The prevalence of asthma, one of the most common chronic childhood illnesses in the United States, has steadily increased among children in the last 2 decades (1-3) and was 8.9% in 2005 (4). Studies have reported significantly higher rates of asthma prevalence, hospitalization, and death in minority children than in white children (5-14), and studies of Medicaid records show that black and Hispanic children have more emergency department visits for asthma and use fewer preventive asthma medications than do white children (7-11,15). Because childhood asthma results in a large number of physician visits, hospital admissions, and school absences (4,16-20), it disproportionately affects economically disadvantaged families (21-23).

Studies of asthma hospitalization and emergency department visits have concluded that ethnic disparities persist, even after adjusting for socioeconomic status (SES) (10,12,15); however, many past studies used limited samples, such as children on Medicaid (7-11,15), so results are not generalizable to all children or to children who lack health insurance. Furthermore, few studies have considered categories of health service utilization other than emergency department visits or hospitalization (10-13,15,24). Because most children with asthma do not use the emergency department or require hospitalization, inferences from such studies afford a limited view of the overall use of health services for asthma.

The goal of our study was to examine the extent to which race/ethnicity, SES, and health insurance status are associated with differences in utilization of 4 categories of health services (office visits, emergency department visits, prescribed medications, and general checkups) by children with asthma. A better understanding of the factors associated with disparities in utilization of health services for asthma may help policy makers develop health policies that can reduce these disparities.

Approximately one-fourth of children with asthma were black, and approximately one-third lived below the federal poverty line (Table 1). Asthma was less common in girls and in children who were younger than 5 years.

Approximately 7.5% of children with asthma used no health services during the 2-year study period. The most frequently used type of health service was prescription fills, followed by general checkups and office visits (Table 2). Approximately 14% had an asthma-related emergency department visit during the study period.

Distribution of health service utilization varied significantly by family characteristics (Table 2). Black children had the lowest level of office visits and the highest level of emergency department utilization. Children in low-income families had the lowest level of prescription fills. Children with mothers who had the least education had the lowest level of prescription fills and checkups. Uninsured children had the lowest level of health service utilization across all categories except for office visits.

During the study period, black children and children whose mothers had less than a high school education used health services the fewest number of times, except for the emergency department, which they used the most number of times (Table 3). Because few children used the emergency department, the average number of visits was small. Of those who used the emergency department, 70% used it only once, and the maximum number of visits was 3. Overall, black children, children from low-income families, children whose mothers had less than a high school education, and children who had no insurance had the fewest checkups.

Minority children were more likely to use the emergency department for asthma than were white children (Table 4). Mothers who had the most education were most likely to have their children's prescriptions filled and take their children for checkups. Children who had health insurance were more likely to have checkups than were uninsured children. Family income variables were not significantly associated with the odds of using health services, but the pattern suggests that children in poorer families (income <200% of the poverty line) were less likely to use health services than were children in middle- and high-income families. For control variables (data not shown), children aged 5 to 10 years and girls were less likely to have checkups than were older children and boys, respectively. Children who lived in regions other than the West were more than twice as likely to use the emergency department and were also more likely to have checkups.

IRRs shown in Table 5 confirm the pattern of OR results in Table 4, but they provide additional information about the extent of health service utilization. Black children had more emergency department visits and fewer checkups than did other children. Hispanic children had twice as many emergency department visits as did white children and more office visits than either black or white children. With respect to family income, children in low-income families had significantly fewer visits for checkups than did children in high-income families. A similar pattern was found for prescribed medications, although it did not reach significance (P = .07). Children whose mothers had more education visited the emergency department less often and had checkups more often. Children who had only public health insurance used more of all types of health services (although this difference did not reach significance for emergency department utilization).

Most control variables were not significantly associated with frequency of health service utilization, except that younger children had more emergency department visits and fewer prescription fills and checkups. Children who lived in regions other than the West had more health service utilization in general (data not shown).

We found that race/ethnicity, health insurance coverage, and family income and mother's education (as proxies for SES) are associated with differences in all categories of health service utilization for children with asthma. In general, minority children, poor children, children who lack health insurance, and children whose mothers are less educated use more emergency department care and less preventive care.

Consistent with past studies of race/ethnicity, emergency department utilization for asthma is significantly higher for black and Hispanic children than for white children, even after controlling for other socioeconomic factors (10). Part of the explanation may be that black and Hispanic (especially Puerto Rican) children have more asthma attacks than do whites (as supported by trend data [3,6]), but the explanation may also lie with the lack of a usual source of care, with level of home asthma management skills, or with attitudes (13,15). If parents rely on emergency department care for children who are having acute asthma episodes because they lack home management skills, then intervention programs designed to improve caregivers' asthma management skills or attitudes would be useful. However, if parents seek emergency department care for asthma episodes because they lack access to usual care, then more systems-oriented changes may be needed.

Our findings do not show significant differences in prescription filling by race/ethnicity. This finding is in contrast to those of previous studies of children on Medicaid (9,15), which reported that minority children are less likely to use antiinflammatory or controller medications to prevent asthma exacerbations. The discrepancies between these findings may come from differences in the samples (national vs Medicaid). We do not report the analyses of antiinflammatory or controller medications alone because MEPS provides data only on which medications were prescribed, not which medications were used.

We found that Hispanic children have a significantly higher level of office visits than do children of other racial/ethnic groups. Puerto Ricans have a much higher prevalence of asthma than do other racial/ethnic groups or Hispanic subgroups (6), which may explain why Hispanic children have more office visits for asthma than do white or black children. We tested this explanation in a separate analysis by including Puerto Ricans as a separate ethnic group, but we found that non-Puerto Rican Hispanic groups had significantly more office visits than did other groups (data not shown). The small sample sizes of other Hispanic subgroups prohibited further investigation of subgroup differences.

The associations with family income are less straightforward, and most of the associations in multivariate analysis did not reach significance. Only children in low-income families (125%-199% of the poverty level) had significantly fewer checkups. Nevertheless, the overall pattern suggests that children from families at less than 200% of the poverty level use less preventive and urgent care than do children from high-income families. Low-income families may lack the resources needed to manage their children's asthma and prevent exacerbations.

Children with health insurance are more likely to use health services and use more services than children without insurance, although the association usually reached or approached significance in children with only public insurance. The reasons for this finding may be complex. Some children in low-income families who do not have private insurance are likely to be ineligible for public insurance. Consequently, they may not use health services as much as children of poor or near-poor families who have public insurance. In our sample, approximately 17% of children in low-income families were uninsured, compared with 9% in poor families and 10% in near-poor families. Approximately 65% of children in poor families and 40% in near-poor families had only public insurance, compared with 20% of children in low-income families. Public health insurance appears to provide support for poor families, which suggests that children with asthma in low-income families that lack insurance may be in the most vulnerable position.

Children whose mothers had the most education received more preventive care and had fewer emergency department visits than children whose mothers had less education. This finding reinforces the idea that mothers with more education focus more on prevention and rely less on urgent care (15).

A strength of our study is that we used nationally representative data from MEPS. MEPS provides extensive information on health care utilization during 2 complete calendar years and concurrent socioeconomic and health insurance information for children with asthma.

We conclude that minority children of socioeconomically disadvantaged families use more urgent care and less preventive care for asthma, and families without health insurance use fewer health services overall. Providing health insurance to children who do not have it may be crucial to managing asthma, especially for preventive services such as filling prescriptions and having general checkups. To meet the needs of more uninsured children, the federal government should support public insurance programs for underserved children, such as the State Children's Health Insurance Program. Future research should address how additional factors — such as caregiver time, skills, and attitudes — relate to health services utilization in children with asthma.