective-payment system to the NYPHRM case payment system. Medicare was no longer a payer. Three hospitals—most notably, The Genesee Hospital—had withdrawn as participants. The initial agreements emphasizing community control over planning and hospital services were rescinded with discontinuation of community and hospital budget limits.

In this concluding chapter on the Rochester experiment, the focus is on what HEP accomplished and whether it was successful. Was Ryan's initial optimism about the objectives of the prospective-payment program borne out when the experiment ended? Because HEP III was a distinct departure from HEP I and HEP II, the evaluation is mainly limited to the first two HEP iterations.