|
Children’s Ability To Access And Use Health Care
Under new regulations for Medicaid and the State Children’s Health Insurance Program (SCHIP), states can extend health insurance to child enrollees’ uninsured parents. We compared the extent to which child-only and family coverage (child and parent insured) ensure health care access and use for children in working-poor families. Among these children, 21 percent were uninsured, as were 30 percent of their parents. Children with no family coverage encountered more access barriers than insured children. Extending insurance to children markedly increases access and use. The additional benefits of family coverage over child-only coverage seem less pronounced, but family coverage expansions may narrow disparities in access.
Almost ten million children (13 percent) and more than twenty-six million adults (16 percent) lacked health insurance coverage in 1999.1 Children and adults living below 200 percent of the federal poverty level ($33,400 for a family of four in 1999) were more than three times less likely than their higher-income counterparts to have health insurance. Working-poor families faced the highest risk in securing access to care. Children of the working poor are far more likely than children of nonworking-poor families to be uninsured.2 For these children, parental employment does not provide adequate access to employment-based coverage, yet it reduces the likelihood of eligibility for Medicaid. New expansions in health insurance coverage through the State Children’s Health Insurance Program (SCHIP) are making coverage more available to children in working-poor families. In contrast, coverage for working-poor parents has been eroding. A strong economic expansion in the 1990s brought only small declines in adults’ uninsurance rates, attributable to employer-sponsored coverage. Meanwhile, Medicaid participation by low-income families was declining.3 Parents earning just enough to bring them to the poverty level are not eligible for Medicaid in most states. Parents who leave welfare may only qualify for transitional Medicaid, leaving most with no continuous health insurance coverage.4 Extending coverage to parents may benefit their personal access and use but not that of their children. Alternatively, since children depend on parents to enroll them in insurance programs and to arrange for care, health care financing to meet parents’ needs may be an effective way to improve access and use for their children. Previous studies have found a relationship between children’s and mothers’ use of health care.5 These relationships have led policy analysts to infer that by neglecting insurance for adults, SCHIP initiatives may have the unintended effect of reducing their potential impact because they make children and parents enroll in different health plans.6 Under new Medicaid and SCHIP federal regulations, states now have the opportunity to extend coverage to enrollees’ uninsured parents. Policymakers need to examine whether these initiatives indeed improve access to care for children. In this paper we use data from a random probability sample of children and parents from the 2000 National Health Interview Survey (NHIS) to address two questions: (1) To what extent does insurance coverage targeted only at children and not their parents secure access to and use of care by children of the working poor? (2) Would children’s access and use increase if their parents were also insured?
Sample and data source. The NHIS is a continuing survey of the civilian, noninstitutionalized U.S. population, sponsored by the National Center for Health Statistics (NCHS). Different households are sampled each year, and in-person interviews are administered by trained Census Bureau staff using computer-assisted technology.7 In 2000 one child per family was randomly selected for a more in-depth interview on access to and use of health care. Of the 14,744 sampled families with eligible children under age eighteen, 13,376 had interview data on a selected child. Mothers or other adult family members provided information about their children. Compared with children whose parents responded to the survey, children of nonrespondents were more likely to be Asian, uninsured, and in good rather than excellent health. They were also less likely to be Latino or disabled and did not differ on receipt of public assistance. Sample children’s records were linked to their parents’ records, and children who did not live with a parent (n = 461) or were missing information on parental living situation (n = 1) were excluded. Ten children’s records were missing parental work status, and sixty-two were missing health insurance status information for either the parent or the child. Parents were considered working poor if they were either currently working or had been employed for at least six months in the past year, and if their combined family income was at or below 200 percent of poverty. We imputed poverty level for 1,285 children who had missing information on income and could not be otherwise classified as working poor. Imputation used a hot-decking approach, randomly assigning a poverty level category (in 25 percent increments) characterizing individuals with similar geographic location, parental education, race/ethnicity, and broad income levels, if available. The final sample included 4,746 working-poor children and their parents. Measures of access and use. Access to care was measured by financial and structural characteristics. Financial access included type of health coverage at the time of the survey (whether private or public) and continuity in coverage obtained by asking, "In the past twelve months, was there any time when [your child] did not have any health insurance coverage?" Structural measures of access referred to whether the child (1) delayed or missed getting medical care because the parents could not afford it; (2) needed medical or dental care, prescriptions, eyeglasses, or mental health care but was unable to get them; and (3) had a particular person or place to which he or she usually went for health advice or care. We also examined whether the usual place of care was a doctor’s office or a health maintenance organization (HMO), rather than a clinic, health center, or some other place. Health care utilization variables were stratified by perceived health status of the children and included (1) entry into care, measured by the lack of a visit to a physician or other health care professional in the previous twelve months and among those with at least one visit in the past year; (2) whether children had one visit, two to three visits, or four or more visits; and (3) whether they had two or more emergency department (ED) visits.
Independent variables.
We used several questions from the health insurance supplemental questionnaire to classify children and their parents into insured (any private or public coverage) and uninsured status (if not covered by any insurance and used public or free clinics exclusively). In two-parent households, if at least one parent was insured, the "parent" was classified as insured. Based on their insurance status, we created three study groups: child and parent insured (Family Coverage, Group 1); child insured/parent uninsured (Child-Only Coverage, Group 2); and child and parent uninsured (No Family Coverage, Group 3). The small group of uninsured children with insured parents (n = 186) was excluded from most analyses (Exhibits 2
Other independent variables included the child’s sex, age, race/ethnicity, perceived health status, and disability status, defined as a reduction in daily-living activities. Household characteristics consisted of family structure and size and the highest educational attainment of the household head. Geographical characteristics included region of the United States and metropolitan statistical area (MSA). Data analysis. We first compared children’s insurance status and type of coverage by parental insurance status and type of coverage for all working-poor children. We used sample weights to extrapolate to the U.S. population of working-poor children who met our study criteria. Subsequently, we compared the sociodemographic characteristics of our three study groups and assessed statistical significance using chi-square tests. Where statistically significant (p < 05), we report these differences in the exhibit. Finally, we estimated unadjusted percentages for each measure of access to and use of care for each of the three study groups. We conducted multivariate logistic regression analyses to determine whether the differences in dichotomous outcomes between the three study populations persisted after the child’s age, race/ethnicity, health status (except where already stratified) and disability, parental education, family structure, family size, poverty level, region, and MSA were adjusted for. We calculated marginal effects and 95 percent confidence intervals as adjusted differences from the prevalence rates in the reference group. Because the design of the NHIS is a complex, multistage sample, the standard errors and significance tests are weighted and corrected for stratification and sampling clustering. The analyses were conducted using SAS version 8.2 and SUDAAN version 8.0. The NHIS data have some shortcomings. The measure of health insurance is subject to parental reporting error and reflects coverage at the time of the interview, whereas health care use is measured in the past twelve months. Also, health care use over a twelve-month period may be difficult to remember, especially for parents whose children use care frequently.8
Children’s and parents’ insurance status and type. Among children in working-poor families, 21 percent were uninsured, and among their parents, 30 percent lacked coverage (Exhibit 1  ). There was a strong relationship between parents’ and children’s insurance status and type. Among uninsured children, 84 percent had parents who lacked coverage, whereas 13 percent had at least one parent who was privately insured, and 3 percent had a parent who was publicly insured. Among publicly insured children, 53 percent had at least one parent who was also publicly insured, 16 percent had at least one parent who was privately insured, and 32 percent had parents who were uninsured. Finally, among privately insured children, 95 percent had at least one parent who was also privately insured, whereas only 1 percent had a parent who was publicly insured, and 4 percent had parents with no insurance. Overall, one out of six insured children had an uninsured parent.
Characteristics of children in the three coverage categories. Compared with children with child-only coverage, children with no family coverage were older, more likely to be of Latino origin, and not disabled (Exhibit 2  ). Children with no family coverage were also more likely to live in larger two-parent families, in West Coast cities. Compared with children who had family coverage, children who had no family coverage were more likely to be Latino and less likely to be white non-Latino or African American or to be disabled. Furthermore, they were more likely to live in large families with a very low income and low parental educational attainment and to reside in cities in the South and West. Children with child-only coverage also differed from children with family coverage in that they were much younger and had less-than-excellent health. They were also more likely to be Latino; to live in smaller, single-parent and poorer families, with a head of household who had not attained a high school diploma; and to reside in nonmetropolitan areas and in the South.
Differentials in access to and use of care.
As expected, children without family coverage were far more likely than the two groups of insured children to experience difficulties in access to and use of health care (Exhibit 3 Compared with children with family coverage, children with no family coverage had a higher probability of delaying or missing care, of reporting unmet health care needs, of not having a regular source of care, and of using a community clinic or health care center as a regular place for care. Despite large inequalities in access, the disparities in health care use between children with and without family coverage were narrower than those between children with no family coverage and those with child-only coverage. Compared with children having family coverage, those having no family coverage had a higher probability of forgoing a doctor’s visit in the previous year, regardless of their health status. Among those who had seen a doctor and were in less-than-excellent health, the probability of having four or more visits was 17 percent lower for children with no family coverage. It is noteworthy that uninsured children of insured parents had similar rates of access to and use of care as children with no family coverage (data not shown).
Comparison among insured children.
There were few disparities in access to and use of health care between children who had family coverage and those who had child-only coverage (Exhibit 3
Child-only versus family coverage. Our findings suggest that extending health insurance coverage to children of the working poor who have no family coverage markedly increases their access to and use of health services. In fact, on almost every indicator that we examined, children with no family coverage encountered more barriers to care than children with child-only or family coverage. Once the child is covered, it appears that extending coverage to the parent does not confer many additional benefits. However, when coverage was extended to parents and children rather than exclusively to children, children were more likely to obtain continuous coverage and to be seen regularly at a doctor’s office or HMO. This may partially explain why among children in less-than-excellent health, those with child-only coverage were less likely than children with private family coverage to have visited a physician in the past year. With respect to other indicators, our findings show that the gaps in access to care and use of physician services between those with child-only coverage and those with family coverage were quite comparable. Impact of Medicaid/SCHIP expansions. According to our results, 21 percent of children of working-poor families in 2000 were uninsured. This is a drop of one percentage point since health insurance expansions through SCHIP and Medicaid were mandated by law in 1997.9 However, the estimate could also be affected by a higher uninsurance rate among those excluded from our sample. Although SCHIP plans are progressing in most states, and states have taken steps to make enrollment easier, enrollment has been slower than expected. The complexity of administrative and eligibility structures, cost-sharing provisions that may discourage enrollment, poorly designed outreach programs, insufficient efforts to ensure that enrollees remain covered, and poor coordination among SCHIP and other welfare programs are some of the difficulties encountered.10 These roadblocks, in addition to cultural and linguistic barriers, need to be addressed if SCHIP is to cover its target population. Similar improvements are needed to simplify Medicaid eligibility procedures and facilitate enrollment in and maintenance of Medicaid coverage. Participation of eligible children has lost ground since welfare reform was implemented.11 However, nonwelfare Medicaid is increasing as older children in poverty are phased in. Coverage expansions for parents. Efforts are under way to build upon recent insurance expansions for children by adding coverage expansions for their parents. As a result of policy changes, states that can meet the guidelines now can request waivers to use federal matching funds to provide health insurance to uninsured parents of SCHIP enrollees.12 One requirement is that states show that it is not more expensive to cover the parents than it would have been to cover the child in SCHIP. Although several states recently approved Section 1115 demonstration projects that extend Title XXI SCHIP funding for parents of eligible children, actual implementation of these projects is threatened by budgetary restrictions. In the 1990s Delaware, Hawaii, Minnesota, Oregon, Massachusetts, Tennessee, and Vermont used Section 1115 waivers of Medicaid rules to greatly expand family coverage. New flexibility in the 1996 federal welfare law that delinked Medicaid eligibility from welfare eligibility offers states the opportunity through Section 1931 to expand Medicaid eligibility to more low-income parents, by allowing broad discretion in defining what counts as income and assets and by extending eligibility to more two-parent families.13 California, Connecticut, and the District of Columbia, for instance, have used Section 1931 authority to expand family Medicaid coverage, while others such as New York have combined Section 1931 and Section 1115 waivers to extend coverage to parents of Medicaid-eligible children. In addition, states still have the option to expand family coverage through state funds. Two expectations drive these initiatives: that they will reduce the number of uninsured adults and that they will encourage the enrollment of children.14 Preliminary evidence from our study suggests that when health insurance is extended to uninsured, low-income parents, they have greater ability to obtain regular care in a private office or HMO for their children and avoid breaks in coverage. Furthermore, evidence from case studies of broad Medicaid eligibility expansions that include parents, such as those in three pioneering states (Oregon, Hawaii, and Tennessee) in 1994, have shown that these expansions help stimulate moderately higher Medicaid enrollment rates among children under age six.15 However, take-up rates for Medicaid coverage remain low among adolescents.16 Link between insurance and health care need. While insurance may help children get health care, some children may have health insurance because they need care. The causal relationship between insurance status and health status could not be ascertained with cross-sectional data. In general, the effect would be to overestimate the benefit of a program to extend coverage to uninsured children. By adjusting for several covariates and stratifying by health status, we tried to equalize other factors that could influence the outcomes, but this may not be a true simulation of the conditions that could occur were we to extend coverage to uninsured parents or children under new health insurance regulations. Furthermore, parents who secure health insurance for their children may be different from parents who do not make an effort or are unable to secure coverage, in ways that we were unable to address in this study. Based on our study findings, we conclude that expansion of coverage for parents and children does not confer many additional benefits over child-only coverage. However, by reducing breaks in coverage and increasing access to HMOs and private physicians’ offices as a regular source of care, new public expansions in family insurance coverage are likely to narrow disparities in access to and use of health care for children of the working poor.
This study was partially funded by a grant from the Henry J. Kaiser Family Foundation. The authors thank Kristen Marchi for assistance with statistical analysis, Helen Halpin and Steve Samuels for their suggestions, and Lora Santiago for clerical support. Sylvia Guendelman (sylviag{at}uclink4.berkeley.edu) is professor and chair of the Maternal and Child Health Program in the School of Public Health, University of California, Berkeley. Michelle Pearl is a research scientist at the Sequoia Foundation, Genetic Disease Branch, in Richmond, California.
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Study Methods
Study Results
