Overview.

This issue features child and adolescent health care, focusing especially on the effectiveness of the 1989 Omnibus Budget Reconciliation Act (OBRA 89), which expanded health benefits to more children and pregnant women in Medicaid. Also featured: the effectiveness of some managed health care plans for Medicaid-eligible children, and injury hospitalizations in California in 1992. Some of the material is particularly relevant to the Children's Health Insurance Program (CHIP), which is the current effort to insure the Nation's working poor.

Most of the major causes of childhood mortality and morbidity have been eliminated with improvements in water, food quality, housing, and other public health measures including effective vaccines (Dubos, 1959; Yankauer, 1973). However, many children in the United States had only limited access to health insurance until the enactment of Medicaid in the 1960s. Since then, several expansions in the Medicaid program have occurred to insure even more low-income children and pregnant women. This issue of the Health Care Financing Review contains several evaluations of Medicaid expansions as well as other articles about child health services. This issue is particularly timely in light of current activity to establish the State CHIP, which expands insurance coverage to children of the working poor. The results of the studies in this issue may offer insights into assessing the potential effectiveness of the new coverage expansions.

An expansion of insurance coverage for low-income children occurred when OBRA 89 expanded Medicaid eligibility, created incentives for provider participation, and allowed providers to broaden their scope of services. The analyses in this issue reveal the following results of the 1989 expansions. Adams and Graver tracked child/provider ratios before and after OBRA 89. They found substantial growth in enrollment in four States (California, Georgia, Tennessee, and Michigan) and higher child/provider ratios after the expansion. Michigan in particular had a striking increase in the number of Medicaid providers. Herz, Chawla, and Gavin evaluated changes in use of preventive services in fee-for-service (FFS) Medicaid in the same four States. They found increased enrollment and modest improvements in rates of use of preventive services. There was a notable increase in use of dental services in Georgia, but not the other three.

This issue also contains two evaluations of Medicaid managed care before and after OBRA 89. Gavin, Farrelly, and Simpson evaluated controlled Medicaid demonstrations in Florida and New Mexico. Use of children's preventive services marginally increased in both States compared with contemporaneous FFS controls. The second study, by Lo Sasso, evaluated changes in two “mature” California managed health care plans between 1989 and 1992. They found small improvements in immunization levels, but not in use of well-child care visits. Epstein and Newhouse studied expansions in prenatal care in South Carolina and California during the same period. Their analysis revealed that, while more women became eligible for prenatal services, use of these services increased modestly in one State, and not at all in another, and that birth outcomes remained unchanged.

These studies suggest that the OBRA 89 expansions increased enrollment and number of providers and made modest improvements in use of services. However, use of these services was still considerably lower than what was recommended by the American Academy of Pediatrics (1988) at that time. Readers are cautioned that major limitations affect interpretation of these evaluations. The studies had serious data constraints. Reliable claims-based data on individual enrollees were available from only four States, which is not enough to measure national trends. The evaluations were also unable to explain State-to-State or county-to-county variations. In addition, the Medicaid program has very little information about levels of participation and quality of care provided to children in capitated managed care plans because they do not generate bills for individual episodes of care. Because of this, evaluations of managed care for children have been sporadic and cannot be generalized. This problem is highlighted in the Florida and New Mexico evaluations. The studies' reliance on administrative data from a single year prevents a direct assessment of whether the expansions improved the health status of the newly enrolled children and pregnant women. However, the available data allowed a limited evaluation of effectiveness in the article by Gavin. The authors report a concomitant decline in emergency department visits and hospitalizations for ambulatory care-sensitive conditions among demonstration and non-demonstration counties in Florida. Encouragingly, the decline was greater in the demonstration counties, suggesting a positive effect of the Florida demonstration program on child health. This result suggests that measurement of health outcomes should be an intrinsic part of Medicaid program evaluations.

Currently, 60 percent of States in the U.S. send Medicaid claims data to the HCFA. HCFA transforms the claims into annual research files called the State Medicaid research files (SMRF). These files are accessible to researchers, simplifying the job of evaluation. By 1999, all States are required to provide Medicaid claims data to HCFA, which will support a national evaluation effort. In addition, the CHIP legislation requires enrollment outreach plans such as outstationing enrollment workers in the community, instituting mail-in applications, using presumptive eligibility, extending the length of eligibility to 12 months despite income changes, and simplifying the enrollment process. CHIP policymakers are developing a system for monitoring indicators of success of the program. The first indicator of success is increased enrollment of children and reduction of the number of uninsured low-income children. The States will also evaluate effectiveness of CHIP by collecting information on key indicators such as child characteristics, immunization timing and levels, well-child visits, and followup visits such as visits for dental and vision screening and treatment.

Other articles about child health services in this issue include an evaluation of patterns of enrollment and use of services in Pennsylvania's CHIP by Lin and Lave. The authors found that length of enrollment was directly related to consumer cost of the program. They also found evidence of pent-up demand for services such as dental, vision, and hearing care among newly enrolled patients, and increased emergency department use. Perrin, Kuhlthau, Ettner, McLaughlin, and Gortmaker found a fourfold increase in enrollment of children from the supplemental security income program in Medicaid. Finally, this issue contains a data view of injury hospitalizations among Medicaid-enrolled children in California during 1992. Baugh, Rotwein, Hakim and Boschert describe patterns of non-fatal inpatient injuries in California in 1992. This is a preliminary analysis of injuries among Medicaid-enrolled children. Because of the importance of intentional and unintentional injuries during childhood, HCFA will continue to develop both inpatient and outpatient injury data from more States and years.