 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
SPECIAL ARTICLE |
a Department of Pediatrics, University of Rochester School of Medicine and Dentistry, Rochester, New York
b Department of Community and Preventive Medicine and Strong Children's Research Center, University of Rochester School of Medicine and Dentistry, Rochester, New York
d the School of Nursing, University of Rochester School of Medicine and Dentistry, Rochester, New York
c School of Public Health, University of Illinois at Chicago, Chicago, Illinois
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
OBJECTIVES. We sought to assess the impact of New York's State Children's Health Insurance Program (SCHIP) on health care for children with asthma.
DESIGN. Parents of a stratified random sample of new enrollees in New York's SCHIP were interviewed by telephone shortly after enrollment (baseline, n = 2644 [74% of eligible children]) and 1 year later (follow-up, n = 2310 [87%]). Asthma was defined by parent report using questions based on National Heart, Lung, and Blood Institute criteria. A comparison group (n = 401) who enrolled in SCHIP 1 year later was interviewed as a test for secular trends.
MAIN OUTCOME MEASURES. Access (having a usual source of care [USC], unmet health needs, problems receiving acute asthma care), asthma-related medical visits, quality (continuity of care at the USC, problems receiving chronic asthma care, use of antiinflammatory medications), and asthma outcomes (change in asthma care or severity) were the main outcome measures used. Bivariate and multivariate analyses compared measures at baseline (year before SCHIP) versus follow-up (year during SCHIP).
RESULTS. Three-hundred eighty-three children (14%) had asthma at baseline, and 364 had asthma at follow-up (16%). No secular trends were detected between the baseline study group and the comparison group. After enrollment in SCHIP, improvements were noted in access: lacking a USC (decrease from 5% to 1%), unmet health needs (48% to 21%), and problems getting to the USC for asthma (13 to 4%). Children had fewer asthma-related attacks and medical visits after SCHIP (mean number of attacks: 9.5 to 3.8: mean number of asthma visits: 3.0 to 1.5; hospitalizations: 11% to 3%). Quality of asthma care improved for general measures (most/all visits to USC: 53% to 94%; mean rating of provider: 7.9 to 8.8 of 10) and asthma-specific measures (problems getting to the USC for asthma care when child was well: 13% to 1%). More than two thirds of the parents at follow-up reported that both quality of asthma care and asthma severity were "better or much better" than at baseline, generally because of insurance coverage or lower costs of medications and medical care.
CONCLUSIONS. Enrollment in New York's SCHIP was associated with improvements in access to asthma care, quality of asthma care, and asthma-specific outcomes. These findings suggest that health insurance improves the health of children with asthma.
Key Words: SCHIP • health insurance • children • asthma
Abbreviations: NHLBI—National Heart, Lung, and Blood Institute • SCHIP—State Children's Health Insurance Program • FPL—federal poverty level • USC—usual source of care • ED—emergency department
Asthma is the most prevalent chronic medical condition of childhood1,2 and affects >6 million US children annually3 with substantial morbidity4–6 and cost.7 Low-income and minority children have the greatest morbidity from asthma8–13 and are less likely to receive asthma care6 that prevents asthma exacerbations.12,14–18
Health care–system, provider, and family barriers to asthma care have been identified.19 Health care–system barriers include lack of health insurance,20–22 primary care,22 care coordination,23 or acute care services24,25 and the high cost of asthma medications and services.26–29 Health care–provider factors include lack of recognition of the child's asthma severity30,31 and suboptimal compliance with recommended guidelines for asthma management.32–37 Family-based barriers38 include confusion about asthma symptoms39,40 and therapies.41–44
Recent models of care management for chronic disease45–47 have pointed out that all 3 components (system, provider, and family) are important in improving outcomes among patients with chronic disease such as asthma. Most studies of childhood-asthma interventions have evaluated provider-level strategies48–55 or family-based interventions56; few have evaluated health care–system changes such as provision of health insurance. A study in the mid-1990s of a state pilot health insurance program for previously uninsured children noted some improvements but still suboptimal levels of access to asthma care after the provision of insurance coverage.57 At the time of that study, new childhood-asthma guidelines from the National Heart, Lung, and Blood Institute (NHLBI) were just being disseminated.58 Given recent improvements in our understanding of childhood asthma as well as widely distributed and measurable guidelines,59,60 it is now possible to better assess the degree to which provision of health insurance plays a role in improving childhood-asthma care.
The State Children's Health Insurance Program (SCHIP) provided a unique opportunity to examine these questions. SCHIP was enacted in 1997 as Title XXI of the Social Security Act61 to provide health insurance coverage to low-income children who are neither eligible for Medicaid nor covered by private health insurance. Studies of SCHIP have noted improved access to health care among the enrolled population.62–67 However, little is known about the impact of SCHIP on children with chronic diseases68 such as asthma. This study assessed the impact of New York's SCHIP on access, utilization of services, quality of care, and asthma-related outcomes among children with asthma.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Setting: New York's SCHIP
In 2001, New York had 590000 SCHIP enrollees (18% of US SCHIP enrollees).69 New York's SCHIP was provided by 32 managed care plans.70,71 Children were eligible if they were 0 to 18 years old, at or below 230% of the federal poverty level (FPL), New York residents, not covered by other insurance, and not eligible for Medicaid. Monthly premiums varied by income, from none (for those at <160% of the FPL) to $9 to $15 per child (family maximum of $45 for 
3 children). A few families at >230% of the FPL purchased SCHIP for the full monthly premium. The state-defined benefit package included ambulatory, emergency, inpatient, and prescription medications, which are similar to benefits of commercial plans but less comprehensive than those available through New York's Medicaid plan.72 Coverage of nebulizer machines and spacers varied by insurer.
Study Design
We used a baseline versus follow-up–cohort design.66,67 We interviewed parents of New York's SCHIP enrollees shortly after enrollment (baseline interview) in 2001 and again 13 months after enrollment (follow-up interview). The baseline interviews reflected the child's experience in the 1-year period before SCHIP, and the follow-up interviews reflected the first year during SCHIP enrollment.
Study Subjects
New York's SCHIP administrative files were analyzed to identify new enrollees. A stratified random sample (1 child per family) was selected from 4 geographic regions, 3 age groups (0–5, 6–11, and 12–18 years), and 3 race/ethnicity groups based on parent self-report (white non-Hispanic, black non-Hispanic, and Hispanic; other groups were excluded).73 The data were weighted to account for this complex sampling design so that estimates were representative of SCHIP enrollees statewide.
We also interviewed parents of a separate comparison group of children66 who enrolled in SCHIP 1 year after the study group enrolled. This comparison group comprised a simple random sample of SCHIP enrollees throughout New York State; it was designed to test for secular trends because their pre-SCHIP period was identical to the SCHIP time period for the study group. During the study time period, there were no changes in national asthma guidelines.
Telephone Interviews
The National Opinion Research Center,74 using Computer Assisted Technology, Inc75 (Buffalo Grove, IL) conducted telephone interviews. Baseline interviews were conducted between March 15, 2001, and September 15, 2001, and follow-up interviews were conducted between December 1, 2001, and May 4, 2002. Interviews were conducted in English and Spanish, day and evening, 7 days per week.
General Dependent Measures
Key questions were obtained from standard instruments or developed for the study. Demographic measures included patient age, gender, race/ethnicity, single-parent household, family income, and parent education and employment. Overall health status was measured as excellent, good, fair, or poor.76 Prior health insurance was assessed as the number of months that the child was insured during the year before enrollment and the type of insurance that was held before SCHIP (private, public, none). Access measures included the presence of a usual source of care (USC), accessibility of the USC (4 measures77), and overall and specific unmet health care needs. Utilization measures involved use of health services during the relevant years (before and during SCHIP) for emergency department (ED), outpatient, or hospital care. Quality measures included use of the USC for none, some, most, or all care (a measure of continuity) and parents' ratings of specific aspects of health care and their primary provider using Consumer Assessment of Health Plans items.77
Asthma-Specific Measures
Because there is no gold-standard set of questions for identifying children with asthma, we adapted 5 questions from NHLBI criteria59 that have been used in other studies.6,30,31,57 These questions were asked at both baseline and follow-up, because children could "grow into" or "grow out of" asthma over the 2-year study time frame. Questions pertained to either the 12-month period before (baseline) or after enrollment (follow-up). Parents were asked (1) whether a physician said the child has asthma and, for times other than when the child had a cold, whether the child had (2) wheezing or whistling in the chest, (3) chest sounding wheezy during or after exercise, (4) waking from sleep because of coughing, wheezing, or trouble breathing, and (5) wheezing severe enough to limit speech. A positive response to any question identified the child as having asthma and invoked a series of asthma-related questions.
For each time period, parents were asked 3 additional severity questions based on the 3 NHLBI criteria30,31,59 to classify the child's asthma as "mild intermittent" or "mild persistent to severe" (hereafter designated "mild" or "moderate/severe"): (1) frequency of wheezing, coughing, or shortness of breath during a typical week or month; (2) limitation of physical activity resulting from asthma; and (3) nighttime awakening from sleep because of coughing, wheezing, or trouble breathing. These questions were used for asthma-severity classification only.
Asthma-related–dependent measures included (1) access (problems getting care or obtaining medications during an asthma attack and specific reasons for these problems), (2) utilization (hospitalizations and visits for asthma attacks), (3) quality (problems getting asthma chronic-care checkups or medications, use of antiinflammatory medications or asthma action plans), and (4) outcomes (change in severity of the child's asthma and the quality of care for asthma between the 2 years). The last 2 questions were asked only at follow-up. All other questions were asked identically at baseline and follow-up.
Analyses
For bivariate and multivariate analyses, we restricted the analyses to children with asthma.
Comparisons
We present analyses that included all children who were identified as having asthma at baseline and all children who were identified as having asthma at follow-up regardless of their asthma status at the other interview. This analytic strategy was selected because it reflects the reality that children can grow into or out of asthma (or into or out of "moderate/severe asthma") and that preexisting yet undiagnosed asthma could be newly diagnosed after obtaining insurance. It best represents the population of SCHIP enrollees with asthma at either time period. We also compared baseline versus follow-up measures by limiting analyses to (1) those children with asthma who were identified at both baseline and follow-up and (2) the cohort of children who were identified at baseline only. Although each method has different potential strengths and weaknesses, results from all 3 methods were nearly identical.
We compared antiinflammatory medication use and provision of a written asthma action plan before versus during SCHIP coverage. These measures were assessed for all children with asthma (including all levels of asthma severity reported at the time of interview), although at any point in time only children with moderate/severe asthma should have these interventions. We did this to minimize the bias in our estimates of the effect of SCHIP on these measures. If asthma care improved as a result of SCHIP and this improvement left fewer children with moderate/severe asthma symptoms during SCHIP coverage, analyses that included only children with moderate/severe asthma would generate selection biases. By considering all children with asthma, we avoid this concern. Also, our severity classification may not correlate precisely with criteria used by physicians in implementing antiinflammatory medicines or action plans, resulting in misclassification of who should have been prescribed these interventions.
Statistics
In all analyses, we used sample weights to obtain estimated means (and odds ratios) for children with asthma. We used Stata 8.278 to account for the complex sampling design.
We first summarized baseline demographics, health characteristics, and insurance measures (baseline survey) for children with asthma in the SCHIP population. Subsequent analyses compared before SCHIP versus during SCHIP using survey-design–based F tests and t tests. All results shown in the tables are for New York's SCHIP population with asthma.
To assess whether bivariate results of key outcomes were affected by confounding demographic and prior insurance variables, we estimated multivariate logistic-regression models of key unmet needs, continuity at the USC, and asthma-specific quality measures, controlling for the demographic characteristics and prior insurance status that are shown in Table 1. We included an identifier for moderate/severe versus mild asthma and calculated adjusted odds ratios and confidence intervals for follow-up versus baseline separately for each subgroup. We then compared these adjusted results with the unadjusted bivariate results. Generalized estimating equations79 were used to estimate logistic population-averaged models and to account for the correlation between baseline and follow-up responses for same child.
|
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Parents of 401 children completed the comparison-group interview.
Prevalence of Asthma
At baseline, 383 (14%) of 2644 children had asthma. Of these, 155 (40%) had moderate/severe asthma. At follow-up, 364 (16%) of 2310 children had asthma, including 126 (35%) with moderate/severe asthma. Forty-nine children interviewed at baseline were lost to follow-up (and excluded). The final samples comprised 334 at baseline and 364 at follow-up. Altogether, we analyzed 472 children with asthma who were identified at either baseline or follow-up. The prevalence of asthma was not statistically different between the baseline and follow-up periods.
Comparison Group Versus Study Group
Key measures for the pre-SCHIP period were compared for the study group (n = 2644) versus the comparison group (n = 401) to assess potential secular trends. Of 34 measures tested including demographics, access, unmet needs, utilization of outpatient services, and quality of care, statistically significant differences were noted for only 3 measures (results are available from the corresponding author). There was no statistical difference in asthma prevalence: 72 comparison-group children (17.9%) had asthma. We also compared 23 measures of access, unmet needs, and utilization for children with asthma in the baseline study group (n = 334) versus the comparison group (n = 72). There were no significant differences between the 2 groups. These results suggest that there were no major secular trends during this 1-year time span.
Demographic Characteristics of SCHIP Enrollees With Asthma
SCHIP enrollees with asthma were from families with low income, and many children were from minority, single-parent, working households (Table 1). Many children had poor health status.
Access and Utilization of Health Care Among Children With Asthma
The proportion of children with asthma who had no USC (Table 2) decreased during SCHIP coverage from 5% to 1%. It is notable that 45% of the children changed their USC after enrollment in SCHIP. Two indicators of accessibility improved after SCHIP enrollment: travel time of 30 minutes and difficulty getting an appointment at the USC.
|
|
|
|
Multivariate Analyses
None of the results from the multivariate analyses were substantively different from those of the bivariate analyses, which suggests that the changes observed after SCHIP enrollment were not explained by differences in other variables. For example, we calculated the unadjusted and adjusted (for the variables listed in Table 1) odds ratios for having "problems getting asthma medications" (from Table 6) during SCHIP versus before SCHIP separately for children uninsured and insured before SCHIP. For previously uninsured children with severe asthma, the unadjusted and adjusted odds ratios were 0.05 (P < .001) and 0.04 (P = .001), respectively, and for previously insured children, the odds ratios were 0.09 (P < .05) and 0.1 (P < .05), respectively. Similar findings were noted for children with mild asthma, with similar odds ratios for unadjusted and adjusted analyses (if uninsured: 0.16 and 0.12, respectively [P < .01]; if insured: 0.32 and 0.29, respectively [P > .05]). This result suggests that (1) controlling for child characteristics did not affect the findings, and the effect of SCHIP was significant and stronger for uninsured children; and (2) the SCHIP effect was greater for children with severe asthma who also had more problems at baseline.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Impact of SCHIP on Children With Asthma
Asthma Prevalence
The prevalence of asthma in New York's SCHIP (14–16%) was at least as high as the prevalence in the general population. If enrollment in New York's SCHIP is representative, >500000 children with asthma are enrolled nationally in SCHIP, and SCHIP can play an important role in addressing health care needs of children with asthma.
Asthma Morbidity
The number of asthma attacks that are experienced by children was markedly reduced during SCHIP compared with before SCHIP coverage, as were urgent health care visits and asthma-related hospitalizations. The parents reported improvements in their child's asthma severity during SCHIP. Together, these findings strongly suggest that receipt of SCHIP coverage was associated with a reduction in asthma morbidity.
SCHIP and the Potential Role of the Medical Home
Although more than one third of SCHIP enrollees with asthma were insured before SCHIP and 95% already had a USC, their unmet needs were high before SCHIP, and asthma visits were scattered across health care sites. During SCHIP, the pattern of care changed. More than 90% of children made most or all visits to their USC, and unmet needs were far lower, which suggests that provision of health insurance may facilitate use of the USC as a "medical home"80 for children with asthma, providing more accessible and coordinated care by the primary care provider that does not usually require specialty, ED, or hospital care.81,82 Children had equivalent service needs before and during SCHIP, but more of these needs were addressed successfully during SCHIP. Most parents attributed reasons for these improvements to better access to health care or medications.
Mechanisms by Which SCHIP Seemed to Improve Asthma Care
Although our study was not designed to determine how health insurance improves asthma care, our findings shed light on possible mechanisms. Because 45% of children changed their USC after enrollment in SCHIP, we anticipated that this may be a major factor. However, neither our subgroup analysis nor multivariate analysis found that children whose USC changed experienced greater improvements in asthma measures than those who did not change their USC. Second, although there was an overall trend toward more use of antiinflammatory medications (from 19% before SCHIP to 39% during SCHIP), greater use of antiinflammatory medications by themselves did not explain the improved outcomes. More intensive interventions83,84 beyond simply providing health insurance are likely needed to ensure optimal use of antiinflammatory medications. We therefore suspect that multiple factors resulted in improved asthma care within the USC. These other factors85 include receipt of acute reliever asthma medications at the USC, improved compliance with medications that are covered under SCHIP, and education about asthma that leads to earlier recognition of symptoms or changes in the environment (such as reduced exposure to tobacco). The parents stated that overall insurance coverage for medicines and improved overall asthma-related care were the key factors.
Strengths and Limitations
Our study was based on parent self-report without verification by providers, claims data, or medical chart review. Nonetheless, for some sentinel events such as hospitalizations or ED visits, self-report has been found to have good validity.86,87 Second, we could not control for possible regression to the mean (eg, children with particularly poor access at baseline or high unmet needs may have been more likely to enroll in SCHIP, and their care may have improved even in the absence of SCHIP). Our finding that pre-SCHIP access (eg, having a USC or unmet needs) were better than rates reported for uninsured children nationally88 makes regression to the mean less likely. Third, although we did not find any evidence for secular trends by using our comparison-group sample, secular trends still may have affected the results. Fourth, although the second interview captured 87% of those who were interviewed initially and nonresponders seemed similar to responders on baseline measures, some bias may have occurred from loss to follow-up. Fifth, we included all children who completed both interviews, although only 36% of the SCHIP enrollees with asthma were insured for 12 months before SCHIP and one sixth had become uninsured by the time of the second interview. However, both of these factors result in conservative estimates of the effect of SCHIP coverage on the population. Finally, it is possible that some of the improvements noted could have been a result of the child's SCHIP coverage being provided by managed care plans and not solely to the provision of insurance.
External validity is limited in that we studied only 1 state, although New York had 18% of the nation's SCHIP population. We studied white, black, and Hispanic children and not children of other races. Our study also focused on children who enrolled in SCHIP rather than children who were eligible but not enrolled.89,90 Thus, our study findings may not be generalizable to the entire population of near-poor children nor to children on Medicaid.
Implications for Clinicians and Policy Makers
Clinicians should be aware that many children with asthma are enrolling in SCHIP; their baseline unmet needs are high and quality of asthma care is poor. Enrollment in SCHIP may result in greater reliance on the USC and less scattering of care, which would enhance the opportunity for care within a medical home. More intensive office-based interventions are needed to improve key quality measures such as receipt of antiinflammatory asthma medicines.
Health plans that administer SCHIP should note that broad benefits are important for parents, who cited being able to afford asthma medicines as a major reason for improved asthma care.
State and federal policy makers should note that SCHIP coverage seems to improve access to asthma-related services as well as asthma-specific quality and outcome measures. Provision of health insurance seems to change the pattern of utilization of health services for children with asthma (with greater reliance on the USC) yet without greater use of high-cost specialty, emergency, or hospital services. SCHIP contributes to substantial benefits in quality of care, reduced parental worry, and perhaps better functional health status.
In sum, this study suggests that provision of health insurance to low-income children with asthma improves their health care.
|
ACKNOWLEDGMENTS |
|---|
We acknowledge Sally Findley, PhD, Mattie Irigoyen, MD, Cindy Brach, MPP, and Eugene Lewit, PhD, for their assistance. This study was part of the Child Health Insurance Research Initiative, the objective of which is to supply policy makers with information to improve access to and quality of health care for low-income children.68,91–93
|
FOOTNOTES |
|---|
Address correspondence to Peter G. Szilagyi, MD, MPH, Department of Pediatrics, Box 632, Strong Memorial Hospital, 601 Elmwood Ave, Rochester, NY 14642. E-mail: peter_szilagyi{at}urmc.rochester.edu
The authors have indicated they have no financial relationships relevant to this article to disclose.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  G. M. Kenney SCHIP Data: An Author Responds Health Aff., July 1, 2008; 27(4): 1199 - 1199. [Full Text] [PDF]
|
|
|
|
 |
|
 C. HOFFMAN and J. PARADISE Health Insurance and Access to Health Care in the United States Ann. N.Y. Acad. Sci., June 1, 2008; 1136(1): 149 - 160. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. J. Ungar, A. Kozyrskyj, M. Paterson, and F. Ahmad Effect of Cost-Sharing on Use of Asthma Medication in Children Arch Pediatr Adolesc Med, February 1, 2008; 162(2): 104 - 110. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Clancy, J. P. Kiley, and K. B. Weiss Eliminating Asthma Disparities Through Multistakeholder Partnerships Chest, November 1, 2007; 132(5): 1422 - 1424. [Full Text] [PDF]
|
|
|
|
 |
|
 Committee on Child Health Financing State Children's Health Insurance Program Achievements, Challenges, and Policy Recommendations Pediatrics, June 1, 2007; 119(6): 1224 - 1228. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. D. Klein, L. P. Shone, P. G. Szilagyi, A. Bajorska, K. Wilson, and A. W. Dick Impact of the State Children's Health Insurance Program on Adolescents in New York Pediatrics, April 1, 2007; 119(4): e885 - e892. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Kenney and J. Yee SCHIP At A Crossroads: Experiences To Date And Challenges Ahead Health Aff., March 1, 2007; 26(2): 356 - 369. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Butz, M. Tsoukleris, M. Donithan, V. D. Hsu, K. Mudd, I. H. Zuckerman, and M. E. Bollinger Patterns of Inhaled Antiinflammatory Medication Use in Young Underserved Children With Asthma Pediatrics, December 1, 2006; 118(6): 2504 - 2513. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||
