Background. The recently enacted State Children's Health Insurance Program (SCHIP), designed to provide affordable health insurance for uninsured children, was modeled in part on New York State's Child Health Plus (CHPlus), which was implemented in 1991. All SCHIP programs involve voluntary enrollment of eligible children. Little is known about characteristics of children who enroll in these programs.
Objectives. To provide a profile of children enrolled in CHPlus between 1993 and 1994 in the 6-county upstate New York study area, and to estimate the participation rate in CHPlus.
Methods. A parent interview was conducted to obtain information about children, 0 to 6.9 years old, who enrolled in CHPlus in the study area. Two school-based surveys and the Current Population Survey were used to estimate health insurance coverage. Enrollment data from New York State's Department of Health, together with estimates of the uninsured, were used to estimate participation rates in CHPlus.
Results. Most children enrolled in CHPlus in the study area were white. Although 17% of all children in the study area who were <13 years old and living in families with incomes below 160% of the federal poverty level were black, only 9% of CHPlus-enrolled children were black. Twenty-one percent of enrolled children were uninsured during the entire year before enrollment and 61% of children had a gap in coverage lasting >1 month. Children were generally healthy; only 4% had fair or poor health. Eighty-eight percent of parents of enrolled children had completed high school or a higher level of education. Parents reported that loss of a job was the main reason for loss of prior health insurance for their child. Most families learned about CHPlus from a friend (30%) or from their doctor (26%). The uninsured rate among children in the study area was approximately 4.1%. By 1993, the participation rate in CHPlus was about 36%.
Conclusion. Blacks were underrepresented in CHPlus. Because the underlying uninsured rate was relatively low and parental education and family income were relatively high, the effects of CHPlus observed in this evaluation may be conservative in comparison to the potential effects of CHPlus for other populations of children. Participation rates during the early years of the program were modest.
The State Children's Health Insurance Program (SCHIP), enacted in August 1997, is the largest investment in child health in nearly 30 years. SCHIP was designed to provide health insurance coverage for the estimated 11 million uninsured1–3children in the United States. An additional 7.9 million children are underinsured2 and the SCHIP legislation permits states to use state funds to provide coverage for some underinsured children. New York State has chosen this option. Over the past decade the number of uninsured and underinsured children in the United States has not declined, despite Medicaid expansions and tax credits for health insurance.3,,4
In the early 1990s, states began developing health insurance programs for children.5 New York State has the largest program—Child Health Plus (CHPlus)—which was implemented throughout the state in 1991.6 Several other states also established state-funded programs, including MinnesotaCare, Pennsylvania's Children's Health Insurance Program, and Florida's Healthy Kids Program. In other states, low-income children were offered health insurance programs by private insurers; examples include the Alabama Caring Program and the Western Pennsylvania Caring Program for Children.
In this article we examine results of an evaluation of CHPlus in Monroe County, New York (including the city of Rochester), and 5 surrounding counties, henceforth referred to as the study area. We present the demographic and prior health insurance characteristics of the study population, the results of a survey designed to estimate the number of uninsured children in the study area, and an analysis of theCurrent Population Survey to estimate participation rates in CHPlus.
Profile of Children Enrolled in CHPlus
To obtain information about the characteristics of children enrolled in CHPlus, we conducted a parent interview after the child's enrollment. The methodology used to gather this information is described in detail in an accompanying article.7 Most outcome measures were stratified by type of prior insurance and gap in coverage. Type of prior insurance was categorized by the extent of services covered. Four categories were created: 1) uninsured; 2) underinsured (plans providing inpatient and/or catastrophic or major medical coverage but limited coverage for ambulatory services); 3) fully insured (plans, such as a health maintenance organization, that provide full coverage of most ambulatory and inpatient services); and 4) Medicaid. Parents then were asked whether their child had experienced a gap in coverage—that is, a period of time without health insurance coverage before enrollment in CHPlus. The length of the gap in coverage was recorded in months and was categorized as none, <6 months, and 6 or more months.
The Uninsured Rate in the Study Area
To gather information about the uninsured rate in the study area, we used data from the Current Population Survey, and we conducted 2 separate surveys of families in the study area. Estimation of health insurance coverage based on the Current Population Survey has limitations. Among these are: 1) the data are not consistent across years because health insurance questions were modified during the study period; 2) lack of health insurance is not asked directly, but rather is determined as a residual category; 3) health insurance questions refer to experiences during the previous 12-month period; 4) the Current Population Survey samples are constructed to be representative at the state level, but not for a smaller geographic area such as the study area; and 5) the health insurance questions were not designed to determine enrollment in programs such as CHPlus, and, as a result, CHPlus enrollees could be categorized as uninsured (the residual) or as privately insured.Current Population Survey estimates of health insurance coverage rates must be viewed with these limitations in mind. Despite these issues, the Current Population Survey is among the most frequently used source of data used to estimate the number of uninsured children,1,,8,9 and analytic techniques described below can improve the accuracy of Current Population Surveydata.
We pooled 5 years of data (the 1990 through 1994 cross-sections that contain health insurance data for the period 1989 through 1993) and reweighted by strata using census data to make the data representative of the study area and to bolster sample sizes and increase the precision of estimates. All standard errors are calculated using Huber/White/sandwhich estimators to incorporate probability weights.10,,11 We categorized health insurance coverage in the Current Population Survey as employer-based, which includes parents' employer or federal government for US military dependents (Civilian Health and Medical Programs of Uniformed Services) insurance; other privately insured, which includes commercial insurance; Medicaid; and uninsured.
To support the Current Population Survey estimates of the uninsured rate, we conducted 2 separate local surveys of health insurance coverage: 1) among children in kindergarten through fifth grade in a rural county in the study area (Livingston County); and 2) among children attending kindergarten or pre-kindergarten programs in the Rochester City School District ([Eighty-five percent of age-eligible children in the city of Rochester attend the Rochester City School District, with the remaining children attending private or parochial schools] Monroe County; personal communication, A. McGowan, Evaluation and Testing, Rochester City School District). At the beginning of each school year, all schools in the study area require parents to complete a health form. We obtained permission from all school districts in Livingston County and the Rochester City School District to include 2 additional questions on the 1993 health form. The first question asked, “Is your child now covered by any health insurance (including Medicaid)?” The second question asked, “If yes, what is the name of your child's health insurance?” Analysis of these surveys provided estimates of the uninsured rate in the study area.
Participation Rate in CHPlus
To estimate the participation rate in CHPlus, we first had to estimate the number of CHPlus-eligible children in the study area. Because the 5-year pooled Current Population Survey sample of uninsured children is small in the study area, we limited our criteria for CHPlus eligibility to children who were <13 years old and uninsured. Although premium subsidies were linked to income, the only children explicitly excluded from eligibility were those eligible for Medicaid.
We calculated the total number of eligible children by summing the estimated total number of uninsured children and the total number of CHPlus enrolled children. The estimated participation rate, then, is given by the total number of enrolled children divided by the total number of eligible children. We calculate participation rates using each of the 3 estimates of the number of uninsured children in the study area.
Profile of Children Enrolled in CHPlus
Results were obtained by parent interview and are for all children, 0 to 6.9 years old at the time of enrollment in CHPlus (N = 1853). Results for older children 7 to 12 years old are not included in this article because we only obtained information for a subset of these children who had asthma; an accompanying article describes that subgroup of children.12Table 1 shows demographic characteristics of children enrolled in CHPlus in the study area. Fifty-two percent of CHPlus-enrolled children were male and most children were white.
Although 17% of children who were <13 years old and living in families with incomes below 160% of the federal poverty level were black (according to Current Population Survey data), only 9% of CHPlus-enrolled children were black. Similar proportions of CHPlus-enrolled children and children <13 years old and living in families with incomes below 160% of the federal poverty level in the study area were Hispanic (5%). A large proportion of parents of CHPlus-enrolled children in the study area had a high school or higher level of education (88%). Eighty-nine percent of CHPlus-enrolled children had at least 1 employed parent. Sixty-four percent of CHPlus-enrolled children lived in families with annual incomes below 160% of the federal poverty level.
Table 2 shows health insurance characteristics of children enrolled in CHPlus. Twenty-one percent of children in the study area who enrolled in CHPlus were uninsured before enrollment. Nineteen percent were reported to be uninsured since birth in the study area. Sixty-one percent of CHPlus-enrolled children in the study area had a gap of 1 or more months, but only 38% had a >6-month gap in insurance before CHPlus. Thirty-nine percent of children had no gap in coverage, and enrolled in CHPlus immediately after losing or dropping their prior insurance coverage.
Parents in the study area were asked to state the reasons for the loss of their child's prior insurance. Reasons included: loss of parents' job (27%), income too low to afford private insurance but too high for Medicaid (12%), premium of prior insurance too costly (24%), and desire for better/different health insurance (5%).
Parents also were asked about insurance coverage concurrent with CHPlus: 3% of children were reported to be receiving CHPlus and Medicaid simultaneously; 5% had concurrent commercial insurance; and <2% had concurrent full insurance (covering both ambulatory and inpatient care). Thirty-five percent had concurrent inpatient coverage only (at the time of this study, CHPlus did not cover inpatient care, but has since expanded benefits to include inpatient services.13)
We asked parents several questions about their child's health status, including general health status, specific questions about functional health, and the presence of specific chronic conditions.7Table 3 shows the health status of this CHPlus-enrolled population. Two-thirds were reported to be in excellent health. Only 4.1% had fair or poor health, which is similar to the health status found on a nationally-representative population of uninsured children.14 More than two-thirds of children had no chronic conditions, and the proportion of children with 1 or more chronic conditions was also similar to findings from national samples.14 Parents were asked to describe their qualitative impressions of the relative health of their child compared with other children (Table 3) and most parents viewed their CHPlus-enrolled child as relatively healthy.
The Uninsured Rate in the Study Area
The school-based survey about health insurance coverage in Livingston County (N = 5807) and in the city of Rochester (N = 3203) resulted in response rates of 70% (4066/5807) and 53% (1694/3203) respectively. In Livingston County, 5.3% (standard error [SE]: .5%) of surveyed children were uninsured and in the Rochester City School District 3.8% (SE: .4%) of surveyed children were uninsured. In the Livingston County sample, 9.3% of children were reported to be covered by Medicaid, while in the Rochester City School District sample, >50% of children were reported to be covered by Medicaid.
Using pooled Current Population Survey data from 1989 through 1993, we estimated health insurance coverage for children <13 years old in the study area: 4.1% (SE: .9%) were uninsured, 17.5% (SE: 1.7%) had Medicaid, and 82.2% (SE: 1.6%) had some type of private insurance. These results do not sum to 100% because of dual insurance coverage for some children.
In summary, we obtained 3 estimates of the uninsured rate in the study area, which yielded the following results:
Current Population Survey ⇒ 4.1%
Livingston County ⇒ 5.3%
City of Rochester School District ⇒ 3.8%
Participation Rate in CHPlus
Based on pooled Current Population Survey data (1989–1993), approximately 259 000 children <13 years old resided in the study area. An estimated 4.1% (SE: .9%) were uninsured. Thus, an estimated 10 619 children were eligible for, but not enrolled in CHPlus in the study area. The total number of eligible children, then, is given by 10 619 (eligible but not enrolled) + 5876 (enrolled) or 16 495, and the implied participation rate is 36%. Using the Livingston County and the City of Rochester School District estimates of the uninsured rates, the estimated total number of eligible children are 19 603 and 15 718, respectively, and the implied participation rates are 37% and 30%, respectively.
An estimated 64 000 adolescents 13 to 19 years old also resided in the area and their uninsured rate was 4.7% (SE: 1.3%) resulting in an estimated 3008 additional adolescents who would have been eligible had CHPlus been expanded to include adolescents (as is the case for SCHIP). In fact, in 1997 CHPlus eligibility was expanded to age 19.13
New York State's health insurance program for low-income children—CHPlus—enrolled a significant number of age- and income-eligible children who were or would have become uninsured within the study area. The participation rate 2 years after implementation did, however, remain modest.
Enrollment in CHPlus was voluntary and therefore selective. Black children were underrepresented in CHPlus. During the study period, several neighborhood health centers, serving a significant segment of black children in the study area, did not participate in CHPlus. In addition, child advocates, in the community, expressed concerns that parents of children no longer eligible for Medicaid were not receiving assistance or information about CHPlus and may not therefore have known about the program. Indeed, most families reported learning about the program by word of mouth. These findings point to the need for effective, community-based outreach efforts, particularly targeting minority populations. Providing information about SCHIP to outreach workers in community agencies, churches, and stores frequented by underrepresented populations is essential.
Because nearly a third of families indicated that they learned about CHPlus from their physician, health care providers may be an effective source to identify eligible children and enroll them. One possible mechanism to systematically identify uninsured children within a primary care practice is to use a practice's computerized billing file. Using such available technology, providers could identify and contact uninsured patients and inform them about SCHIP programs.
Although few families learned about CHPlus from their prior insurer or employer, most children had an employed parent, and a significant proportion of children became uninsured because their parent lost employer-based health insurance benefits. This suggests that providing information about SCHIP through personnel/benefits offices would be an effective way of disseminating information to potentially eligible families.
Voluntary enrollment in CHPlus did not appear to result in enrollment of either a very high-risk or low-risk patient population. The health status and chronic health conditions of CHPlus-enrolled children were similar to that of the general child population.14 This may have accounted for the rather modest per-patient health care costs during CHPlus that are described in an accompanying article.15
This study has some limitations with respect to the profile of the SCHIP population in other states. The results of this study may have somewhat limited generalizability to large metropolitan areas. The uninsured rate was lower in this study area than in the rest of New York State (4.1% vs 9%). A higher proportion of CHPlus enrollees had prior health insurance during the preceding 12 months than was reported in the Florida Healthy Kids Program.16 Demographic and health insurance characteristics of children enrolled in CHPlus in the study area, such as a relatively low baseline uninsured rate, relatively few children who were uninsured before enrollment, and high levels of education, employment, and family income, may have contributed to less dramatic effects of CHPlus for this study population than for other populations of children with less favorable characteristics. Thus, the impact of CHPlus noted in this study17 may in fact underestimate the potential impact in areas having poorer baseline health insurance coverage.
Another limitation of this study and a major problem, in general, is the lack of data that are available to estimate and characterize uninsured populations at local levels. Although we pooled 5 years ofCurrent Population Survey data to investigate the demographic characteristics of uninsured children in the 6-county study area, estimates were too imprecise to report.
Finally, national surveys are not ideal for estimating the number of uninsured children in many communities. In the 6-county study area, however, pooled Current Population Survey estimates of the uninsured corresponded closely to 2 separate local population-based surveys. One means of monitoring and tracking the implementation of SCHIP is to use a combination of estimates from secondary data and local population-based surveys.
In summary, we conclude that, although participation rates in CHPlus were rather modest at the time of the study, children who enrolled in CHPlus tended to be healthy, from working, relatively educated parents who often experienced a job loss, life change, or financial change resulting in loss of prior health insurance coverage. Improved outreach efforts and systematic strategies to reach eligible populations are needed to increase participation in SCHIP programs.
- SCHIP =
- State Children's Health Insurance Program •
- CHPlus =
- Child Health Plus •
- SE =
- standard error
- ↵US General Accounting Office. Health Insurance for Children: Declines in Employment-based Coverage Leaves Millions of Uninsured; State and Private Programs Offer New Approaches. Washington, DC: US General Accounting Office; 1997. Publication No. GAO/T-HEHS-97-105
- ↵US General Accounting Office. Tax Policy: Health Insurance Tax Credit Participation Rate Was Low. Washington, DC: US General Accounting Office; 1994. Publication No. GAO/GGD-94-99
- ↵US General Accounting Office. Health Insurance for Children: Many Remain Uninsured Despite Medicaid Expansion. Washington, DC: US General Accounting Office; 1995. Publication No. GAO/HEHS-95-175
- ↵US General Accounting Office. Health Insurance for Children: State and Private Programs Create New Strategies to Insure Children. Washington, DC: US General Accounting Office; 1996. Publication No. GAO/HEHS-96-35
- ↵New York State Public Health Law, Article 25, Title I—A Child Health Insurance Plan (Chapters 922 and 923 of the Laws of 1990) as amended
- Szilagyi PG,
- Shone LP,
- Holl JL,
- et al.
- ↵US Bureau of the Census. Current Population Survey. Washington DC: US Bureau of the Census; 1994
- ↵Huber PJ. The behavior of maximum likelihod estimates under non-standard conditions. In: Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability. Berkeley, CA: University of California Press; 1967:221–233
- Szilagyi, PG, Holl JL, Rodewald LE, et al
- ↵Trafton S, Shone LP, Zwanziger J, et al. Evolution of a children's health insurance program: lessons from New York State's Child Health Plus. 2000;105(suppl):692–696
- Zwanziger J,
- Mukamel DB,
- Szilagyi PG,
- et al.
- Shenkman E,
- Bucciarelli R,
- Hope Wegener D, Naff R, Freedman S
- Holl JL,
- Szilagyi PG,
- Rodewald LE,
- et al.
- Copyright © 2000 American Academy of Pediatrics