PEDIATRICS Vol. 118 No. 4 October 2006, pp. 1676-1682 (doi:10.1542/peds.2006-0004)
ARTICLE |
The Impact of Public Insurance Expansions on Children’s Access and Use of Care
a Department of Social and Behavioral Sciences
b Institute for Health Policy Studies
c Center on Social Disparities in Health, University of California, San Francisco, California
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
OBJECTIVE. Our goal was to examine the impact of the State Children’s Health Insurance Program nationally on children’s access and use of health care.
OBJECTIVE. Our data source was the National Health Interview Survey, using 1997 as a baseline, which predates the implementation of the State Children’s Health Insurance Program, and 2003 as the end point of the analysis. We analyzed 25 734 children aged 0 to 18 years (1997 and 2003 combined) to examine changes in health insurance coverage rates, health care access, and utilization for children in the State Children’s Health Insurance Program target population, defined here as those living in families with incomes between 100% and 199% of the federal poverty level.
RESULTS. Children in the State Children’s Health Insurance Program target income group showed the largest reduction in rates of uninsurance among 3 income groups (<100%, 100%–199%, and 
200% of the federal poverty level) between 1997 and 2003 (15.1%–8.7%). Significant reductions occurred in the proportion of children without a usual source of care in the target income group (9.4%–7.3%) and in the proportion of children without a provider visit in the past year (10.8%–9.8%). Other measures (unmet needs, delayed care, volume of provider visits, receipt of well-child care, and dental care) showed no significant changes over this time period. A separate multivariate analysis restricted to the State Children’s Health Insurance Program target population in 2003 showed that children with continuous public coverage had significantly better access and utilization on all measures studied when compared with uninsured children and performed as well or better than children with continuous private coverage.
CONCLUSIONS. Implementation of the State Children’s Health Insurance Program is associated with substantial gains in public coverage for children in the target income group. Although some of these gains were offset by losses in private coverage, our findings demonstrate that public health insurance provides significant benefits in terms of access and utilization for children living in the target income group.
Key Words: insurance • medicaid • SCHIP • access • use
Abbreviations: SCHIP—State Children’s Health Insurance Program • FPL—federal poverty level • NHIS—National Health Interview Survey • NCHS—National Center for Health Statistics
The State Children’s Health Insurance Program (SCHIP), enacted in 1997, represents the largest single expansion of health insurance coverage for children in the United States in >30 years.1 In 1997, there were >11 million uninsured children in the United States representing 15% of the pediatric population.2 SCHIP was established to reduce the number of uninsured children and targets children from birth to 18 years of age primarily living in near-poor families whose incomes exceed Medicaid eligibility limits but fall below 200% of the federal poverty level (FPL).*
The Balanced Budget Act of 1997, which created SCHIP, allocated $40 billion over 10 years to extend coverage to low-income, uninsured children who were ineligible for Medicaid.3 The SCHIP legislation allowed states to provide coverage by expanding their existing Medicaid program, creating a freestanding separate insurance program more like employer-sponsored private insurance, or a combination of these approaches. States that chose to expand their Medicaid program were required to adhere to the eligibility and benefits required under existing federal Medicaid regulations. States that created separate programs were allowed greater flexibility in program eligibility, benefits, and cost sharing and could design benefit packages less generous than required under Medicaid.
Initial federal eligibility requirements were designed to cover 
5 million of the >11 million uninsured children. By the completion of fiscal year 2003, the number of children ever enrolled in SCHIP rose to 5.8 million.4 Enrollment has grown most rapidly in those states that chose to create separate SCHIP programs. Thirty-five states are currently operating either separate SCHIP programs or a combination of separate SCHIP and Medicaid expansion programs.4 Although Medicaid serves many more children than SCHIP, the enhanced level of federal matching funds along with the state autonomy allowed in program design and implementation has generated considerable interest and support for SCHIP among states.5
The SCHIP legislation was motivated by the belief that expanding health insurance coverage would increase access to health care services, raise the use of preventive medical services, and improve the health of children.6 Uninsured children have decreased access to health care services, often use more expensive medical services, and have more serious health problems than children in privately or publicly insured programs.7,8 Uninsured children are more likely to lack a usual source of care, have unmet health needs, delay seeking health care, have lower immunization rates, and use fewer well-child care services than insured children.9,10 Measuring the impact of health insurance on health has proved challenging for researchers.10–12
Recent state studies after the implementation of SCHIP have reported improvements in access to care, changes in the patterns of health care with a greater proportion of health care provided within the usual source of primary care, and improved continuity and quality of care for children in the target group.13–16 These state-level analyses were based on interviews with parents of SCHIP enrollees in large and small states and demonstrate a larger effect than found in national samples.17,18 Although encouraging, the level of unmet health needs remained high (18.8% in New York state) even after SCHIP.15
The purpose of this article is to present an analysis of data from the National Health Interview Survey (NHIS) examining the impact of SCHIP on children’s access to care and use of health care services. This is the first study to assess trends in coverage, access, and use for the SCHIP target population using nationally representative survey data. We used 1997 data, which predates the implementation of SCHIP, as the baseline for our analyses, and 2003 as the end point of the analysis. Our study also uses a broad set of access and use measures from the NHIS that are not available in other data sets. We hypothesized that at the 6-year point, the SCHIP target population of children in families with incomes between 100% and 199% of the FPL would experience greater improvements in access and use compared with their peers in higher income groups. We also expected to see gains in access and use for children living in poor families (incomes <100% of the FPL) for 2 reasons. First, the final phase-in of Medicaid eligibility requiring all states to provide coverage to 100% of the FPL occurred during the study period. Second, SCHIP enrollment efforts had a halo effect, such that the outreach activities for SCHIP often resulted in increased rates of Medicaid enrollment by children with family incomes below the SCHIP income eligibility floor.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Data Source
The NHIS is a continuing nationally representative household survey sponsored by the National Center for Health Statistics (NCHS).19 The annual sample includes
40000 households with 30000 children <18 years of age.20 Trained personnel from the US Bureau of the Census conducted face-to-face interviews with an adult caretaker, usually the mother, who provided proxy responses for children <17 years of age in the survey. Seventeen-year-olds were permitted to respond for themselves. The survey instrument was redesigned in 1997 and consists of core questionnaires for household, family, person, and child. We used data from the 1997 and 2003 family core files and sample child files for children <18 years of age for our SCHIP analysis. Missing income data are a problem in all of the household surveys. To address this issue, NCHS has developed a set of data files with imputed values for cases with missing income data. We used these files in our analysis. The sample included 13579 children in 1997 and 11900 children in 2003. The final response rates for the 1997 sample child and family surveys were 84.1% and 90.3%, respectively; the respective response rates for 2003 were 81.1% and 87.9%, respectively. Table 1 reports sample sizes and weighted national population estimates.
|
Theoretical Model
The behavioral model of health services use by Aday and Andersen21 was used as the conceptual framework for selecting independent variables for our multivariate analysis of access to care and use of health care services. The Aday and Andersen model postulates the use of health services as being dependent on factors of an individual’s predisposition to use the health services, the individual’s level of need for health services, and the presence of factors that enable or facilitate the use of health care services.
Independent Variables
The predisposing factors included in our analysis were age, gender, race and ethnicity, family structure, family size, and education of the family reference person in the household. Measures of need included perceived health status, and enabling factors included poverty status, health insurance status, region, and place of residence.
Measures of Access and Use
We used the sample child data to obtain information on access to health care as indicated by the presence of 1 unmet health need (including prescription costs, mental health, and dental care costs), a delay in seeking health care services, and absence of a usual source of care. For information on the use of health care services, data from the sample child and family core provided measures of the number of health provider visits in the past year, absence of any health provider visits in the past year, absence of any dental visits in the past year, and absence of any preventive well-child care visits in the past year. Item nonresponse rates for the independent and dependent variables did not exceed 5%.
Statistical Analysis
We used 2 methods of analyzing the data to demonstrate the impact of SCHIP. First, we used a predesign and postdesign to assess the impact of SCHIP on access and use for the target population of children in families with incomes between 100% and 199% of the FPL. This analysis used 1997 NHIS data as a baseline, which predate the implementation of SCHIP, and 2003, the 6-year point of the program, as the end point. Second, we used multivariable analyses to assess the impact of SCHIP within the target population during 2003. Specifically, multivariable models were used to compare children with full-year public coverage to those with full-year private coverage, part-year coverage, and full-year uninsured while controlling for confounding variables, including child’s age, gender, race and ethnicity, family structure, family income, region and place of residence, parental education, health insurance, and parental report of health status. All of the analyses were conducted using the SUDAAN statistical software. This statistical software program adjusts for the complex sample design of the NHIS in computing SEs.22
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Changes in Health Insurance Coverage From 1997 to 2003
Overall, there was a significant reduction in the proportion of US children without health insurance between 1997 and 2003 (Table 1). In 1997, before the enactment of SCHIP, 8.4% of all children were uninsured for the entire year. By 2003, that proportion had fallen to 5.3% (P < .001). Children living in families with incomes between 100% and 199% of the FPL, the SCHIP target income population, showed the largest percentage reduction in rates of uninsurance among all of the income groups (15.1%–8.7%; P < .001). Children living in families with incomes <100% of the FPL also saw gains, with the rate of uninsurance declining from 14.5% to 10.2% between 1997 and 2003 (P < .001). A more modest but still significant gain occurred for children in families with incomes
200% of the FPL. All of the reductions in uninsurance are attributable to growth in public insurance coverage. Whereas children in all 3 poverty categories gained public coverage, children living in families with incomes between 100% and 199% of the FPL showed the most substantial percentage gains in obtaining full-year public coverage, growing from 18.7% to 39.5% between 1997 and 2003 (P < .001). Some of these gains in coverage for children were offset by substantial erosion in private coverage. Children living in families with incomes between 100% and 199% of the FPL experienced the largest percentage decline in private health insurance, from 53.5% to 40.9% between 1997 and 2003 (P < .001).
We also assessed the prevalence of part-year coverage to determine whether changes in rates of insurance discontinuity occurred between study years and assessed the impact of partial-year coverage on access to health care and use of health care services. Partial-year coverage includes children with public and/or private coverage who were covered less than the full year before the interview date. Children living in families with incomes <100% of the FPL experienced a significant decrease in partial-year coverage between 1997 and 2003 (13.0% in 1997 and 9.3% in 2003; P < .001). Children in the SCHIP target income group of 100% to 199% of the FPL had a higher prevalence of partial-year coverage in 2003 than those <100% of the FPL and showed no significant change over the study period (12.7% in 1997 and 10.9% in 2003). The prevalence of partial-year coverage among children living in higher income groups was much lower and showed no significant change over the study period (4.3% in 1997 and 4.7% in 2003).
Changes in Access and Use of Services
Improvements in the measures of access to health care for children on the whole have occurred since the implementation of SCHIP, although the effects are small. As shown in Table 2, between 1997 and 2003, a significant reduction occurred in the proportion of children without a usual source of care in the target income group of 100% to 199% of the FPL (9.4% in 1997 and 7.3% in 2003; P < .05). Children living in families with incomes <100% of the FPL had a similar decline in the proportion of children without a usual source of care (11.3% in 1997 and 8.9% in 2003; P < .05). No significant changes occurred in the other access measures for children in the SCHIP target income group over the study period.
|
There were also some small changes in the use of health care services for children as a group. There was a significant reduction in the proportion of children overall without a provider visit in the past year, from 10.8% in 1997 to 9.8% in 2003 (P <. 05), although no significant change occurred in children in the SCHIP target income group over the study period. No significant changes occurred in the use measures for children in the SCHIP target income group.
To ensure that the findings described above were not subject to the influence of confounding, we replicated the bivariate analyses using multivariable analysis. The results from the multivariable analyses were consistent with those of the bivariate analyses. Results for the adjusted analyses are available on request from the corresponding author.
The Role of Insurance
In the SCHIP target income group, we found an absence of substantial improvements in access and use for children, which may have been attributable to the gains in public coverage being diluted by the losses in private coverage seen in Table 2. Hence, we undertook a closer examination of the effect of public insurance coverage for children in the SCHIP target income group using multivariable analysis. Table 3 presents the results of this analysis for children in families with incomes between 100% and 199% of the FPL in 2003. In a series of logistic and ordinary least-squares regression analyses, we compared access and use outcomes for children with continuous public coverage (the SCHIP target group) to children with continuous private coverage, children with part-year private or public coverage, and continuously uninsured children. We controlled for race and ethnicity, age, gender, family composition, family income, region and place of residence, health status, and education level of parent.
|
Our analysis shows that for virtually all of the access and use measures, children with full-year public coverage experience at least as good and sometimes better outcomes than children with full-year private coverage. In contrast, children in the SCHIP target income group without health insurance during the past year were 5 times more likely to experience an unmet health need, 15 times more likely to experience a delay in seeking health care, and 11 times more likely to lack a usual source of care than children with full-year private coverage. Near-poor children without health insurance coverage during the past year were also far more likely to have gone without preventive well-child care or a dental visit during the year.
Effect sizes for the group of children who were uninsured for part of the year were not as strong as those for full-year uninsured children but were nevertheless substantial in many cases. Taken together, these results indicate that continuous public insurance coverage is associated with significant and substantial improvements in access and use when compared with no coverage or partial-year coverage.
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
A large proportion of previously uninsured children gained health insurance coverage as a result of SCHIP. Children living in families with incomes between 100% and 199% of FPL showed the largest percentage reduction in rates of uninsurance among all of the income groups. The increase in public coverage among poor children can be attributed in part to the outreach efforts under SCHIP, which identified increasing numbers of Medicaid-eligible children after the enactment of SCHIP.
The pattern of decline in private health insurance coverage presented in our findings is a continuation of a larger secular trend of reduced provision of private insurance for children’s health care. Private insurance coverage for children on the whole has declined from a high of 69.8% in 1977 to 48.5% in 2001.23 Although losses in private coverage began well before the implementation of SCHIP, some analysts and policy-makers have argued that expanded public insurance has contributed to losses of private coverage for children.23–25 Regardless, SCHIP has played the important role of preventing further erosion in health insurance coverage averting an even greater number of uninsured children, particularly in working poor families.
Our findings demonstrate that children in the SCHIP target income group without health insurance coverage remain at significant risk. They are much more likely to have an unmet health need, a delay in seeking health care, and to lack a usual source of health care than those with full-year public coverage. One of the unique contributions of our analysis is the finding that public coverage is as protective as private coverage if it is available on a continuous basis. In 2003, children with full-year public coverage were on a par with those children with full-year private coverage in all measures except for the absence of a provider visit. These findings suggest that SCHIP may play an important role in narrowing health disparities over time.
The NCHS recently published a study on trends in health insurance coverage examining data from a similar time period as our analysis.18 Our study examined a greater number of access and use outcome measures using multivariable analysis to adjust for confounding and imputation to account for the high rate of item nonresponse for family income.
This is a cross-sectional study, and the estimated effects of SCHIP on children before and after enrollment may be influenced by factors not controlled for in the analysis. These factors could include parental error in recall data, bias because of nonresponse, and secular trends unrelated to SCHIP that may have affected the outcome measures. In particular, the design of our 1997–2003 comparison (sometimes referred to as an untreated comparison group design with separate pretest and posttest samples) is not as powerful as a design using the same individuals in the pretest and posttest periods. Our design implicitly assumes that no other major changes or "shocks" occurred to the treatment and comparison income groups other than the insurance expansions.
Although the NHIS is designed to be representative of the US civilian population, children who are institutionalized or reside in correctional facilities, hospitals, and long-term care facilities are excluded. Homeless and other transient children without a permanent address are also excluded from the survey. These children are more likely to experience greater difficulties in access to health care and use of health care services than children included in the NHIS sample. This may result in our estimates of access difficulties being somewhat understated in both time periods.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
SCHIP has been highly successful in reducing the number of uninsured children in the United States since its introduction. Implementation of SCHIP is associated with substantial gains in public coverage for children in the target income group. Although some of the gains were offset by losses in private coverage, our findings demonstrate that public health insurance is a significant protective factor for children living in the target income group. Importantly, the analytic work that we reported here will also provide a template for continued monitoring of health care access and use for children in the SCHIP target population. This will provide long-term value by providing a baseline for assessing the impact of future changes in public programs and private health insurance.
|
ACKNOWLEDGMENTS |
|---|
This study was supported by funding provided by the Centers for Disease Control and Prevention (CDC/ATPM Cooperative Agreement TS-0842). Additional funding was provided by a cooperative agreement from CAN-DO: Center to Address Disparities in Children’s Oral Health, sponsored by the National Institute for Dental and Craniofacial Research and the National Center on Minority Health Disparities at the National Institutes of Health (U54 DE14251).
We thank Lena Libatique for her assistance in the preparation of this article.
|
FOOTNOTES |
|---|
Accepted Apr 19, 2006.
Address correspondence to Karen G. Duderstadt, PhD, RN, PNP, Department of Social and Behavioral Sciences, University of California San Francisco, 3333 California St, Suite 455, San Francisco, CA 94118. E-mail: karen.duderstadt{at}nursing.ucsf.edu
The authors have indicated they have no financial relationships relevant to this article to disclose.
* A yearly income of $14995 for a family of 4 was the FPL guideline when SCHIP was established. Poverty level in 2005 is expressed as a yearly income of $19350 for a family of 4.26 States commonly use 200% of the FPL as the cap for SCHIP eligibility, although some states set slightly higher or lower limits. 
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
- Centers for Medicare and Medicaid Services. The State Children’s Health Insurance Program Annual Enrollment Report. Washington, DC: Centers for Medicare and Medicaid Services; 2002
- Feder J, Uccello C, O’Brien E. The Difference Different Approaches Make: Comparing Proposals to Expand Heath Insurance. Menlo Park, CA: The Henry J Kaiser Foundation; 1999
- Shi L, Oliver TR, Huang V. The children’s health insurance program: expanding the framework to evaluate state goals and performance. Milbank Q. 2000;78 :403 –446, 340–341.[CrossRef][Web of Science][Medline]
- Center for Medicare and Medicaid Services. State Children’s Health Insurance Program Enrollment Report. Washington, DC: Center for Medicare and Medicaid Services; 2004
- Moore J. SCHIP in the Formative Years: An Update. Washington DC: National Health Policy Forum; 2000
- Kaestner R, Joyce TJ, Racine AD. Medicaid eligibility and the incidence of ambulatory care sensitive hospitalizations for children. Soc Sci Med. 2001;52 :305 –313[CrossRef][Web of Science][Medline]
- Byck GR. A comparison of the socioeconomic and health status characteristics of uninsured, state children’s health insurance program-eligible children in the United States with those of other groups of insured children: Implications for policy.
Pediatrics. 2000;106; 14
–21
[Abstract/Free Full Text] - Cunningham P. Targeting communities with high rates of uninsured children. Health Aff (Millwood). 2001:W20 –W29
- Newacheck PW, Hughes DC, Cisternas M. Children and health insurance: an overview of recent trends. Health Aff (Millwood). 1995;14 :244 –254.[Abstract]
- Newacheck PW, Stoddard JJ, Hughes DC, Pearl M. Health insurance and access to primary care for children.
N Engl J Med. 1998;338
:513
–519
[Abstract/Free Full Text] - Guendelman S, Pearl M. Children’s ability to access and use of health care.
Health Aff (Millwood). 2004;23; 235
–244
[Abstract/Free Full Text] - Racine AD, Kaestner R, Joyce TJ, Colman GJ. Differential impact of recent Medicaid expansions by race and ethnicity.
Pediatrics. 2001;108; 1135
–1142
[Abstract/Free Full Text] - Dick AW, Brach C, Allison RA, et al. SCHIP’s impact in three states: how do the most vulnerable children fare? SCHIP improved access to and satisfaction with care for all enrollees, even the most vulnerable.
Health Aff (Millwood). 2004;23; 63
–75
[Abstract/Free Full Text] - Holl JL, Szilagyi PG, Rodewald LE, et al. Evaluation of New York State’s Child Health Plus: Access, utilization, quality of health care, and health status. Pediatrics. 2000;105(3 suppl E) :711 –718
- Szilagyi PG, Dick AW, Klein JD, et al. Improved access and quality of care after enrollment in the New York State Children’s Health Insurance Program (SCHIP).
Pediatrics. 2004;113; e395
–e404
[Abstract/Free Full Text] - Damiano PC, Willard JC, Momany ET, Chowdhury J. The impact of the Iowa S-SCHIP program on access, health status, and the family environment. Ambul Pediatr. 2003;3 :263 –269[CrossRef][Web of Science][Medline]
- Cohen RA, Coriaty-Nelson Z. Health Insurance Coverage: Estimates From the National Health Interview Survey, 2003. Washington, DC: The National Center for Health Statistics; 2004
- Cohen RA, Bloom B. Trends in Health Insurance and Access to Medical Care for Children Under 19 Years: United States, 1998–2003. Atlanta, GA: Centers for Disease Control and Prevention; 2005: April 15, Report No. 355
- Schoenborn CA, Adams PF, Schiller JS. Summary Health Statistics for the US Population: National Health Interview Survey, 2000. Washington, DC: US Department of Health and Human Services; 2003
- National Center for Health Statistics. National Health Interview Survey. Hyattsville, MD: National Center for Health Statistics; 2003
- Aday LA, Andersen R. A framework for the study of access to medical care. Health Serv Res. 1974;9(3) :208 –220
- Research Triangle Institute. SUDAAN User’s Manual, Release 8.0. Research Triangle Park, NC: Research Triangle Institute; 2001
- Cunningham P, Kirby J. Children’s health coverage: a quarter-century of change.
Health Aff (Millwood). 2004;23
:27
–38
[Abstract/Free Full Text] - Cunningham P, Reschovsky J, Hadley J. SCHIP: Medicaid Expansions Lead to Shifts in Children’s Coverage. Washington, DC: Center for Studying Health Systems Change; 2002
- Lo Sasso AT, Buchmueller TC. The effect of the state children’s health insurance program on health insurance coverage. J Health Econ. 2004;23 :1059 –1082[CrossRef][Web of Science][Medline]
- Department of Health and Human Services. 2005 HHS Poverty Guidelines. Washington, DC: Department of Health and Human Services; 2005: Report No. 33
PEDIATRICS (ISSN 1098-4275). ©2006 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  J. R. Cummings, S. A. Lavarreda, T. Rice, and E. R. Brown The Effects of Varying Periods of Uninsurance on Children's Access to Health Care Pediatrics, March 1, 2009; 123(3): e411 - e418. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. K. Hoilette, S. J. Clark, A. Gebremariam, and M. M. Davis Usual Source of Care and Unmet Need Among Vulnerable Children: 1998-2006 Pediatrics, February 1, 2009; 123(2): e214 - e219. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. H. Brickhouse, R. G. Rozier, and G. D. Slade Effects of Enrollment in Medicaid Versus the State Children's Health Insurance Program on Kindergarten Children's Untreated Dental Caries Am J Public Health, May 1, 2008; 98(5): 876 - 881. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
