Health Care Use and Health and Functional Impact of Developmental Disabilities Among US Children, 1997-2005

Objective  To present nationally representative estimates of health-related limitations, needs, and service use among US children with and without developmental disabilities (DDs).

Design  Retrospective analysis of data from a sample of US households from the 1997-2005 National Health Interview Surveys.

Participants  Children aged 3 to 17 years (n = 95 132).

Main Outcome Measures  Parents or other knowledgeable adults reported on their children's DDs, health needs, and use of health and education services. Developmental disabilities included attention-deficit/hyperactivity disorder, autism, blindness, cerebral palsy, deaf/a lot of trouble hearing, learning disability, mental retardation, seizures, stuttering/stammering, and other developmental delay.

Results  Among children with 1 or more DDs, prevalence estimates for limitations in movement (6.1%), needed help with personal care (3.2%), needed special equipment (3.5%), received home health care (1.4%), and regularly took prescription medication(s) (37.5%) were 4 to 32 times higher than for children without DDs. Children with DDs were 2 to 8 times as likely to have had more than 9 health care visits (14.9%), received special education (38.8%), had a surgical or medical procedure (7.5%), and recently visited a medical specialist (23.9%), mental health professional (26.6%), therapist/allied health professional (19.6%), and/or emergency department (10.3%). Effects were generally stable during the study interval and independent of age, race, sex, and family income. Cerebral palsy, autism, mental retardation, blindness, and deafness/a lot of trouble hearing were associated with the highest levels of health and functional impact indicators.

Conclusions  Developmental disabilities profoundly affect children's health and functioning. These data can inform evidence-based targeted prevention strategies for minimizing functional limitations and lifetime disability. Additional study of unmet needs and access to care is needed.

Developmental disabilities (DDs) are chronic physical, cognitive, speech or language, psychological, or self-care conditions that typically originate during childhood; are likely to continue indefinitely; and require additional coordinated services, support, or other assistance for an extended duration or during a lifetime.^1,2 Developmental disabilities, including conditions such as intellectual disability, deafness, blindness, cerebral palsy, and autism,² represent a subset of conditions that affect children with special health care needs.³ Although patterns of health care use among children with special health care needs have been previously described in general,^4-6 there is limited information of this type for children with individual DDs and a paucity of population-based data covering the full range of DDs in children.

Boyle et al⁷ previously calculated national estimates of the health impact of DDs among children and reported significantly higher rates of physician visits, hospitalizations, and lost school days among children with DDs compared with children without; however, these data are nearly 20 years old. More recently, Msall et al⁸ reported that neurodevelopmental disorders such as cerebral palsy, seizures, and autism spectrum disorders were related to multiple functional limitations, and approximately 35% of children with neurodevelopmental disorders needed or received special education services compared with 2.4% of children with no identified medical disabilities. Other studies based on nationally representative samples and data from large group-model health plans document increased rates of outpatient and emergency department visits and prescription medication use among children with autism and attention-deficit/hyperactivity disorder (ADHD)^9-13 and increased frequencies of physician contacts and hospitalizations among children who are blind or have vision loss or are deaf or have hearing loss.¹⁴

Updated and more comprehensive population-based estimates of functional limitations and health services utilization among children with a range of disabilities are needed to better inform health and education planning for this population. Furthermore, as we move toward earlier identification and intervention, it is crucial to have high quality data on this population's current health status to guide decisions on program development and resource allocation and to measure the potential impact of intervention strategies. We used national data to describe overall and condition-specific health and functional status and health care and special educational services use among children 3 to 17 years of age with a broad range of DDs.

The data used for this analysis were derived from the 1997-2005 National Health Interview Survey (NHIS). The NHIS is a multistage probability sample survey that is conducted annually to monitor the health of the US civilian noninstitutionalized population.¹⁵ Basic demographic and health information on all household members is collected during an in-person interview with a knowledgeable household member. One child younger than 18 years is randomly selected from each family with children, and information pertaining to that child is collected from a knowledgeable adult, usually (>90% of interviews) a parent or other legal guardian. (For the duration of this article, we use the term parent in place of parent or legal guardian when describing survey respondents.) The Child Sample Core questionnaire was the primary data source for our analysis and was supplemented by demographic and other household information. The unweighted sample size for the 1997-2005 NHIS was 117 520 children, of whom 96 345 were 3 to 17 years of age. Of those, approximately 98.7% (n = 95 132) provided information on the DDs included in this analysis. Sample sizes were similar for all 9 survey years.

Children younger than 3 years were excluded because many conditions are not routinely diagnosed in infancy and very early childhood. All DDs listed in the survey were included in the analysis: attention-deficit disorder (ADD) and ADHD, autism, blindness, cerebral palsy, deaf/a lot of trouble hearing, learning disability, mental retardation, seizures, stuttering or stammering, and other developmental delay. Data about most conditions (ADD/ADHD, autism, cerebral palsy, learning disability, mental retardation, and other developmental delay) were ascertained from identically worded questions that asked about any past medical diagnosis. The condition was considered to be present if the knowledgeable adult answered in the affirmative when asked, “Has a doctor or health professional ever told you that [child's name] had [condition]?” For seizures and stuttering or stammering, the questions pertained to episodes occurring during the past 12 months only. Blindness was ascertained from the question, “Is [child's name] blind or unable to see at all?” Deafness/a lot of trouble hearing was positively identified if the respondent indicated that “deaf or a lot of trouble hearing” best described the child's hearing without a hearing aid. The current accepted term for mental retardation is “intellectual disability”¹⁶; however, since the NHIS survey specifically queried parents about mental retardation, we used this term in the tables and text. Because substantial collinearity was noted between learning disability and mental retardation, children with reported learning disability and mental retardation were not included in the learning disability category. These children were included in the mental retardation category; however, learning disability was not counted as a co-occurring DD for children reported as having mental retardation.

We examined the variation in the prevalence of each condition by the child's age and sex, household income (poverty was defined as annual income <200% of the federal poverty level, or approximately $39 000 for a family of 4 in 2004 dollars¹⁷), maternal educational level, and race. We also examined the frequency of co-occurring DDs, and, for each relevant DD, we report the most frequent other co-occurring DDs. Cerebral palsy was excluded from the analyses of the 2004 and 2005 survey years because of known data collection errors.

Health and functional impact was assessed via a series of 5 questions that were similar to those used in other national studies¹⁸ and that were related to the definition created by the Maternal and Child Health Bureau.³ The questions address functional limitations lasting or expected to last longer than 12 months, use of special equipment (excluding ordinary eyeglasses or corrective shoes), needing help with personal care needs (eg, eating, bathing, or dressing), use of prescription medication for at least 3 months for a current problem, and use of home health care in the past year. A separate category (≥2 health status indicators) was created to further identify those children for whom an affirmative response was noted for 2 or more of the health effect variables. We also assessed parent report of the child's health status; however, this variable was not included in the 2 or more health status indicators category.

Health care use during the past 12 months was evaluated according to the types of health care providers seen (eg, medical specialist other than an obstetrician/gynecologist, psychiatrist, or ophthalmologist; allied health professional such as physical, speech, respiratory, occupational, or audiology therapist; and mental health professional such as a psychiatrist, psychologist, or clinical social worker) and the frequency of visits to a physician's office or clinic (excluding hospital and emergency department visits) and to a hospital emergency department. Use of special educational or early intervention services was also examined. Because of the high prevalence of co-occurring conditions, we also conducted analyses of both health and functional impact and health care utilization limited to children with an isolated DD (ie, those without a co-occurring DD), with the exception of autism, mental retardation, and other developmental delay, which were isolated in only 4.6%, 17.1%, and 23.9% of affected children, respectively.

The estimates and standard errors presented in this article were weighted to reflect the national noninstitutionalized population of children 3 to 17 years of age. SUDAAN,¹⁹ a statistical software package designed to analyze data from complex sample surveys, was used to generate all estimates, as well as standard errors and 2-tailed χ² tests. Effect modification was assessed by comparing all associations across sex, age, maternal educational level, poverty level, and race. Because no remarkable variations from the unstratified analyses were observed, we do not present the stratified results here. We evaluated temporal trends in the health impact and utilization variables; only regular prescription medication use showed a clear pattern of trends across several DDs. Therefore, we further assessed temporal trends in prescription medication use with odds ratios, comparing regular prescription medication use during early vs late study years. Odds ratios were adjusted for child's age, income level, sex, race, and maternal educational level.

Table 1 depicts the weighted prevalence and demographic characteristics of DDs among children 3 to 17 years of age from 1997 through 2005. Approximately 13% (13.2%) of children were reported to have 1 or more DDs, whereas 1.6% had 3 or more DDs. The most frequently reported conditions were learning disability (in the absence of mental retardation) (7.0%), ADD/ADHD (6.4%), and other developmental delay (3.5%). Approximately 67% of children with 1 or more DDs were male. Among the individual DDs, the predominance of males was highest for autism (82.8%) and lowest for cerebral palsy (51.0%). Children with reported autism and stuttering were, on average, younger than children without a DD, whereas children with several other DDs, including ADD/ADHD and learning disability, were generally older. With the exception of children with autism, children with a DD were more likely than children without a DD to live in a household with an annual income less than 200% of the federal poverty level.

Among children reported to have a DD, 41.4% had multiple disabilities (Table 2). The percentage with co-occurring conditions ranged from 43.3% for children who were blind to 95.6% for children with autism. With the exception of other developmental delay, learning disability (without report of mental retardation) was the most frequently reported co-occurring condition for children with ADD/ADHD, autism, blindness, cerebral palsy, deafness/a lot of trouble hearing, seizures, and stuttering. Other commonly reported co-occurring conditions among children with 2 or more DDs included ADD/ADHD, mental retardation, seizures, and stuttering.

The proportion of children with 1 or more DDs who were limited in their ability to walk, run, or play; needed help with personal care; needed special equipment; reported long-term use (>3 months) of prescription medication; and received home health care during the past year was 4 to 32 times greater than among children without DDs (Table 3). For every DD assessed, we observed a significant increase for each health and functional status indicator in comparison with children with no reported DDs. However, a fairly wide variation was observed. Children with ADD/ADHD were least affected for 4 of 5 indicators assessed. In contrast, 40.8% of children with cerebral palsy were limited in their ability to crawl, walk, run, or play; children with mental retardation, cerebral palsy, or autism most often needed help with personal care (31.2%, 29.3%, and 28.0%, respectively); children with cerebral palsy and/or who were deaf/had a lot of trouble hearing most frequently used special equipment (31.0% and 28.4%, respectively); and children with cerebral palsy were the most frequent users of home health care in the past year (10.3%). Prevalence of prolonged prescription medication use was more than 30% for all DDs and was highest among children with seizures (64.1%), ADD/ADHD (55.9%), mental retardation (50.8%), and autism (49.1%). Children with cerebral palsy, mental retardation, or blindness were most likely to have 2 or more of the health and functional status indicators (≥30% for each condition). Parental report of health status substantiated these findings, with 6.1% of children with 1 or more DDs and 17.5% of children with 3 or more DDs reported to have fair or poor health. Although some variation was observed, parents of children with each individual DD reported higher levels of fair or poor health status than did parents of children without DDs.

In analyses limited to children with isolated DDs, statistically significant increases were still observed for most health and functional status indicators. However, for each DD assessed, the prevalence estimates of nearly all indicators were substantially reduced among the subset with isolated DDs compared with the total sample, indicating that the severity of individual conditions is likely greater in cases in which they co-occur with other disabilities.

All measures of health care use were higher among children with DDs compared with children without DDs (Table 4). Children with DDs were 2.2 times more likely to have been seen by a medical specialist, 7.8 times more likely to have been seen by a mental health professional, and 5.9 times more likely to have visited a therapist or allied health professional during the past year. A 2-fold increase in the number of visits to emergency departments and surgical or medical procedures in the past year for children with a DD compared with children without a DD was also noted. Children with seizures were most likely to have seen a medical specialist in the past year (56.7%), and children with autism were most likely to have recently seen a mental health care professional (54.8%) or therapist (60.5%). Frequent office visits (>9 in the past year) were common across all DD categories, with approximately one-fifth to one-quarter of children with DDs falling in this range. Children with seizures and those reported to be deaf/have a lot of trouble hearing were more likely than children with other DDs to have 2 or more visits to the emergency department in the past year. Surgery or medical procedures in the past year were most common among children with blindness.

Use of special education services was significantly higher among children in all disability groups (range, 36.6%-87.1%) than among children without a DD (1.5%). Approximately 38.8% of children with 1 or more DDs and 78.6% of children with 3 or more DDs received special education or early intervention services. Special education was reported for most children with autism (87.1%) and mental retardation (80.8%) and was less frequent in the other disability groups.

As noted with health and functional status indicators, the rates of most health care utilization indicators were lower among children with isolated DDs than among the total sample of children with each DD. In some instances, the differences were substantial. For example, when the total sample of children with cerebral palsy were examined, the rates of each health utilization indicator examined were 2 to 4 times higher than the rates among children with isolated cerebral palsy; use of special education services was 6.5 times higher among all children with cerebral palsy than among those with isolated cerebral palsy. Nevertheless, the children with each of the isolated DDs still had significantly higher service utilization than children without DDs.

We examined temporal trends for each health and functional status measure and health or education services use indicator in Tables 3 and 4. We observed only a few noteworthy trends in health services use; there were significant (P <.05) increases in the use of therapeutic services (ie, physical, respiratory, speech, or occupational therapy) for children with seizures (1997-2000, 23.9%; 2001-2005, 34.4%) and frequent office visits (>9) for children who were blind (1997-2000, 13.0%; 2001-2005, 32.6%). For most health and functional status indicators there was little change in prevalence during the study period for any of the disabilities (data not shown). However, temporal trends were noted for prescription drug use (Table 5). Children with autism were 2.4 times more likely to have regularly taken a prescription medication for at least 3 months from 2001 through 2005 than from 1997 through 2000. Slight but statistically significant temporal increases in prescription medication use were also noted for children with ADD/ADHD, learning disabilities, and other developmental delays.

Developmental disabilities have a profound effect on a child's health and functioning, as evidenced by the comparatively high rates of health care and special education service utilization noted in this report. With a few exceptions, this effect was independent of age, race/ethnicity, sex, and family income and changed little during the 9-year study period. The effect varied across the spectrum of DDs assessed, with autism, blindness, cerebral palsy, deafness/a lot of trouble hearing, and mental retardation showing the greatest impact. Although the impact was reduced among children with isolated DDs, notable increases in many health and special education services remained, compared with children without DDs. Because health promotion for this population is an important public health goal, these data provide vital information for ongoing monitoring of the health and functional impact of these chronic disorders and can help inform prevention strategies for minimizing impairment and lifetime disability and reducing health disparities.

Our findings substantiate, update, and expand on a previous study using the NHIS and several other studies. An earlier analysis based on 1988 NHIS data indicated that the prevalence of DDs was approximately 16.8%,⁷ whereas our study reported a prevalence of 13.2%. Differences in the overall prevalence rates in the 2 surveys is likely because certain rates of DDs (eg, autism, ADD/ADHD, and mental retardation) were not directly assessed in the 1988 survey but instead were determined through broad questions on functioning. Data from various other studies also document major health care and education needs among children with various specific disabilities. Halfon and Newacheck²⁰ reported that 79.9% of children 5 to 17 years of age with disabling mental health conditions, including autism, ADHD, a developmental delay, a learning disability, or mental retardation, received special education services from 1992 through 1994 and that 5.3% of children with these DDs were hospitalized during the previous year, compared with 2.8% of children without these DDs. Gurney et al²¹ analyzed data from the 2003 National Survey of Children's Health and reported that children with autism had higher frequencies of emergency department visits; physician visits; physical, occupational, or speech therapy; and treatment or counseling for an emotional, developmental, or behavioral problem than children without autism. Increased rates of prescription medication use among children with autism^10,11,13,21 and ADD/ADHD²² also have been reported, and there is some evidence that medication use among individuals and youth with autism has increased over time.²³

In the present study, we also assessed secular trends of health impact indicators. The level of effect has remained stable for most indicators; however, the use of prescription medications among children in several disability groups has increased. Although it is difficult to determine if the few observed increases in service and prescription medication use reflect an increased severity in conditions, increased access to appropriate services, or temporal changes in clinical practice (eg, use of psychotropic medications for children with autism), the data nevertheless demonstrate a substantial level of need for children with DDs. Moreover, the continued stable high rates of health impact factors, such as limited ability to walk, run, and play and the need for help with personal care, might provide a strong argument for increased attention to the development of effective early intervention services and further study of the barriers families currently face in accessing existing services.

The findings presented in this article are subject to several limitations. Because the data are limited to noninstitutionalized children, some measures may be underestimated. However, US Census data indicate that only 0.2% of children were living under formally authorized, supervised care or in the custody of institutions in 2000.²⁴ Because these data are cross-sectional, the causal effect of DDs on health status and service use cannot be inferred. Furthermore, although our analyses of isolated DDs provided insights, we lacked specific data to determine whether the health and functional impact observed in the total population were attributable to the cumulative effect of multiple conditions or the severity of a given condition. We also did not evaluate underlying medical conditions that might co-occur with DDs, such as metabolic disorders, sickle cell disease, or asthma. Because the data were ascertained via reports from parents or other knowledgeable caregivers, there is a possibility of a degree of inaccurate recall owing to parental distress,^25,26 stigma associated with certain conditions, misunderstanding the survey question(s), or variations in professional vs parental terminology.²⁰ Although data on the validity of parental report of developmental conditions are limited, comparison of the parent-reported autism rate in the NHIS and the National Survey of Children's Health suggested high reliability and internal convergent validity.²⁷ A separate study reported moderate to high sensitivities for parent-reported vision impairment and behavioral disorders but low sensitivity for parent-reported hearing impairment.¹² Another study found moderate levels of agreement between maternal report and medical records for conditions such as asthma, cerebral palsy, and spina bifida but poor agreement for eating and swallowing problems and delayed speech.²⁸ Although the discordance for some of the conditions assessed in the study may have been attributable to maternal overreport, it was noted that some serious conditions were not identified in the medical records. Therefore, the authors concluded that the discrepancies may be owing to access to care, communication issues, variations in understanding about children's health issues, and recall problems.

This study provides important information on the patterns of health care use and the health and functional impact of DDs in US children. Elevated rates of health care use, functional limitations, and medication use among children with DDs may be reflective of the potential impact of these conditions on children, families, and society. Compared with children without DDs, children with DDs are more likely to have poorer self-esteem, more depression and anxiety, and more problems with learning.²⁹ In addition, families of children with DDs can face substantial financial and caregiver burdens.^30,31 Furthermore, these findings suggest that children with DDs use disproportionately high rates of health services relative to children without DDs. Additional research is warranted and should include further evaluation of early intervention services and potential population effects, trends in access to care, and the prevalence of unmet health care needs among children with DDs.

Correspondence: Sheree L. Boulet, DrPH, MPH, National Center on Birth Defects and Developmental Disabilities, Centers for Disease Control and Prevention, 1600 Clifton Rd, Mail Stop E86, Atlanta, GA 30333 (sboulet@cdc.gov).

Accepted for Publication: May 27, 2008.

Author Contributions: All authors had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Boulet, Boyle, and Schieve. Acquisition of data: Boulet.Analysis and interpretation of data: Boulet, Boyle, and Schieve. Drafting of the manuscript: Boulet and Schieve. Critical revision of the manuscript for important intellectual content: Boulet, Boyle, and Schieve. Statistical analysis: Boulet and Schieve.Study supervision: Boyle and Schieve.

Financial Disclosure: None reported.

Funding/Support: This study was supported in part by an appointment to the Research Participation Program at the National Center on Birth Defects and Developmental Disabilities (Dr Boulet), administered by the Oak Ridge Institute for Science and Education through an interagency agreement between the US Department of Energy and the Centers for Disease Control and Prevention.

Disclaimer: The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.