Objective. To examine health care utilization and expenditure patterns for children with disabilities.
Methods. Secondary data analysis was conducted of the 1999 and 2000 editions of the Medical Expenditure Panel Survey (MEPS), a nationally representative survey conducted in 5 rounds by household interview. Two years of MEPS data were combined in this analysis to improve the precision of estimates. Disability was defined by the presence of a limitation in age-appropriate social role activities, such as school or play, or receipt of specialized services through the early intervention or special education programs. The survey sample included 13 792 children younger than 18 years. The overall response rate was 65.5%.
Results. Our findings demonstrate that the 7.3% of US children with disabilities used many more services than their counterparts without disabilities in 1999–2000. The largest differences in utilization were for hospital days (464 vs 55 days per 1000), nonphysician professional visits (3.0 vs 0.6), and home health provider days (3.8 vs 0.04). As a result of their greater use, children with disabilities also had much higher health care expenditures ($2669 vs $676) and higher out-of-pocket expenditures ($297 vs $189). We also found that the distributions of total and out-of-pocket expenses were highly skewed, with a small fraction of the disabled population accounting for a large proportion of expenditures: the upper decile accounted for 65% of total health care expenses and 85% of all out-of-pocket expenses for the population with disabilities. Health insurance was found to convey significant protection against financially burdensome expenses. However, even after controlling for insurance status, low-income families experienced greater financial burdens than higher income families.
Conclusions. The skewed distribution of out-of-pocket expenses found in this and earlier studies indicates that the financial burden of childhood disability continues to be shared unevenly by families. Low-income families are especially vulnerable to burdensome out-of-pocket expenses. Additional efforts are needed to protect these high-risk families.
Estimates of the number of US children with chronic conditions vary widely depending on how the population is defined.1–5 Using less restrictive definitions of chronic conditions, estimates range upwards of 30%.3 Most of these children are only mildly affected by their conditions, and many of their conditions are self-limiting.3–5 An estimated 13% of children have a chronic physical, behavioral, or emotional condition that results in a special health care need (2001 National Survey of Children with Special Health Care Needs, unpublished data). Approximately half of these children have severe enough conditions to result in limitations of their social role activities (play or school, depending on age), or disabilities in common parlance.6
Several studies have reported on the costs of health care for children with chronic conditions. Lozano et al7 provided estimates of the costs of caring for children with asthma. Chan et al8 examined health care costs for children with attention-deficit disorder, and Ringel and Sturm9 estimated the costs of childhood mental health disorders. However, no recent studies have examined health care expenditures for the broader population of children with disabilities.
It is widely known that children with disabilities use far more health care services than other children.4,10 Consequently, their health care expenditures are higher.10,11 Although more recent analyses of expenditures for low-income children who have disabilities and receive Supplemental Security Income have been published by Neff et al12 and Kuhlthau et al,13 the last comprehensive national study of health care expenditures for children with and without disabilities was conducted using 1980 data.11 That study found that annual expenditures for children with disabilities averaged 3 times those for children without disabilities. The findings from that study also showed that a relatively small group of children with disabilities accounted for a disproportionate share of health care spending. Among children with disabilities, those in the upper decile of the expenditure distribution accounted for 65% of all charges.
These findings concerning the added costs of caring for children with disabilities and the concentration of health care expenses within a small subset of the population have important implications for health policy. Understanding the nature of health care expenses for this population is important for health planning efforts by public agencies such as the federal and state Title V programs for children with special health care needs.1 Information on the distribution of health care expenditures is also important for devising appropriate benefit packages and efficient payment mechanisms under public and private health insurance. Finally, information on the predictors of high-cost children, especially those with high out-of-pocket expenses, is important in developing strategies to protect families against experiencing catastrophic health care expenses.
Data collected in earlier expenditure surveys predate the nationwide shift to managed care from fee-for-service14–16 and do not reflect changes in the nature and number of children with disabilities occurring during the past 2 decades.17 Moreover, current data on health care expenditures and out-of-pocket financial burden are particularly important given state-level budget shortfalls. New data from the Medical Expenditure Panel Survey (MEPS) are now available to fill the existing gaps in our knowledge base on health care expenditures for children with disabilities. This report uses data from the 1999 and 2000 panels of the MEPS to provide current information on the distribution and composition of health care expenditures for children with and without disabilities. Two years of data were combined to increase the precision of the estimates reported here.
MEPS is designed to produce national estimates of the health care use, expenditures, and insurance coverage of the US civilian, noninstitutionalized population. It is composed of 4 component surveys: the Household Component, the Medical Provider Component, the Insurance Component, and the Nursing Home Component. MEPS uses an overlapping panel design in which data are collected through a preliminary contact followed by a series of 6 rounds of interviews over a 2.5-year period.18 The sampling frame for the MEPS was drawn from respondents to the previous year's National Health Interview Survey (NHIS) sample. The survey collected data in each round on use and expenditures for office and hospital-based care, home health care, dental services, vision aids, and prescribed medicines. Data were collected for each sample person at the event level and summed across rounds to produce the annual utilization and expenditure data. The response rate was 65% in 1999 and 66% in 2000.18
All 13 792 children who were younger than 18 years in the 1999–2000 MEPS were included in our study. A knowledgeable adult, usually a parent, answered questions about health and health services use for children younger than 17. When present, 17-year-olds answered the questions for themselves. All interviews were conducted in-person through computer-assisted personal or telephone interview method. The survey is conducted in English and Spanish.
Children with disabilities are defined as children who were limited in or unable to perform age-appropriate social roles because of a chronic physical or mental health problem. For children younger than 5, parents were asked about whether the sample child was limited in or unable to engage in the kind or amount of play activities typical for the child's age and whether the child received any early intervention services. For children 5 to 17, parents were asked whether the sample child was limited or unable to go to school, limited in other nonschool activities, or received special education services. Children who were reported as limited in or unable to conduct their major activity or require early intervention/special education services were classified as having a disability in this study. It is possible that some children who receive early intervention services may reside in states where early intervention services are provided to children who are at risk of a disability and thus may not have had a disability at the time of the survey. However, no children in the sample were identified as having a disability solely on the basis of this criterion; all children who were reported to be receiving early intervention services were also reported to have a limitation of activity.
MEPS is the only national survey that provides detailed information for estimating the health care expenditures and insurance coverage for special populations of interest. Total health care expenditures, or charges, were calculated for all individuals in the sample, including expenditures for hospital inpatient and outpatient services, physician services, dental services, services provided by health care professionals other than physicians (eg, nurse practitioners, physical therapists, psychologists, social workers), prescribed medications, diagnostic tests, and certain types of medical equipment and supplies including eyeglasses and contact lenses.
For improving accuracy of expenditure data, MEPS also contains a Medical Provider Survey (MPS). The MPS is designed to collect payment information from providers seen by household respondents, including office-based physicians, hospitals, and paid home health care providers, as well as prescription drug expenses. The main purpose of the MPS is to provide missing data or replace unreliable data reported by or for household respondents. Expenditure data collected in the MPS are generally regarded as more accurate than information collected in the household survey and are used as the source of data for imputations for missing expenditures. Typically, when MPS data are available for a given event, they are used to construct the expenditures for that event. Otherwise, household-reported data are used. Only household data are available for expenditures for nonphysician visits, dental and vision services, and medical supplies and equipment; these services were not covered by the MPS.
Of special interest in this study is the financial burden of out-of-pocket expenses experienced by families. Out-of-pocket expenditures are defined as the payments made by families for health care and include out-of-pocket spending on deductibles and other forms of cost sharing, such as copayments and coinsurance, and direct expenditures for health care services, equipment, and supplies not covered by insurance. We measured financial burden as the ratio of the child's annual out-of-pocket health care expenses to annual family income multiplied by 100 (ie, the percentage of family income spent out of pocket on health care). Also, we used out-of-pocket expenditures for the child exceeding 5% of family income as a threshold for indicating catastrophic expenses.
The MEPS oversampled certain populations, including children in low-income families. Also, because the MEPS panel was originally drawn from the NHIS, the MEPS also reflected an oversample of Hispanics and blacks in the NHIS. Estimates presented in the tables and text have been statistically weighted to account for these oversamples, and the estimates presented reflect national population totals. The weights, provided by the data collection agency, are equal to the inverse of the sampling probability for each case, adjusted for nonresponse. We present our results in the form of univariate distributions and bivariate (Tables 1–3) and multivariate (Table 4) statistical analyses. To address the issue of outliers, we used the log of expenditures in our multivariate analyses. Standard errors and test statistics were derived using software that takes into account the complex sample design of the survey.19
Prevalence of Limitation of Activity
An estimated 5 million (7.3%) children younger than 18 years had a disability in 1999–2000. Prevalence of disability varied according to the sociodemographic characteristics of children and their families (Table 1). Disability was more prevalent among boys (P < .001) and among children in families with income <200% of the federal poverty level (FPL) than among children in higher income groups (400% of the FPL and above; P = .044). Prevalence was also higher among white non-Hispanic children compared with other racial and ethnic groups (P < .001). Prevalence was also positively associated with presence of insurance (P = .028).
Use of Health Services
Analysis of data from the MEPS indicates that children with disabilities used much higher levels of inpatient hospital services than other children (Table 2). Specifically, children with disabilities had >4 times the number of hospitalizations (P < .001) and spent 8 times as many days in hospitals as other children (P < .001).
MEPS data also show that children with disabilities used higher levels of physician and nonphysician services. Children with disabilities made more than twice as many physician visits (P < .001) and 5 times as many nonphysician visits as children without disabilities (P < .001). Nonphysician health professionals include nurse practitioners and physician assistants who work on their own or with a physician, as well as psychologists, social workers, physical therapists, and other professionals (see footnote in Table 2). Such large differences in use of nonphysician professionals are expected given the greater need for mental health and rehabilitative services among chronically ill children.
Children with disabilities had approximately twice as many emergency department visits as children without disabilities (P < .001). Children with disabilities also received ∼3 times the number of prescribed medications as other children (P < .001) and used substantially more home health provider days than other children (P < .001). No significant differences were found for use of dental services.
Expenditures for Health Services
Table 3 shows that total health care expenditures averaged $2669 for children with disabilities—∼4 times the average of $676 for children without disabilities (P < .001). Although children with disabilities comprised 7.3% of the 0 to 17 population, they accounted for 22.7% of total health care expenditures for children during 1999–2000.
Comparison of expenditures by type of health service shows differences that parallel those described above for utilization. For example, children with disabilities had almost 8 times higher expenses for hospital inpatient care as their counterparts without disabilities (P < .001). Children with disabilities had ∼3 times the average expenditure on physician services and nonphysician services as children without disabilities (P < .001 for both). Average expenditures on prescribed medications were ∼5 times higher than for children without disabilities (P < .001). As expected, expenditures for home health services were far greater for children with disabilities (P < .001). Average expenditures for children with disabilities on “other” medical services were ∼3 times those for other children (P < .001). In contrast, there were no significant differences in average expenditures for dental services.
Table 3 shows that average out-of-pocket expenditures (those paid by the family on behalf of the child) are ∼50% higher for children with disabilities than for other children (P < .001). Among children with disabilities, the largest component of out-of-pocket expenditures (in total dollars) was for prescription medications, comprising one third of their out-of-pocket expenses. In contrast, for children without disabilities, the largest component of out-of-pocket expenditures was for dental services, comprising more than half of their total out-of-pocket expenses.
Children with disabilities seemed to enjoy better financial protection against out-of-pocket health care expenses than children without disabilities, at least relative to their total health care expenses. That is, families paid ∼11% of health care bills out of pocket for a child with a disability and 28% for a child without a disability. However, as pointed out in the preceding paragraph, absolute out-of-pocket expenses are higher for children with disabilities.
Among children with disabilities, the share of health care bills paid out of pocket varied by type of service. Table 3 shows that ∼14% of physician services, 12% of nonphysician services, 30% of prescription medicine, and 67% of dental expenses were paid directly by families. In contrast, <1% of inpatient hospital expenses and home health services were paid out of pocket for children with disabilities. Hence, children with disabilities are best protected against hospital care and home nursing expenses and least protected for prescription drugs and dental care needs.
Total expenditures and out-of-pocket expenses for children with disabilities are highly skewed (Fig 1). The median total expense for health care was $462. When total health care expenditures are ranked, the upper decile of children with disabilities accumulated expenses of $4940 or more. This upper decile of children accounted for 64.9% of all health care expenses for children with disabilities. Out-of-pocket expenses for children with disabilities were also distributed in a skewed manner. The median annual out-of-pocket expense for health care was $57. When ranked according to out-of-pocket expenses, the bottom 20% of children with disabilities had no reported expenses, whereas the top 10% had expenses totaling $644 or more. This upper decile accounted for 84.8% of out-of-pocket expenses for children with disabilities. Hence, the burden of out-of-pocket expenses is highly concentrated among a small subset of the population with disabilities.
Predictors of Health Care Spending
We conducted 4 multivariate analyses to assess how characteristics of children with disabilities are related to total health care expenses and out-of-pocket expenses. The results of these analyses are presented in Table 4. Linear regression was used in the first 3 equations, and logistic regression was used in estimating the fourth equation. The results in Table 4 are presented as regression coefficients; for easier interpretation, the effect sizes are presented in percentage terms or odds ratios (for equation 4) in the text. In the first equation, we see that younger, black and Hispanic, and uninsured children with disabilities experienced significantly lower total health care expenses than other children with disabilities. The largest effect size is for insurance status; insured children had 172% percent higher health care expenses than their uninsured counterparts. After adjusting for other variables, poverty status and gender were not associated with total health care expenditures among children with disabilities.
The second column of Table 4 shows the results of our multivariate analysis of predictors for out-of-pocket expenses. As for total health care expenditures, younger children as well as black and Hispanic children experienced lower out-out-of pocket expenses. There is a steep gradient in out-of-pocket expenses by poverty status. Compared with children who live in the highest income families, children in low-income families had 59% lower out-of-pocket costs (P < .001). Children in households at 200% to 399% FPL had 31% lower out-of-pocket costs (P = .002). Insurance status was also closely related to out-of-pocket expenses. Out-of-pocket expenses for insured children were 46% lower than those for uninsured children (P = .004). No significant difference is indicated for gender.
We conducted 2 multivariate analyses to identify characteristics of children who experience high levels of financial burden. These analyses are shown in the last 2 columns of Table 4. Financial burden is expressed as the ratio of out-of-pocket expenses to family income (the log of the percentage of family income spent on the child's health care) in the third column and the likelihood that a child has potentially catastrophic out-of-pocket expenses (in excess of 5% of family income) in the fourth column. In both of these analyses, children with disabilities in low-income families experience significantly higher risk of incurring burdensome out-of-pocket expenses than children with disabilities in higher income families. Compared with households with incomes >400% FPL, children in households <200% FPL spent 172% more of their family income on health care for their children with disabilities (P < .001). There was an especially strong effect in the case of potentially catastrophic out-of-pocket expenses; children with disabilities in low-income families were ∼19 times more likely than their counterparts in higher income families to experience expenses that exceeded 5% of their families' income. Insurance status was significantly related to out-of-pocket expenses as a share of family income. Families with an insured child with disabilities spent 51% less of their income on health care for their child with disabilities as did families whose child was without insurance.
The MEPS provides new and current information concerning use and expenditures for a broad range of health services for the population of children with disabilities. Our findings demonstrate that children with disabilities used many more services than their counterparts without disabilities. As a result of their greater use, children with disabilities also had much higher health care expenditures. Out-of-pocket expenses were higher as well. We also found that the distributions of total and out-of-pocket expenses were highly skewed, with a small fraction of the population with disabilities accounting for a large proportion of expenditures. Finally, poverty status and insurance status were found to be important predictors of out-of-pocket expenses as well as financial burden.
Although differences in methods limit our ability to compare results from this study with the last in-depth national study of expenditures for children with disabilities conducted using 1980 data,11 some broad similarities are present between the findings from the 2 studies. Both studies show large differences in total expenditures between children with and without disabilities—on the order of 3-fold in the 1980 data set and 4-fold using the 1999–2000 data set. The share of total expenditures paid out of pocket for children with disabilities seems to have declined between 1980 and 1999–2000—perhaps a salutary effect of managed care. Both studies also demonstrate that total and out-of-pocket expenses are highly skewed, such that high expenses are concentrated within a small segment of the population with disabilities. Indeed, most children with disabilities had low or modest expenses in both studies. Finally, both studies demonstrate the vulnerability of low-income families and the importance of health insurance in providing financial protection for families of children with disabilities.
The skewed distribution of out-of-pocket expenses found in the present and earlier study indicates that financial burdens continue to be shared unevenly by families of children with disabilities. In fact, expenditures seem to be becoming even more skewed over time, indicating that families of children with disabilities continue to be exposed to significant risks to their financial well-being. Insurance coverage is designed to equalize these risks and to protect families against undue financial burden. Although insurance clearly is helpful, results from the present study show that low-income families experience heavier financial burdens when measured relative to their ability to pay, even after controlling for the presence of health insurance. Additional efforts are needed to protect these high-risk families. The expansions of public coverage for low-income families through the State Children's Health Insurance Program (SCHIP) should be helpful here. The effects of that program—enacted in 1997—are not fully reflected in our data. However, the benefit package in most separate SCHIP plans is not nearly as extensive as Medicaid,20 so even expanded coverage through SCHIP may not confer the financial protection needed by low-income families with children with disabilities. These programs could be strengthened by providing a wraparound package of benefits suitable to the needs of children with disabilities. Another approach is to provide greater funding under the Title V program to enhance the direct delivery of services for children with disabilities, especially for those services that are not well covered by insurance.
We would have liked to analyze differences in utilization and expenditures among children with disabilities by type of insurance coverage (eg, private, Medicaid, SCHIP). However, the small sample of children with disabilities, combined with the naturally large variations in health care expenditures within this population, precluded this level of analysis. It may be possible to combine future editions of the MEPS to conduct such analyses.
The financial burden of childhood disability continues to be shared unevenly by families. Low-income families are especially vulnerable to burdensome expenses. Additional efforts are needed to protect these high-risk children and their families.
We appreciate the editorial assistance provided by Lena Libatique.
This study was funded by the Maternal and Child Health Bureau, US Department of Health and Human Services (Grant No. U93MC 00180).
- Cadman D, Boyle MH, Offord DR, et al. Chronic illness and functional limitation in Ontario children: findings of the Ontario Child Health Study. CMAJ.1986;135 :761– 767
- ↵Newacheck P, Strickland B, Shonkoff J, et al. An epidemiologic profile of children with special health care needs. Pediatrics.1998;102 :117– 123
- ↵Eberhardt MS, Ingram DD, Makuc DM, et al. Urban and Rural Health Chartbook. Health, United States, 2001. Hyattsville, MD: National Center for Health Statistics; 2001:235– 237
- ↵Lozano P, Fisherman P, VonKorff M, Heet J. Health Care utilization and cost among children with asthma who were enrolled in a health maintenance organization. Pediatrics.1997;99 :757– 764
- ↵Chan E, Zhan C, Homer J. Health care use and costs for children with attention-deficit/hyperactivity disorder. Arch Pediatr Adolesc Medic.2002;156 :504– 511
- ↵Ringle J, Sturm R. National estimates of mental health utilization and expenditures for children in 1998. J Behav Health Res.2001;28 :319– 333
- ↵Newacheck PW, McManus M. Financing health care for disabled children. Pediatrics.1988;81 :385– 394
- ↵Neff JM, Muldoon JH, Gay JC, Anderson GF. Profile of the SSI population in Washington state medicaid program with a diagnosis-based classification system. J Cost Qual.2001;7 :11– 36
- ↵Kuhlthau K, Perrin JM, Ettner SL, McLaughlin TJ, Gortmaker SL. High-expenditure children with Supplemental Security Income. Pediatrics.1998;102 :610– 615
- ↵Kuhlthau K, Walker DK, Perrin JM, et al. Accessing managed care for children with chronic conditions. Health Aff (Millwood).1998;17 :42– 52
- ↵Cohen SB. Sample Design of the 1997 Medical Expenditure Panel Survey Household Component. Rockville, MD: Agency for Healthcare Research and Quality; 2000 (MEPS Methodology Report No. 11; AHCPR Publication No. 01-0001)
- ↵Hamilton L. Statistics With Stata. Version 7. Belmont, CA: Wadsworth Group; 2003
- ↵Fox H, Limb S, McManus M. SCHIP Innovations for Children With Special Health Care Needs in Managed Care. Resource Paper. Washington, DC: Center for Health Care Strategies; 2002
- Copyright © 2004 by the American Academy of Pediatrics