OBJECTIVES: In this study we examined geographic disparities in medical home access among US children with special health care needs (CSHCN) aged 0 to 17 years.
METHODS: The 2005–2006 National Survey of Children With Special Health Care Needs was used to estimate prevalence and odds of not having a medical home and 5 component outcomes according to state. Logistic regression was used to examine individual-level and state-level determinants of access.
RESULTS: Medical home access varied substantially across geographic areas. CSHCN in Alaska, Arizona, Washington, DC, Florida, Illinois, Massachusetts, New Jersey, Nevada, and Virginia had at least 50% higher adjusted odds of not having a medical home than CSHCN in Iowa. The adjusted prevalence of CSHCN lacking a medical home varied from a low of 46% in Iowa and Ohio to a high of 59% in Alaska and 61% in New Jersey. CSHCN in several western and southwestern states experienced greater problems with access to a personal doctor/nurse, a usual source of care, specialty care referrals, care coordination, and family-centered care. Adjustment for age, gender, race/ethnicity, household socioeconomic status, language use, insurance coverage, and functional limitation reduced state disparities in access. CSHCN in states with higher immigrant and non–English-speaking populations had significantly lower medical home access. Increases in state health care expenditure and infrastructure and Medicaid/State Children's Health Insurance Program eligibility were associated with increased access to a personal doctor/nurse.
CONCLUSIONS: Although individual-level sociodemographic and state-level health policy variables are important predictors of access, substantial geographic disparities remain, with CSHCN in several western and northeastern states at high risk of not having a medical home.
Health and health care inequalities in the United States remain substantial,1–3 and in some instances such inequalities have widened over time.4–6 Reducing social and geographic inequalities in health and health care continues to be a high priority for the US Department of Health and Human Services.7 Providing increased access to a medical home is an important policy objective toward reducing health care disparities and improving health and well-being among all children.8,9 A medical home is defined as a source of ongoing, comprehensive, coordinated, family-centered care in the child's primary health care environment.8–11 Access to a medical home has been associated with increased use of preventive health services, treatment adherence, and increased care coordination among both the general population and children with special health care needs (CSHCN).8,9 Medical home access has also been associated with fewer hospital admissions and emergency department visits, shorter length of hospital stays, reduced familial burden, increased access to needed services, and reduced risks of delayed/forgone care, unmet health care needs, and school absence.8,9,12–14
Racial/ethnic and socioeconomic disparities in medical home access have been examined among CSHCN, who comprise 14% of all US children.9,10 However, geographic disparities in medical home access among CSHCN have not yet been examined. The purpose of this article is twofold: (1) to estimate the observed and adjusted prevalence and odds of not having access to a medical home among CSHCN across the 50 states and the District of Columbia and (2) to identify individual-level sociodemographic and state-level social and health policy determinants of medical home access among CSHCN.
To analyze disparities in medical home access, we used the 2005–2006 National Survey of Children With Special Health Care Needs (NS-CSHCN), a nationally representative telephone survey of 40723 CSHCN <18 years old.15–17 Substantive and methodologic details of the survey are described elsewhere.15,16
Our analysis was based on 38886 CSHCN for whom the composite medical home variable could be defined. Medical home was operationalized by using questions related to its 5 components: (1) having a usual place for sick/well care; (2) having a personal doctor or nurse; (3) experiencing no difficulty in obtaining needed referrals; (4) receipt of needed care coordination; and (5) the presence of family-centered care.9,10
On the basis of previous research, we used child's age, gender, race/ethnicity, primary language spoken at home, household income/poverty levels, insurance coverage at the time of the survey, child's functional limitation, and state of residence as individual-level covariates.8–10 These covariates were measured as shown in Table 1.
Income/poverty status was missing for 9% of the households and was imputed by using a multiple-imputation technique.16,18 For all other covariates, there were very few missing responses, which were excluded from the multivariate analyses. The χ2 statistic was used to test the overall association between each covariate and medical home access. Prevalence (%) estimates of medical home access were computed for all 50 states and the District of Columbia. Multivariate logistic regression was used to examine the association between selected individual-level sociodemographic factors and the binary outcome variables of overall medical home access and its 5 components. Adjusted prevalence estimates were predicted marginals derived from the fitted logistic models. To account for the complex sample design of the survey, SUDAAN19 software was used to conduct multivariate logistic analyses and to determine crude and adjusted prevalence estimates.
A series of fixed-effects multilevel logistic models were also fitted by using SUDAAN software to estimate the effects of state-level social and health policy factors on the individual likelihood of not having access to a medical home or its 5 component outcomes after adjusting for individual-level covariates such as age, gender, race/ethnicity, language use, household poverty status, insurance coverage, and functional limitation.20–22 Considered as state-level factors were poverty rate, percentage immigrant or non–English-speaking population, Medicaid/State Children's Health Insurance Program (SCHIP) eligibility criteria, Medicaid expenditure per child, and several health care expenditure and infrastructure variables, including overall health expenditure and primary care physician supply rates.4,5,8,23–26 These variables have been indicated as factors that influence geographic disparities in health and health care.4,5,8,25,26 Because of high correlations among the area-level health care expenditure and supply variables, we constructed an index of health care expenditure and infrastructure by combining and factor-analyzing 4 variables: total number of physicians per capita; number of nurses per capita; number of pediatricians per child; and total health expenditure per capita, with factor loadings of 0.94, 0.89, 0.93, and 0.90, respectively. The health care infrastructure index (Cronbach's α, the reliability coefficient = 0.93) had a mean of 100 and an SD of 20, and the index scores ranged from a high of 207.66 for the District of Columbia to a low of 74.52 for Idaho.
State Variation in Medical Home Access
Overall, 53% of CSHCN in the United States did not have access to a medical home. The observed prevalence varied greatly across the states, with 43% of CSHCN in Iowa and 63% of CSHCN in Washington, DC, not having a medical home (Table 1). The map of overall access shown as Fig 1 indicates relatively higher rates of medical home access in the midwestern states and lower access rates in the coastal states.
Multivariate adjustment for individual-level covariates reduced state disparities in access (Table 1). Compared with those in Iowa, CSHCN in a number of states had at least 50% higher adjusted odds of not having access to a medical home, including Alaska, Arizona, Washington, DC, Florida, Illinois, Massachusetts, New Jersey, Nevada, and Virginia. The adjusted prevalence of CSHCN without a medical home varied from a low of 46% in Iowa and Ohio to a high of 59% in Alaska and 61% in New Jersey.
Although state patterns differed by individual medical home components, CSHCN in a number of western and southwestern states had relatively less access to a personal doctor or nurse, a usual source of care, referrals for specialty care, receipt of needed care coordination, and family-centered care (Fig 1). For example, 10% of the CSHCN in Alaska did not have access to a personal doctor or nurse compared with only 2% of the CSHCN in Rhode Island. Similarly, at least 28% of the CSHCN in California and Arizona had difficulty getting needed referrals for specialty care compared with 10% of the CSHCN in Rhode Island. The CSHCN in Alaska, Montana, North Dakota, Arizona, Oregon, Michigan, Arkansas, Oklahoma, Florida, and Georgia had at least threefold higher adjusted odds of not having access to a personal doctor or nurse than the CSHCN in Rhode Island. CSHCN in California, Arizona, Washington, DC, Pennsylvania, Delaware, Nevada, and Florida had at least 3 times higher adjusted odds of experiencing problems with needed referrals for specialty care than their Rhode Island counterparts (data not shown).
Although 40% of CSHCN nationally did not receive effective care coordination when needed, this percentage varied substantially according to state from a low of 32% in Indiana and 33% in Iowa to a high of 49% in Nevada and New Jersey. The CSHCN in Nevada, New Jersey, Alaska, Arizona, Washington, DC, Delaware, and Massachusetts had at least 50% higher adjusted odds of not receiving effective care coordination than CSHCN in Iowa. The percentage of CSHCN not receiving family-centered care varied from 25% for Iowa to 44% for Washington, DC. CSHCN in California, Maryland, New Jersey, and Nevada had at least 55% higher adjusted odds of not receiving family-centered care than the CSHCN in Iowa.
In addition to geographic disparities, the substantial effects of other covariates listed in Table 1 are worth mentioning. CSHCN aged 12 to 17 years had 26% higher adjusted odds of not having a medical home than those aged 0 to 5 years. Hispanic and black CSHCN had 56% and 57% higher odds, respectively, of not having access than their non-Hispanic white counterparts. CSHCN in non–English-speaking households had 97% higher odds of not having a medical home than those from English-speaking households. CSHCN who lived below the poverty threshold had 67% higher odds of not having a medical home than their most affluent counterparts. CSHCN whose condition greatly affected their activities had 190% higher odds of lacking access to a medical home than those without activity limitation. CSHCN without health insurance had 98% higher odds of not having a medical home than those with health insurance.
State-Level Social and Health Policy Influences on Medical Home Access
Table 2 lists the results of a series of fixed-effects multilevel logistic models, which show the effects of various social and health policy factors on the individual likelihood of lacking medical home access after controlling for individual-level covariates. Overall and most components of medical home access were inversely related to the size of the state's immigrant and non–English-speaking population. For example, a 10-percentage point increase in the state immigrant population was associated with a 12% increase in the odds of CSHCN not having a medical home, a 28% increase in the odds of experiencing problems with needed referrals for specialty care, a 14% increase in the odds of lacking care coordination, and a 10% increase in the odds of not receiving family-centered care.
The overall health care expenditure and infrastructure index and the variables that make up the index had a substantial influence on access to a personal doctor/nurse by CSHCN, even after controlling for the individual-level factors. For example, a 20-point increase in the health care index score was associated with an 18% decrease in the odds of CSHCN not having access to a personal doctor/nurse. CSHCN in states with more physicians (all specialties), pediatricians, and nurses also had increased access to a personal physician/nurse. There was a 39% decrease in the odds of not having access to a personal doctor/nurse for each additional pediatrician per 1000 children, a 19% reduction in odds of no access for each additional physician per 1000 population, and a 6% reduction in odds for each additional nurse per 1000 population. The odds of not having a personal doctor/nurse were reduced by 18% for every $1000 increase in the overall health spending per capita and by 11% for every $1000 increase in Medicaid expenditure per child. It is interesting to note that a lower nursing supply was associated with significantly higher likelihood of experiencing problems with needed referrals and family-centered care.
As expected, access to a personal doctor/nurse by CSHCN was significantly higher in states with lower poverty rates. However, the likelihood of not having access to a medical home was higher among CSHCN in states with lower poverty rates after adjusting for the individual-level covariates, including household poverty status. Lack of access to needed referrals, care coordination, and family-centered care was also higher among CSHCN from states with lower poverty rates and higher Medicaid/SCHIP eligibility limits.
To our knowledge, this is the first study to document the extent of state disparities in medical home access among CSHCN in the United States and to determine if a set of individual-level sociodemographic variables and state-level health policy factors affect the likelihood of CSHCN having a medical home. We found substantial geographic disparities, with the overall medical home access rates being markedly lower in several western and northeastern states. At least 43% of CSHCN in each state did not have access to a medical home. More than one quarter of CSHCN in each state experienced problems with specialty care referrals, care coordination, and family-centered care. Moreover, 10% to 12% of CSHCN in Washington, DC, Louisiana, Arizona, Montana, and Nevada did not have a usual source of care. Approximately 9% to 10% of CSHCN in Alaska, Montana, Arizona, Washington, DC, Mississippi, Arkansas, and Michigan did not have a personal doctor or nurse.
What explains such marked geographic disparities in access among CSHCN? As might be expected, state differences in individual-level factors such as racial/ethnic composition, English-language use, socioeconomic status, and health insurance coverage accounted for a substantial portion of the state disparities in overall and most components of medical home access. State-level social and health policy factors such as immigrant population size, poverty rate, health care expenditure and infrastructure, and Medicaid/SCHIP eligibility criteria were also significant predictors of either overall or some specific components of medical home access among CSHCN even after adjusting for individual-level covariates. These state-level health policy factors, therefore, are also expected to account for some of the state variations in medical home access rates.
States with higher poverty or deprivation levels may be significantly more constrained from a resource standpoint and/or in their organizational capacity to effectively deliver comprehensive primary care services to a majority of their population or households.5 Furthermore, states with a substantial immigrant population may experience significant barriers to providing accessible health care because of the cultural, linguistic, and health policy challenges that confront the immigrant population.27,28 Recent welfare reforms, including the 1996 Temporary Assistance for Needy Families (TANF) program and the Deficit Reduction Act (DRA) of 2005, which restrict or even exclude legal immigrants' access to public services and benefits, might adversely affect health services access and use among the immigrant population.29,30 It is important to note that states with higher immigrant or non–English-speaking populations experience significantly greater problems with such qualitative aspects of the medical home model as having access to needed referrals, care coordination, and family-centered care.
Increases in state health expenditure per capita, total physician supply, number of pediatricians and nurses per capita, and expanded Medicaid/SCHIP eligibility were positively associated with some aspects of medical home access, particularly access to a personal doctor/nurse. Increasing the supply of health care professionals such as nurses has an additional benefit in that it is associated with improved access to needed referrals and family-centered care among CSHCN. We found that overall access to medical homes, as well as access to specialty care referrals, care coordination, and family-centered care, was lower among CSHCN in states with lower poverty rates and higher Medicaid/SCHIP eligibility limits. This finding may indicate the need for the SCHIP and Medicaid programs to target CSHCN even in more affluent states to ensure increased access to comprehensive primary care services. The health policy variables, analyzed in our contextual analysis, clearly reflect the mechanisms and practices through which access by CSHCN to medical homes can be improved and geographic disparities in their health care access can be reduced.
Lower medical home access in more affluent states was unexpected. However, no significant association between state poverty level and medical home access was found at the bivariate level. The unexpected adjusted effect of area poverty may reflect state-level compositional differences in individual-level factors and/or the confounding influences of rural and immigrant population size in states. These area variables were not considered simultaneously because of multicollinearity. In addition, the unexpected adjusted effect of area poverty rate might reflect the limitation of area deprivation measured at the state level, as discussed below.
Some limitations of our study are worth mentioning. First, our study was a cross-sectional analysis in which we attempted to examine the association between individual-level sociodemographic and state-level health policy variables and the individual likelihood of medical home access among CSHCN. Although our state-level health policy variables temporally precede the survey measurement of access, a longitudinal design would be more appropriate for accurately estimating these effects. Second, the effects of some of the area-level factors, such as poverty rate, health care expenditure, and health professional supply variables, may be underestimated because they were considered at the state level rather than at the county or local community level. This was because the geographic area in the public-use NS-CSHCN file could not be identified at a lower level than the state. Third, in our multilevel models, both individual- and state-level covariates were treated as fixed effects, which might have resulted in underestimating the SEs of state-level effects.31,32 Although intracluster correlations in the data are, to some extent, accounted for by our use of SUDAAN software, which models complex sample design effects, a more general random-effects mixed multilevel model may be appropriate for testing whether individual-level effects are random (ie, they vary significantly across states) or whether state-level effects also vary randomly around an overall mean.31
The medical home is increasingly being recognized as the model of quality comprehensive health care not only for CSHCN but for all children.11,33,34 One of the Healthy People 2010 objectives calls for access to comprehensive health care consistent with the medical home standard for all CSHCN.7 The wide disparities shown here in the overall access to a medical home and in its 5 component outcomes according to race/ethnicity, socioeconomic status, and state of residence, however, should represent major policy concerns, because they may adversely affect the nation's effort to reduce health, disease, and health care disparities among CSHCN.
- Accepted August 3, 2009.
- Address correspondence to Gopal K. Singh, PhD, US Department of Health and Human Services, Health Resources and Services Administration, Maternal and Child Health Bureau, 5600 Fishers Lane, Room 18-41, Rockville, MD 20857. E-mail:
The views expressed are the authors' and not necessarily those of the Health Resources and Services Administration or the US Department of Health and Human Services.
Financial Disclosure: The authors have indicated they have no financial relationships relevant to this article to disclose.
- ↵National Center for Health Statistics. Health, United States, 2007 With Chartbook on Trends in the Health of Americans. Hyattsville, MD: US Department of Health and Human Services; 2007
- Braveman P, Egerter S. Overcoming Obstacles to Health. Princeton, NJ: Robert Wood Johnson Foundation; 2008
- ↵Smedley BD, Stith AY, Nelson AR, eds. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington, DC: Institute of Medicine, National Academies Press; 2003
- ↵Singh GK, Siahpush M. Widening socioeconomic inequalities in US life expectancy, 1980–2000. Int J Epidemiol.2006;35 (4):969– 979
- ↵US Department of Health and Human Services. Healthy People 2010: Understanding and Improving Health. 2nd ed. Washington, DC: US Government Printing Office; 2000
- ↵Health Resources and Services Administration, Maternal and Child Health Bureau. The National Survey of Children With Special Health Care Needs Chartbook, 2005–2006. Rockville, MD: US Department of Health and Human Services; 2007
- ↵American Academy of Pediatrics, Medical Home Initiatives for Children With Special Needs Project Advisory Committee. The medical home. Pediatrics.2002;110 (1 pt 1):184– 186
- ↵Homer CJ, Klatka K, Romm D, et al. A review of the evidence for the medical home for children with special health care needs. Pediatrics.2008;122 (4). Available at: www.pediatrics.org/cgi/content/full/122/4/e922
- ↵National Center for Health Statistics. The National Survey of Children's With Special Health Care Needs (NS-CSHCN), 2005–2006: The Public Use Data File. Hyattsville, MD: US Department of Health and Human Services; 2007
- ↵Blumberg SJ, Welch EM, Chowdhury SR, Upchurch HL, Parker EK, Skalland BJ. Design and operation of the National Survey of Children With Special Health Care Needs, 2005–2006. Vital Health Stat 1.2008;(45):1– 188
- ↵Kogan MD, Strickland BB, Newacheck PW. Building systems of care: findings from the National Survey of Children With Special Health Care Needs. Pediatrics.2009;124 (suppl 4):S333– S336
- ↵Pedlow S, Luke JV, Blumberg SJ. Multiple imputation of missing household poverty level values from the National Survey of Children With Special Health Care Needs, 2001, and the National Survey of Children's Health, 2003. Available at: www.cdc.gov/nchs/data/slaits/mimp01_03.pdf. Accessed April 28, 2008
- ↵SUDAAN: Software for the Statistical Analysis of Correlated Data [computer program]. Release 9.0.1. Research Triangle Park, NC: Research Triangle Institute; 2005
- ↵US Census Bureau. Statistical Abstract of the United States, 2008. 127th ed. Washington, DC: US Government Printing Office; 2008
- Health Resources and Services Administration. Area Resource File, 2007 Release. Rockville, MD: US Department of Health and Human Services; 2008
- ↵Health Resources and Services Administration, Maternal and Child Health Bureau. Child Health USA 2006. Rockville, MD: US Department of Health and Human Services; 2006
- ↵Association of State and Territorial Health Officials. State Policy Options to Establish Medical Homes for Children and Youth. Washington, DC; ASTHD: 2005
- ↵Singh GK, Hiatt RA. Trends and disparities in socioeconomic and behavioral characteristics, life expectancy, and cause-specific mortality of native-born and foreign-born populations in the United States, 1979–2003. Int J Epidemiol.2006;35 (4):903– 919
- ↵Coven M. An introduction to TANF. Available at: www.cbpp.org/1-22-02tanf2.htm. Accessed October 30, 2008
- ↵Rosenbaum S, Markus A. The deficit reduction act of 2005: an overview of key Medicaid provisions and their implications for early childhood development services. Available at: www.commonwealthfund.org/publications/publications_show.htm?doc_id=409144. Accessed October 30, 2008
- ↵Diez Roux AV. A glossary for multilevel analysis. J Epidemiol Community Health.2002;56 (8):588– 594
- ↵American Academy of Family Physicians, American Academy of Pediatrics, American College of Physicians, American Osteopathic Association. Joint principles of the patient-centered medical home. Available at: www.medicalhomeinfo.org. Accessed September 23, 2008
- ↵Beal AC, Doty MM, Hernandez SE, Shea KK, Davis K. Closing the Divide: How Medical Homes Promote Equity in Health Care. Results from the Commonwealth Fund 2006 Health Care Quality Survey. New York, NY: Commonwealth Fund; 2007
- Copyright © 2009 by the American Academy of Pediatrics