Health Insurance Coverage Among Persons With Schizophrenia in the United States

The analytic sample contained 328 records representing 493,006 U.S. noninstitutionalized, civilian adults per year with schizophrenia. Sixty-seven percent reported they had Medicaid, 46% had Medicare, 87% had either Medicaid or Medicare, 26% had both Medicaid and Medicare, 8% received care from the Veterans Health Administration, 6% had other public insurance or TRICARE, and 15% reported having private insurance for at least one day during a given year ( Figure 1 ). About 7% were uninsured all of the year. In the U.S. general population, the analytic sample included 117,532 records representing 220,280,457 adults per year. Twenty-six percent of the general population reported having Medicare or Medicaid, 71% reported having private insurance, and 14% reported being uninsured throughout the year.

Type of insurance coverage was examined across demographic and access-to-care variables among persons with schizophrenia. Persons with schizophrenia who were uninsured tended to be male (92%), nonwhite (54%), and unmarried (97%) ( Table 1 ). Those with private insurance had a higher employment rate (42%) than those with public insurance (15%) or the uninsured (29%). Access to care also tended to be reduced among persons without insurance in terms of having a usual source of care and time since last medical checkup. Thirty percent of persons who were uninsured reported that more than five years had elapsed since their last medical checkup, about three times the rate of those with health insurance.

This study investigated the rates and types of health insurance coverage among adults in a civilian, noninstitutionalized U.S. population-based sample who reported having schizophrenia. We documented the high prevalence of public insurance among those with schizophrenia, with almost 90% having either Medicare or Medicaid. Although there was a lower rate of uninsurance (7%) compared with the general population (14%), almost all persons with schizophrenia should be eligible for public insurance.

Uninsured persons with schizophrenia tended to be men who had never married or who were widowed, separated, or divorced, representing a population of persons with a serious illness who are likely to be socially isolated. Lack of insurance was marginally associated with not having a recent medical checkup (p=.09). Because recent medical checkups are associated with higher completion of prevention recommendations, lack of insurance may affect medical outcomes for this group.

A previous study by Yanos and colleagues ( 8 ) examined the association of insurance status with demographic characteristics and service use in a group of 539 persons with schizophrenia spectrum disorders from a statewide behavioral health care system ( 8 ). The investigators found a 20% rate of unisurance, which was substantially higher than our findings. They also found that uninsurance was associated with younger age, Latino ethnicity, and male sex. However, the sample in Yanos and colleagues' study was not a household survey; rather it captured those who had presented for services, so it included people who may have been homeless or institutionalized and who may be more likely to be uninsured.

Stable and comprehensive health insurance coverage is particularly important for persons with schizophrenia, because they require long-term health services to treat this chronic illness. In addition to improving the quality of life in schizophrenia, effective treatment reduces the burden to caregivers, reduces health care costs, and improves overall physical health ( 9 ). In contrast, persons with schizophrenia who do not take antipsychotic medications as recommended have increased hospitalizations, use of emergency psychiatric services, violence, victimization, and substance abuse and reduced mental functioning, all of which increase costs ( 10 ). Health insurance for persons with schizophrenia also pays for ongoing psychotherapy and engagement in rehabilitation or assertive community treatment programs, which are recommended by the Schizophrenia Patient Outcomes Research Team (PORT) guidelines ( 11 ). A recent examination of Florida Medicaid claims records from 1996 to 2001 for persons with schizophrenia found that only 18% of acute treatment and 7% of maintenance treatment met all PORT quality standards ( 12 ). The investigators found that although there was a greater likelihood of receiving an antipsychotic over time, there was also a lower likelihood of mental health visit continuity.

There were several limitations regarding our sample of persons with schizophrenia. Although it is a population-based survey weighted to the national population, the MEPS is not intended to estimate prevalence of medical conditions; rather it is designed to provide estimates of insurance and health service access and utilization ( 13 ). Although the estimated prevalence of schizophrenia is between .5% and 1.1%, the MEPS estimate was lower ( 14 ). Because it is a household survey, it likely does not capture persons with schizophrenia living in supported arrangements, such as group homes, or institutional settings, such as prisons. It may also underrepresent undocumented immigrants and persons who are homeless, groups that are more likely to be uninsured ( 15 ). It is also possible that schizophrenia was underreported because one respondent provides information for the entire household and the medical condition enumeration comes from an open-ended question. In addition, persons with schizophrenia who do not go to a medical provider would not have that data imputed. Finally, even after several years of MEPS data were combined, the sample of persons with schizophrenia may have been too small to detect significant differences between health insurance type and demographic variables and access to care.

We used a national survey to estimate health insurance coverage among noninstitutionalized adults with schizophrenia. Almost all U.S. adults with schizophrenia are eligible for government health insurance, but a nonnegligible minority were uninsured. This is an especially vulnerable population because they not only have an increased risk of many physical diseases, but they also are in need of lifelong psychiatric care. These health insurance estimates highlight opportunities for improving health service delivery for this vulnerable population, particularly because almost all of those with health insurance were covered by government administered programs. However, the high proportion of persons with schizophrenia covered by public insurance programs suggests that future provision of universal coverage for this diagnostic group would not have a strong effect on overall costs. Future directions for research may include determining the risk factors for uninsurance among persons with schizophrenia and finding opportunities for improving care for the vast majority who are receiving public insurance.

This study was funded by the Psychiatric Epidemiology Training Program and through grant 5-T32-MH014592 from the National Institute of Mental Health.

The authors report no competing interests.