PEDIATRICS Vol. 111 No. 5 May 2003, pp. 1024-1029
The Ecology of Medical Care for Children in the United States
* Robert Graham Center, American Academy of Family Physicians, Washington, DC
American Academy of Pediatrics, Center for Child Health Research and Department of Pediatrics, University of Rochester School of Medicine and Dentistry, Rochester, New York
Department of Research, Olmsted Medical Center, Rochester, Minnesota
¶ Agency for Healthcare Research and Quality, Rockville, Maryland
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Background. Medical care ecology has previously been investigated for adults, but no similar exploration has been made specifically for children.
Objective. To describe proportions of children receiving care in 6 types of health care setting on a monthly basis and to identify characteristics associated with receipt of care in these settings.
Methods. Nationally representative data from the 1996 Medical Expenditure Panel Survey were used to estimate the number of children per 1000 per month who received care at least once in each type of setting. Multivariate analyses assessed associations between receiving care in various settings and children’s sociodemographic factors (age, sex, ethnicity, poverty, parent’s education, urban or rural residence, insurance status, and whether or not the child had a usual source of care).
Results. Of 1000 children aged 0 to 17 years, on average each month 167 visited a physician’s office, 82 a dentist’s office, 13 an emergency department, and 8 a hospital-based outpatient clinic. Three were hospitalized and 2 received professional health care in their home. Younger age was associated with increased proportions of children receiving care in all health care settings except dentists’ clinics. Poverty, lack of health insurance, black race, and Hispanic ethnicity were associated with decreased receipt of care in physicians’ and dentists’ offices. Only age (<2 years and 13–17 years) and poverty status were associated with hospitalization (P < .05 for each). Rural residence was not associated with any significant variation in proportions of children receiving care in any setting. Having a usual source of care was associated with increased receipt of care in all settings except hospitals.
Conclusions. The ecology of children’s medical care is similar to that of adults in the United States in that a substantial proportion of children receive health care each month, mostly in community-based, outpatient settings. In all settings except emergency departments, receipt of care varies significantly by children’s age, race, ethnicity, income, health insurance status, and whether they have a usual source of care.
Key Words: ecology • medical care • health systems • Medical Expenditure Panel Survey • primary care
Abbreviations: MEPS, Medical Expenditure Panel Survey • OR, odds ratio
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
In 1961, White et al1 crystallized a patient-centered and population-based vision of health care in a seminal paper describing the ecology of medical care. The motivation for this research was to demonstrate a more valid perspective of medical care use than one would have from data drawn from the biased samples offered by hospitals or outpatient settings in isolation. The ecology model encourages an appreciation of the overall health care use of the community. Policy makers and educators have repeatedly referred to this model2–5 and it has been tested and updated from time to time,6–8 but only recently with data representative of the whole US population.8
White et al1 deliberately excluded children from their study. They did this because "most decisions about children’s medical care are customarily made by their parents or guardians." Children were included in the most recent medical ecology study,8 but no separate investigation for children has yet been undertaken. A medical ecology view of children’s health care environment may be different from that of adults and similarly useful to policy makers in assessing the need for adjustments in various programs to facilitate insurance coverage for children9,10 (such as the State Children’s Health Insurance Program11 and Medicaid12) and in planning to reduce disparities in children’s access and use of health services.13 An ecologic overview provides the broad context of children’s medical care experiences. Taking this perspective should prompt consideration of medical, nursing, and related educational curricular needs of health professionals to ensure that they are suitably prepared to meet the health care needs of the nation’s children.
We describe the medical ecology for children in the United States using the most comprehensive and representative data currently available, and identify sociodemographic characteristics associated with different proportions of children receiving care in various settings. Increased susceptibility to a variety of acute infectious diseases, lower rates of chronic health conditions, and recommended schedules for routine or well child visits for health promotion, disease prevention, and monitoring of growth and development14 led us to expect that the ecology of medical care for children would differ from the ecology of medical care for adults.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Study Population
The general methods of the updated analysis of the ecology of medical care are reported in detail elsewhere.8 In summary, we used data extracted from the publicly released 1996 Medical Expenditure Panel Survey (MEPS) sponsored by the Agency for Health care Research and Quality. The MEPS survey was designed to provide detailed information on the health care use and expenditures of a representative sample of the civilian, noninstitutionalized US population. It is an ongoing nationally representative survey with a rotating panel design that enables longitudinal study of a large cohort over a 2.5-year period. Its sample consists of households and individuals that had taken part in a previous National Health Interview Survey. MEPS data in 1996 represented 99 892 797 US families and 26.4% of these families included 73 532 889 children aged <18 years. In this study, household component records characterizing individuals and families were linked to data from event files describing their health care encounters in calendar year 1996. MEPS records are weighted to permit the calculation of national estimates, usually with small standard errors. The current analysis is limited to children aged <18 years at December 31, 1996.
Variables
To describe the ecology of health care for children we used variables computed from MEPS health event files and estimated the number of children per 1000 aged <18 years who in an average month received health care in each of 6 types of setting identified in MEPS: 1) a physician’s office; 2) a dentist’s office; 3) an outpatient clinic of a hospital; 4) a hospital emergency department; 5) as an inpatient in a hospital (excluding hospitalization for birth); and 6) in their own home by a professional health care provider.
Predictor variables in multivariate analyses were taken from MEPS household files. We evaluated the following for inclusion as independent variables in multivariate analyses (see Table 1):
|
- Age—4 age groups (<2 years, 2–5 years, 6–12 years, and 13–17 years).
- Sex.
- Race—4 racial groups (white, black, Asian and Pacific Islands, and American Indian) with Aleutians/Eskimos omitted because of small sample sizes and thus large standard errors of estimates.
- Ethnicity—Hispanic or non-Hispanic.
- Urban or rural residence–most suburban residents live in Metropolitan Statistical Areas and were defined as urban. Residents of non-Metropolitan Statistical Area counties were defined as rural.
- Education–3 categories were defined by the highest degree attained by the head of the household in which a child resided.
- Economic status–family income was divided by the Federal poverty line based on family size and composition, with the resulting percentages grouped into 5 categories.
- Health insurance–any health insurance held (including Medicaid and private health insurance) without regard to adequacy of coverage.
- Usual source of care–response to the question: "Is there a particular doctor’s office, clinic, health center, or other place that you go to if you are sick or need advice about your health?"
Records were excluded from multivariate analyses if data were missing on any of the study variables. Education, economic status, and residence location were characteristics of the child’s household. All other variables were characteristics of individual children.
Analytical Strategy
Descriptive analyses were first performed to estimate the number of children per 1000 of the US civilian noninstitutionalized child population in 1996 who, in a month, received health care in each of the 6 settings listed above. We first created person-month records as the unit of analysis. It reflects participation (use or nonuse by an individual person) in a health care setting rather than use (number of health care contacts). Each MEPS child participant had 1 record of participation for each setting for each month during the 1996 survey year. The value in this record was "0" if the child had no contact with the setting in the month and "1" if the child had 1 or more contacts. The individual’s survey weight was applied to each record and the result multiplied by 1000 to produce national estimates.
Bivariate analyses indicated strong associations between each predictor variable and participation by children in 1 or more health care settings. Too few children used home care to produce reliable estimates. Therefore, 5 separate logistic regression analyses were performed to derive adjusted odds ratios (ORs) of the independent association of each predictor variable and children’s participation in health care in each setting in an average month in 1996. In effect, associations between participation and setting were assessed statistically controlling for all other predictor variables. The numerically largest category was designated the reference group in these analyses.
SUDAAN statistical software15 (Research Triangle Institute, Research Triangle Park, NC) was used for all analyses as it adjusts variance estimates because of MEPS survey design complexity, particularly the substantial over-sampling of children and families from racial and ethnic minority groups.16
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Of 1000 children aged 0 to 17 years, on average each month 167 visited a physician in the office setting, 82 visited a dentist, 13 received care in an emergency department, 8 visited a hospital-based outpatient clinic, 3 spent time as an inpatient in hospital, and 2 received professional health services in their home. Fig 1 shows the ecology of medical care for US children aged <18 years contrasted with US noninstitutionalized civilians 18 years of age and over.
|
18 years (B).Table 1 shows the proportions of children receiving care in different health care settings stratified by sociodemographic characteristics. In an average month, health care was received in physicians’ offices by a greater proportion of infants than older children, of white children than black, of non-Hispanic children than Hispanic, of insured than uninsured, of children from wealthier families than poorer, of children living in families with college-educated heads of household, and a greater proportion of children who had a usual care provider. Similar patterns were apparent for hospital outpatient departments and for dentists’ offices, with the exception of the age differential being reversed for dentists. No such patterns of differential health care participation were distinguished in emergency department care, hospital inpatient care, and home care settings.
Table 2 provides the results of the multivariate analyses. Receipt of care in physicians’ and dentists’ offices was sensitive to all sociodemographic characteristics in the model (age, sex, race, ethnicity, insurance status, poverty category, education of head of household, residence location, usual source of car) except living in an urban or rural location. After accounting for all other characteristics, making office visits was significantly more likely for younger children than for children 6 years of age and older (OR = 3.58 [3.14, 4.08] for children aged <2 years; OR = 1.48 [1.36, 1.61] for children aged 2–4 years; P < .001 in both cases) and for children in families where the head of the household had a degree beyond a high school diploma (OR = 1.18, 1.06, 1.32; P = .003). Conversely, children were significantly less likely to make office visits if they were black (OR = 0.57 [0.51, 0.64]; P < .001), Hispanic (OR = 0.81 [0.71, 0.92]; P = .001), uninsured (OR = 0.61 [0.52, 0.70]; P < .001), living in low-income families (see Table 2 for ORs), living in families where the head of the household had not graduated from high school (OR = 0.73 [0.63, 0.83]; P < .001), and if they did not have a usual health care provider (OR = 0.40 [0.33, 0.48]; P < .001). After accounting for all other sociodemographic characteristics, sex, race, ethnicity, and education of head of household did not significantly influence the number of children per 1000 per month having hospital stays and outpatient visits. Insurance status, poverty category, and education of head of household did not influence odds of visiting emergency departments. Rural residence was not independently associated with variation in proportions of children receiving care in any of the health care settings investigated, whereas having a usual source of care was associated with an increased likelihood of children receiving care in all settings other than hospitals.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This study is the first application to children of the classic ecology of medical care model.1,8 It demonstrates that, as is true of adults, a substantial proportion of children make contact with the health care system in a typical month, and most children who are seen in a month are seen in outpatient, community-based settings. Although overall patterns were in general similar for children and adults, for most settings we explored the proportion of children receiving care was smaller than the proportion of adults, probably because children as a group carry less burden of chronic and degenerative disease than adults. The exceptions were emergency departments where proportions were approximately equal, and dentists’ offices, where children were more likely than adults to receive care.
Age strongly affected the ecology of medical care for children. Among all children, those <2 years of age were most likely to receive care at least once in a typical month in all settings except dentists’ offices. Children 2 to 5 years old were more likely than children 6 to 12 years old to receive care in all settings except dentists’ offices. In contrast, greater proportions of children aged 13 to 17 years than 6 to 12 years received care in dentists’ offices, emergency departments, outpatient clinics, and hospitals, but not physicians’ offices, despite recommendations for adolescents to have more frequent preventive health visits than 6 to 12 year olds.14,17
The multivariate analyses in this investigation added a further dimension to the ecology model and confirmed that children’s participation in the health care system is affected more substantially by their age than by any other sociodemographic characteristic we investigated. Of particular interest is the physicians’ office setting where, controlling for other variables, fewer children received care if they were poor, uninsured, not white, Hispanic, living with a head of household with lower educational attainment, or lacking a usual source of care. Given the relative importance of physicians’ offices in providing children with preventive care, wellness checks, and a wide range of acute illness services, these differences merit careful scrutiny.
These data do not demonstrate that children in rural settings were any less likely to receive care in any of the health settings assessed. This is surprising because approximately two thirds of the 2900 primary care physician shortage areas in the United States are in rural areas18 and 
7 million children live in these shortage areas, for whom an additional 2000 to 3500 clinicians have been estimated to be needed to appropriately provide medical care.19 However, this analysis makes no investigation of the number of contacts made by children and it may be that children in rural areas receive fewer services per child, despite equivalent numbers of children in rural and urban areas receiving any services each month.
We found that having a usual source of care is independently associated with greater likelihood of children receiving care in a typical month in physicians’ offices, dentists’ offices, emergency departments and outpatient clinics, but not hospitals. It is possible that the measure of "usual source of care" that we used in this study was affected by recall bias, with children having more recent care being more likely to have been attributed a "usual source" of this care. However, it is plausible that having a single usual source of care is an important factor in facilitating children’s access to health services. Further research is needed to establish this association with greater certainty.
Dental care has not been incorporated into prior analyses of the ecology of medical care. Our findings corroborate current concerns about disparities in access to children’s oral health services in the United States,20–22 demonstrating the same patterns of use and variation seen in physicians’ offices. The American Academy of Pediatrics14 and the American Academy of Pediatric Dentistry23 disagree on the appropriate age for children to routinely begin seeing dentists, with the former suggesting 3 years of age while the latter advocates that dental visits begin at 1 year. Our study found that only 2 children per 1000 aged <2 years received dental services in any month, increasing to only 38 per 1000 among children 2 to 5 years of age. Thus, neither recommendation about dental care for young children had been adopted in practice when these data were collected in 1996.
Two other findings about children’s dental services merit attention. First, almost the same proportions of adolescents received care in dentists’ and office-based physicians’ offices each month. Second, more children older than 2 years of age visited dentists’ offices than any other setting apart from physicians’ offices. We estimated that 82 children per 1000 per month receive care in dentists’ offices. This finding suggests that dental clinics could serve as an important source of health promotion, disease prevention, and screening for nonoral health-related problems if new roles and better integration of children’s dental services and other child health care services could be arranged.
We were unable to reproduce certain strata used in other ecology model research.1,8 Estimates of the numbers of people experiencing symptoms of ill health, considering seeking medical care, and being hospitalized in academic health centers were included in earlier work but we could produce no comparable estimates for children. The population-based literature on children experiencing symptoms of ill health is sparse and possibly outdated.24 The MEPS contains no information about symptoms children may have experienced or about families who considered seeking care for their children but did not. No nationally representative studies providing these estimates for children could be found in the literature. This important work appears yet to be undertaken.
This analysis has important limitations. Because of small numbers, we could not perform multivariate analyses concerning home care. We could not ascertain the proportion of children obtaining primary or subspecialty care from physicians’ offices or hospital clinics, and we could not distinguish community health centers from private physician offices or visits for routine well child care from visits for acute or chronic conditions. It is unfortunate that the data used here from 1996 are the most recent MEPS data permitting usual source of care to be included in the analysis.
This study also has important strengths. We based our analysis on nationally representative samples that were large enough to produce reasonably precise estimates of the proportions of children receiving care. Racial and ethnic minority groups were over-sampled to ensure similar precision of estimates concerning these groups. Data were collected from the surveyed sample using such methods as logs of encounters to minimize recall bias. This analysis extended previous ecology analyses by including dentistry in the array of health care settings investigated.
These results should not be interpreted as setting worthy health care participation goals as the data we used included no measures of unmet need or appropriateness of received care. Instead, this analysis characterized children’s receipt of health care in selected settings during the months of 1996. This first application of the Ecology of Medical Care model to children demonstrates that while fewer children make contact with the health care system each month, the overall pattern of their contacts is similar to that of adults. This is a macro-level model that allows the health care participation of all children to be assessed across an array of settings. The model shows that the health care settings used most by children in the United States are still accessed to a greater extent by white children, non-Hispanic children, by children living in wealthier households, by children in families headed by people with higher levels of education, and by insured children. If equality of health care access is our goal, there is still a distance to be traveled, especially to achieve equality of access in the health care settings used by most children—the offices of physicians and dentists.
|
FOOTNOTES |
|---|
Received for publication May 23, 2002; Accepted Sep 16, 2002.
Address correspondence to Susan Dovey, PhD, MPH, The Robert Graham Center, Policy Studies in Family Practice and Primary Care, 2023 Massachusetts Ave, NW, Washington, DC 20036. E-mail: sdovey{at}aafp.org
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
- White KL, Williams TF, Greenberg BG. The ecology of medical care. N Engl J Med.1961; 265 :885 –892
- McWhinney I. An Introduction to Family Medicine. New York, NY: Oxford University Press; 1981
- White K. The ecology of medical care: origins and implications for population-based healthcare research. Health Serv Res.1997; 32 :11 –21[Web of Science][Medline]
- Task Force on Building Capacity for Research in Primary Care. Putting Research into Practice. Rockville, MD: Agency for Health Care Policy and Research; 1993
- Godwin M, Grzybowski S, Stewart M, et al. Need for an institute of primary care research within the Canadian institutes of health research. Can Fam Physician.1999; 45 :1405 –1409[Web of Science][Medline]
- White K. Life and death and medicine. Sci Am.1973; 229 :23 –33
- Thacker S, Greene S, Salber E. Hospitalizations in a southern rural community: an application of the "ecology model".
Int J Epidemiol.1977; 6
:55
–63
[Abstract/Free Full Text] - Green L, Fryer G, Yawn B, Lanier D, Dovey S. The ecology of medical care revisited.
N Engl J Med.2001; 344
:2021
–2025
[Free Full Text] - Holl JL, Szilagyi PG, Rodewald LE, Byrd RS, Weitzman ML. Profile of uninsured children in the United States.
Arch Pediatr Adolesc Med.1995; 149
:398
–406
[Abstract/Free Full Text] - Simpson G, Bloom B, Cohen RA, Parsons PE. Access to Health Care. Part 1: Children. Hyattsville, MD: National Center for Health Statistics, 1997
- Shi L, Oliver T, Huang V. The Children’s Health Insurance Program: expanding the framework to evaluate state goals and performance. Milbank Q.2000; 78 :340 –341
- Feld P, Matlock C, Sandma DR. Insuring the Children of New York City’s Low-Income Families. New York, NY: The Commonwealth Fund; 1998
- Weitzman M, Byrd RS, Auinger P. Black and white middle class children who have private health insurance in the United States.
Pediatrics.1999; 104
:151
–157
[Abstract/Free Full Text] - American Academy of Pediatrics. Policy statement: recommendations for Preventive Pediatric Health Care (RE9939).
Pediatrics.2000; 105
:645
[Free Full Text] - Shah BV, Barnwell BG, Bieler GS. SUDAAN: Software for the Statistical Analysis of Correlated Data, User’s Manual, Volume 1. Research Triangle Park, NC: Research Triangle Institute; 1997
- Cohen J. Design and Methods of the Medical Expenditure Panel Survey Household Component. MEPS Methodology Report; 1997:1
- Elster AB, Kuznets MJ. Guidelines for Adolescent Preventive Services. Baltimore, MD: Williams & Wilkins; 1993
- North Carolina Rural Health Research Program, University of North Carolina at Chapel Hill Cecil G. Sheps Center for Health Services Research. Facts About Rural Physicians. Washington, DC: Federal Office of Rural Health Policy, HRSA, US Department of Health and Human Services; 1997
- The Future of Pediatric Education II. Organizing pediatric education to meet the needs of infants, children, adolescents and young adults in the 21st century. Pediatrics.2000; 105 :S182
- Mouradian W, Wehr E, Crall J. Disparities in children’s oral health and access to dental care.
JAMA.2000; 284
:2625
–2631
[Abstract/Free Full Text] - US Department of Health and Human Services. Oral Health in America: A Report of the Surgeon General. Rockville, MD: US Department of Health and Human Services, National Institute of Dental and Craniofacial Research. National Institutes of Health; 2000
- Weitzman M, Moss M. About children’s oral health needs. Ambul Pediatr.2001; 1 :162 –165[CrossRef][Web of Science][Medline]
- Berkowitz RJ, Zero D, Edelstein B. Dental decay. In: Campos-Outcalt D, ed. 20 Common Problems in Preventive Health Care. New York, NY: McGraw-Hill; 2000: 65–78
- Roghmann K, Haggerty R. The diary as a research instrument in the study of health and illness behavior: experiences with a random sample of young families. Med Care.1972; 10 :143 –163[CrossRef][Medline]
PEDIATRICS (ISSN 1098-4275). ©2003 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:
|
|
 |
|
  B. Barnet, J. Liu, and M. DeVoe Double Jeopardy: Depressive Symptoms and Rapid Subsequent Pregnancy in Adolescent Mothers Arch Pediatr Adolesc Med, March 1, 2008; 162(3): 246 - 252. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. T. Chien and R. A. Dudley Pay-for-Performance in Pediatrics: Proceed With Caution Pediatrics, July 1, 2007; 120(1): 186 - 188. [Full Text] [PDF]
|
|
|
|
|
|
E. Schainker, M. J. O'Brien, D. Fox, and H. Bauchner School Nursing Services: Use in an Urban Public School System Arch Pediatr Adolesc Med, January 1, 2005; 159(1): 83 - 87. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. J. Chung and M. A. Schuster Access And Quality In Child Health Services: Voltage Drops Health Aff., September 1, 2004; 23(5): 77 - 87. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
The Child Health Business Case Working Group Exploring The Business Case For Improving The Quality Of Health Care For Children Health Aff., July 1, 2004; 23(4): 159 - 166. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. R. Kemper, L. M. Cohn, and K. J. Dombkowski Patterns of Vision Care Among Medicaid-Enrolled Children Pediatrics, March 1, 2004; 113(3): e190 - 196. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
