HEDIS: almost ready for prime time - HEDIS Report Card Pilot Project for health maintenance organizations

Business & Health, March, 1995 by Gary M. Stephenson, Steven Findlay

Gary Stephenson writes frequently about health care. He lives in Balitmore, Md.

In a pilot study of the report-card concept, 22 managed-care plans earned enviable marks. But experts warn employers not to make too much of the results.

The tabulations are in and, the envelope please...everyone won! Or at least nobody lost. That pretty much sums up the results of the $2.1 million HEDIS Report Card Pilot Project, a year-long initial go at measuring the performance and quality of 22 health plans using a standardized set of criteria.

All plans scored high on most of the basic measures, in large part because the majority of them had initiated programs over the past several years aimed at improving quality in exactly those services that this project measured. These include mammography rates, rates of coronary-bypass surgery and cesarean sections, cholesterol screening, childhood immunizations, and enrollee access to physicians and health-care services.

Still, the vast majority of the roughly 10,000 enrollees surveyed as part of the HEDIS pilot gave their health plans high marks. This underscores a clear message to all plans: Customer-oriented improvements get noticed. "The pilot project has been a success in showing that managed-care organizations can be accountable for doing the things that are important to their patients," says John M. Ludden, M.D., senior vice president for medical affairs for Harvard Community Health Plan, Brookline, Mass., one of the plans in the study.

In some areas, however, there remains considerable room for improvement. For example, about 30 percent of enrollees said they had poor access to the plan doctor of their choice. Plans also scored low overall on giving diabetics timely screening for eye problems. And several leading health plans had quite low cholesterol-screening rates.

Producing the pilot project was a meticulous, sometimes torturous process that brought together dozens of people to hammer out hundreds of details about what to measure and how. And in the end the results had more to say about the feasibility of gauging healthplan performance than about how well the tested plans are doing their jobs. Measuring performance can be accomplished, say those involved in this effort. But they also agree that it isn't easy, and that it will take several more years of honing their methods before "the HEDIS system" is sufficiently refined to yield the kind of accurate, objective assessments of health plans that would enable employers and consumers to make intelligent, informed choices among them.

"What we learned is that we can gain enough information to be dangerous," says Joan Marie Vereeke, a member of the pilot study steering committee and manager of managed-care plans for Federated Department Stores in Cincinnati. "That is, we learned enough to know we're headed in the right direction, but not enough to be absolutely sure the numbers are a true characterization of quality. I'd say we're about three years from having a real national report-card system."

Until then, Vereeke and Ludden agree, employers shouldn't use HEDIS data as the sole criteria for purchasing decisions.

EMPLOYERS IN THE LEAD

The HEDIS Pilot Project grew out of the activities of a handful of large employers in the late 1980s. GTE, Xerox, Digital, and Bull HN Information Systems, Inc., in particular, were determined to develop a systematic tool to measure the performance and services of HMOs and guide employers in purchasing health care. The group hooked up with the National Committee for Quality Assurance (NCQA), a Washington, D.C.-based managedcare accrediting body, in 1992, and over the next year developed HEDIS 2.0, the Health Plan Employer Data Information Set. NCQA is currently working on a refined version, due out this spring as HEDIS 2.5, which will give way to HEDIS 3.0 next year.

The "set" consists of 36 performance measures. Released to other employers and the managed-care industry in November 1993, it almost instantly became the gold standard of quality assessment. HMOs around the country adopted its criteria, and employers pressed for its widespread use.

At least a dozen HEDIS-based report-card efforts have been produced so far, some by individual plans, and some by coalitions of employers and managed-care providers. The New England HEDIS coalition, for example, brought together 14 health plans with 17 employers and employer coalitions in the Northeast. Another such effort, by The HMO Group, an alliance of group-practice plans, released HEDIS 2.0 data on 15 of its members last September. And last August, U.S. Healthcare, based in Blue Bell, Pa., became the first multistate HMO to publish HEDIS data on its plans.

But it was clear to all from the beginning that a more formal and objective test of HEDIS was needed. Thus, the pilot project, officially launched in January 1994.

Of the 36 HEDIS measures, 28 were selected for the project. (See page 50.) In addition, researchers polled some 10,000 enrollees on several patient-satisfaction issues.