States grapple with collecting, publishing survey data on managed care organizations

States can use information to discipline the market’ and keep costs down

Don’t expect Medicaid enrollees to change their health plans based on the results of state health maintenance report cards, say the authors of the massive Consumer Assessment of Health Plans Study (CAHPS), a joint effort of the Rand Corporation, Harvard Medical School, and the Research Triangle Institute.

Nevertheless, the information is invaluable for states in evaluating HMOs and directing covered lives to plans that win high marks from enrollees, Rand research scientist Matthew Lewis told state Medicaid directors gathered in Bethesda, MD, in late October. "By doing that, you start disciplining the market," he said.

Mr. Lewis cautioned that survey data do not come cheap, with the cost per completed response likely to come in at the high end of a range between $5 and $20. Comparative data must be based on at least 300 to 400 completed responses per plan in order to be reliable, he said.

Results measured against average

When presenting the results, CAHPS researchers compare answers for a given plan against the average calculated from all health plans. For example, information about whether an enrollee would recommend a plan describes whether enrollees would recommend it less frequently, about as often, or more frequently than all health plans overall.

Don’t be surprised if enrollees disparage individual features of a plan but give it high marks overall, Mr. Lewis said. "It’s a interesting phenomenon. People can go and rag on the plan but give it a pretty high rating. People aggregate the data in their heads and it tends to be a more positive outcome."

While most states provide the results of plan surveys in booklet or newspaper format, the CAHPS study has allowed New Jersey and Florida to experiment with computer kiosks for that function. One advantage of a computer-based system is that it organizes plan information according to the features an individual enrollee defines as the most important.

Lewis pointed out the following key issues officials must consider when implementing computerized plan information:

Motivation. State officials must remember an information kiosk typically has to compete for attention with the noise and cramped conditions of a typical government office waiting room. "People have kids pulling on them and they’re waiting for their number to come up," Mr. Lewis said. "How do you motivate people to actually use this thing?"

Accessibility. Use of a touch screen makes the system easy to use among people at all levels of skill and literacy.

Efficient use of time. "If you’re lucky, you’re going to get 10 or 12 minutes with a person sitting down at a kiosk in a benefits office," Mr. Lewis said. A system should direct an enrollee to relevant information as quickly as possible, he said.

Individualized information. A computerized system is uniquely capable of highlighting plan results that reflect the features an enrollee considers the most important. A printed version of the plan allows enrollees to discuss the information with others before making a decision. Enrollees often return for benefits counseling two or three times before making a decision, said Martha Walters, regional director of benefits counseling firm Benova Inc. in Philadelphia.

Links to other databases. In Florida’s computer system, enrollees can link to a provider database that lists the plans in which individual physicians participate.