Data were taken from the 1998 National Research Corporation Healthcare Market Guide Survey (1), in which household members primarily responsible for family health care decisions completed questionnaires. We examined responses to two questions. "All things considered, how satisfied are you with your health plan?" "Does any household member have any of the following ailments: diabetes, sciatica, depression, arthritis/rheumatism, or asthma?"

We created a four-part variable: "neither," no family member had diabetes or depression; "diabetes only," at least one had diabetes and none had depression; "depression only," no one had diabetes and at least one had depression; and "both," at least one had diabetes and at least one had depression. We cross-tabulated this variable with health plan satisfaction.

The results for families covered by two plan types are shown in F1 and F2. The relatively small differences in satisfaction ratings between types of plan may be due to the fact that traditional plans are increasingly managed, and the differences are therefore blurred. Another explanation may be that enrollees' satisfaction with the enhanced access to care in HMOs offsets their dissatisfaction with the lower intensity of care. Differences also may be blurred by the inability of the survey to assess the use of managed behavioral carve-outs by either type of plan.

The most striking difference was in the satisfaction ratings of enrollees in the diabetes-only group and those in the depression-only group. In traditional plans, 34.7 percent of diabetes-only enrollees were completely satisfied with the plan, compared with 21.6 percent of the depression-only enrollees. Moreover, almost twice as many depression-only enrollees as diabetes-only enrollees were completely dissatisfied.

The pattern of differences between the diabetes-only and depression-only groups might be explained in several ways. Care may be worse for depression than for diabetes. Coverage may be worse for depression. Negative attitudes associated with depressive states may have affected the responses. Finally, diabetes care may engender less ambivalent feelings toward caregivers.

Dr. Haviland and Dr. McGhee are affiliated with the department of psychiatry at Loma Linda University School of Medicine, 11374 Mountain View Avenue, Loma Linda, California 92354-3842 (e-mail, haviland@ix.netcom.com). Dr. Dial is with the National Education Association in Washington, D.C. Dr. Pincus, who is coeditor of this column with Terri Tanielian, M.A., is affiliated with Rand and the department of psychiatry at the University of Pittsburgh School of Medicine.