Health Care Costs Associated With Parent-Reported ADHD: A Longitudinal Australian Population

Abstract

Objective: To examine the health care costs associated with ADHD within a nationally representative sample of children. Method: Data were from Waves 1 to 3 (4-9 years) of the Longitudinal Study of Australian Children (N = 4,983). ADHD was defined by previous diagnosis and a measure of ADHD symptoms (Strengths and Difficulties Questionnaire [SDQ]). Participant data were linked to administrative data on health care costs. Analyses controlled for demographic factors and internalizing and externalizing comorbidities. Results: Costs associated with health care attendances and medications were higher for children with parent-reported ADHD at each age. Cost differences were highest at 8 to 9 years for both health care attendances and medications. Persistent symptoms were associated with higher costs (p < .001). Excess population health care costs amounted to Aus$25 to Aus$30 million over 6 years, from 4 to 9 years of age. Conclusion: ADHD is associated with significant health care costs from early in life. Understanding the costs associated with ADHD is an important first step in helping to plan for service-system changes.

Keywords

ADHD health care costs economic medication children

ADHD is a highly prevalent neurodevelopmental disorder affecting approximately 5% of school-aged children worldwide (Polanczyk, de Lima, Horta, Biederman, & Rohde, 2007) with up to 60% to 100% of children with ADHD presenting with one or more comorbid disorders (Gillberg et al., 2004). It is characterized by developmentally inappropriate levels of inattention and/or hyperactivity/impulsivity, which result in significant impairments in academic, social, and family functioning (American Psychiatric Association, 2000). These impairments persist into adolescence and adulthood (Barkley, Fischer, Edelbrock, & Smallish, 1990; Biederman et al., 2006). Comorbidities are the rule, rather than the exception in ADHD. In Australia, general pediatricians are the main health care providers for children with ADHD, with smaller proportions being managed (or comanaged) by psychologists, psychiatrists, and other health care service providers. A recent 2-week audit of Australian pediatricians’ practice revealed that 18% of all consultations involved children with ADHD and that 77% of children with ADHD were prescribed stimulant medication (Hiscock et al., 2011). It is not surprising that previous studies have demonstrated that ADHD is associated with significant health care costs (Doshi et al., 2012; Pelham, Foster, & Robb, 2007). However, little is known internationally about age-related differences in health care costs or the costs associated with persistent ADHD. Furthermore, the costs to the Australian health care system associated with ADHD are unknown. Knowledge about the health care costs associated with ADHD is important to better understand the societal burden associated with the disorder.

Existing studies have found that ADHD is associated with significant health care costs, with the majority of these studies being conducted in the United States (Doshi et al., 2012; Pelham et al., 2007). These comprise costs associated with purchasing ADHD medications, as well as the cost of health care attendances. Studies examining the health care costs of children within European countries have also found that ADHD incurs significant health care costs (De Ridder & De Graeve, 2006; Hakkaart-van Roijen et al., 2007; Schlander, Trott, & Schwarz, 2010). The majority of these studies have used parent reports of ADHD diagnosis or diagnostic codes (entered into databases by health professionals) to define ADHD status, which provide insight into the costs associated with children who have been formally diagnosed with ADHD (Doshi et al., 2012). Few studies have examined the relationship between actual ADHD symptoms, rated by parents for example, and health care costs. This comparison is important as children may present with significant levels of inattention and/or hyperactivity, especially at younger ages, without a formal diagnosis of ADHD.

Approximately 75% of ADHD cases have an onset by age 8, and 95% by age 11 (Kessler et al., 2005), yet previous research examining the health care costs in this group have mainly examined costs for older children and adolescents or have examined a wide age range of patients with ADHD (e.g., 3-17 years), without examining age-related differences in costs (Chan, Zhan, & Homer, 2002; Guevara, Lozano, Wickizer, Mell, & Gephart, 2000). Very little is known about the specific health care costs for younger children and while ADHD symptoms persist into adulthood for approximately 65% of children with ADHD (Faraone, Biederman, & Mick, 2006), research has yet to examine how the persistence of ADHD symptoms relates to the costs associated with the disorder. Previous studies are also limited by small sample sizes, reliance on parent report of health care attendances, failure to take into account both internalizing and externalizing comorbidities, and failure to examine costs associated with both health care attendances and medication prescriptions (Doshi et al., 2012; Pelham et al., 2007).

This study aimed to quantify the nonhospital health care costs associated with ADHD from 4 to 9 years of age, using data from the population-based Longitudinal Study of Australian Children (LSAC). In Australia, ambulatory public health care is funded predominantly by the federal government through Medicare, a universal subsidized health care insurance scheme; therefore in this study, we measured health care costs using administrative data from the Medicare Benefits Scheme (MBS; health care attendances) and Pharmaceutical Benefits Scheme (PBS; prescription medications). This study aims to add to the existing literature by examining the health care costs for younger children with ADHD and to explore the impact of persistent ADHD symptoms on health care costs. To do this, we examined health care costs for children with ADHD within a nationally representative sample that has been linked to administrative data on health care costs. More specifically, we aimed to investigate the following:

The concurrent MBS, PBS, and total Medicare (MBS and PBS combined) health care costs for children identified with and without ADHD diagnosis/clinical symptoms from ages 4 to 9 years;

The excess total Medicare expenditure associated with ADHD diagnosis/clinical symptoms from age 4 to 9 years, at the Australian population level; and

Whether concurrent total Medicare costs vary according to the persistence of ADHD symptoms between 4 and 9 years of age.

Method

Study Design and Participants

LSAC is a population-based, longitudinal study of children’s health and development. Children were selected from the Australian Medicare database using a two-stage cluster sampling design (Gray & Sanson, 2005; Soloff, Lawrence, & Johnstone, 2005). After stratifying by state of residence and urban versus rural status, postcodes (except the most remote) were sampled. For the Kindergarten (K) cohort, children from the sampled postcodes were selected if they were born between March 1999 and February 2000 and enrolled in the Australian Medicare database (98% of all 4-year-old Australian children are registered). Of the contactable families selected, 4,983 four- to five-year-old children (59% response rate) participated in Wave 1 data collection, which commenced in 2004. Children were included in the current study if they had complete data on either of the ADHD measures (88%-100% depending on the ADHD measure and wave) and successful Medicare data linkage (92%). The K cohort were followed up at 6 to 7 (Wave 2: n = 4,464; 90% retention) and 8 to 9 years (Wave 3: n = 4,331; 87% retention; Misson & Sipthorp, 2007), with linked Medicare (health care cost) data.

Measures

ADHD status

We used two measures to define ADHD status: parent-reported ADHD and a measure of ADHD symptoms.

Parent-reported ADHD: During the parent interview at Waves 1 to 3, parents were asked “Does the study child have attention-deficit disorder or attention-deficit/hyperactivity disorder?” If parents responded “yes,” the child was classified as having ADHD. We have used this case definition in a previous study using LSAC data (Sciberras, Ukoumunne, & Efron, 2011) and found that children with parent-reported ADHD were very similar to those with clinically elevated ADHD symptoms on the Strengths and Difficulties Questionnaire (SDQ). For ease of interpretation, we will refer to this category as “children with ADHD” throughout the manuscript.

ADHD symptoms: Inattention/hyperactivity symptoms were assessed at Waves 1 to 3 using the parent-reported SDQ. The SDQ is a widely used and validated 25-item measure of behavioral/emotional problems for children aged 4 to 16 years (Goodman, 2001). We used the 5-item Hyperactivity-Inattention score in this study. Possible total scores range from 0 to 10, with higher scores indicating more symptoms. We categorized children as having clinical ADHD symptoms if their Hyperactivity-Inattention scores were ≥90th percentile (score of 8 or above) in this sample. The parent-reported Hyperactivity-Inattention scale has good internal consistency (α = .77; Hawes & Dadds, 2004). Children scoring ≥90th percentile on the Hyperactivity-Inattention subscale are approximately 18 times more likely to meet the Diagnostic and Statistical Manual for Mental Disorders (4th ed.; DSM-IV; American Psychiatric Association, 1994) criteria for ADHD than other children (Hawes & Dadds, 2004). For ease of interpretation, we will refer to children in this category as having “clinical ADHD symptoms” throughout the manuscript.

Internalizing and externalizing problems

The Emotional Problems and Conduct Problems scales from the SDQ were used to define clinical levels of internalizing and externalizing problems, respectively (Goodman, 2001). Each scale comprises 5 items with possible total scores ranging from 0 to 10. Children were defined as having an internalizing or externalizing problem if their scores on these scales were ≥90th percentile. This equated to a score of 5 or more on the Emotional Problems scale and a score of 4 or more on the Conduct Problems scale. Children scoring ≥90th percentile on the Emotional Problems scale are 12 times as likely to meet DSM-IV criteria for an internalizing disorder, and those scoring ≥90th percentile on the Conduct Problems scale are 31 times as likely to meet DSM-IV criteria for an externalizing disorder (Hawes & Dadds, 2004).

Health care costs

Data from the national MBS and PBS, together comprising the Australian Medicare subsidized health care scheme, were used to estimate the costs to the Australian federal government for health care attendances and prescription medication outside the hospital setting (Commonwealth Department of Health and Aged Care, 2000; Willis & Reynolds, 2008). Health care in Australia is funded through a combination of state and federal government (70%), private health and other insurance (13%), and individual out-of-pocket expenses (17%) (Australian Institute of Health and Welfare, 2012). Through the MBS, the Federal government subsidizes mainly nonhospital-based medical practitioners for providing care up to a predetermined amount, with any “gap” costs paid by the patient. The majority of services covered by the MBS are visits to health professionals, although it also includes nonhospital diagnostic and pathology services and treatment (Commonwealth Department of Health and Aged Care, 2000; Willis & Reynolds, 2008). In addition to total MBS costs, we report on three subcategories that reflect the three most common MBS service types relevant to ADHD: (a) general practitioner; (b) specialist (e.g., pediatrician, child psychiatrist); and (c) allied health clinician (e.g., psychologist, speech pathologist).

A predetermined list of medications are subsidized by the PBS to approximately 83% of the cost, with the remainder paid by the patient (Willis & Reynolds, 2008). In addition to total PBS costs, we examined costs in three groups:

Core ADHD medications: including methylphenidate immediate release (Ritalin 10), methylphenidate biphasic (Ritalin LA), methylphenidate extended release (Concerta), dexamphetamine immediate release, and atomoxetine (Strattera).

Other centrally acting medications: including other psychotropic and antiepileptic medications that were not classified as core ADHD medications; and

Other medications: including all other medications, which predominantly comprised antibiotic medications.

Health care costs were obtained by linkage with Australian Medicare data, with consent from the child’s primary caregiver. Costs were inflated to 2012 Australian dollars (Aus$1.00 = US$1.04, as of March 18, 2013) using the consumer price index figures provided by the Australian Bureau of Statistics (www.abs.gov.au). Costs for each child were estimated over a 6-year period, from the child’s 4th up to their 10th birthday. Costs were presented as 2-year costs, matched to each of the three biennial collections of data on ADHD status.

Demographic characteristics

The child’s primary caregiver provided information at baseline about the child’s age and gender, the mother’s primary language at home (English/not English), single-parent status (yes/no), and maternal high school completion (yes/no).

Analyses

Analyses were conducted in Stata version 12.0, using survey methods to weight the analyses to account for the unequal probability of participant selection, nonresponse and sample attrition, and the multistage, clustered sampling design (Soloff, Lawrence, Misson, & Johnstone, 2006). Following inspection of the data, we defined outliers as cases with costs more than Aus$10,000 for MBS data or more than Aus$8,000 for PBS data in any single year. Three outliers at each wave were identified and excluded from analyses.

Overall, mean Medicare (MBS + PBS) costs were higher than median costs indicating that the costs data were positively skewed. Our analyses were based on mean group differences as these are regarded as more representative of expected costs at a population level, and the use of linear regression to estimate these parameters is supported in large public health data sets (Lumley, Diehr, Emerson, & Chen, 2002). The bootstrap method would have ideally been used to ensure robustness of the standard error given the skewed data; however, this method could not be used in conjunction with the survey methods (weighting and first-order Taylor linearization) in place to account for the complex survey design. A comparison of mean costs with and without bootstrapping using the data without adjusting for the survey design yielded minimal differences (Aus$1-Aus$5 maximum difference in 95% confidence interval [CI] bounds).

Summary statistics were estimated for each cost category (MBS, PBS, and total combined) for each ADHD case definition: (a) children with and without ADHD at each age; and (b) children with and without clinical ADHD symptoms at each age. General linear model regression analyses were used to estimate the mean difference in total combined Medicare, MBS (including service type), and PBS (including prescription type) costs (and 95% CIs) between groups for each ADHD case definition separately. Adjusted mean differences and 95% CIs were calculated using covariates selected a priori (child age in months and gender, single-parent family, mother’s main language spoken at home, maternal high school completion, internalizing problems, externalizing problems). We were unable to control for other commonly occurring comorbidities, including learning disorders and autism spectrum disorders, due to an absence of reliable measures of these comorbidities across all waves of data collection. Participants were included in adjusted analyses if they had complete ADHD data relevant to the case definition being examined at each age, in addition to complete data for all covariates examined. Yearly population cost estimates were calculated using population size estimates from Australian 2009 population statistics for children aged 4 to 9 years (Australian Bureau of Statistics, 2010).

We also used general linear model regression analyses to estimate the Medicare costs (and 95% CIs) associated with persistent clinical ADHD symptoms across the three waves of data collection (no clinical ADHD symptoms at any wave, clinical symptoms at one wave, and clinical symptoms at two or more waves). We included participants with parent-reported ADHD symptoms data available at all three waves for this analysis.

Results

Sample Characteristics

The prevalence of parent-reported ADHD ranged from 1.3% at age 4-5 years to 2.6% at 8-9 years, while the prevalence of clinical ADHD symptoms ranged from 6.0% at age 4-5 years to 6.4% at age 8-9 years. Table 1 shows the demographic characteristics by ADHD status at age 8-9 years, the age at which the prevalence of ADHD was highest. Children with ADHD were slightly older than those without ADHD and were more likely to be male, to have a mother who did not complete Year 12, and were less likely to have a mother who spoke a primary language other than English. Children with ADHD also had significantly higher rates of internalizing and externalizing problems. Similar patterns were observed for children with clinical ADHD symptoms. Children with clinical ADHD symptoms were also more likely to be living in a single-parent family. Sample characteristics did not differ significantly between children with ADHD and children with clinical ADHD symptoms. Although not shown, consistent patterns were observed in earlier waves.

Table 1.

Sample Characteristics^a for Children With and Without ADHD at 8 to 9 Years.

	Children with ADHD			Children with clinical ADHD symptoms
	Yes (n = 95)	No (n = 3,930)	p value	Yes (n = 238)	No (n = 3,796)	p value
Child age in months, M (SD)	106.3 (2.7)	105.6 (2.9)	.03	105.5 (2.9)	105.5 (2.9)	.97
Male, %	82.0	50.5	<.001	77.3	49.7	<.001
Mother speaks language other than English, %	6.6	17.5	.03	9.6	16.0	.04
Single-parent family, %	15.5	16.9	.44	20.0	16.7	.008
Mother did not complete high school, %	64.7	51.1	.01	63.0	49.0	<.001
Clinical internalizing problems,^b %	29.6	7.4	<.001	22.9	6.9	<.001
Clinical conduct problems,^c %	43.3	9.4	<.001	39.7	8.0	<.001

Note. LSAC = Longitudinal Study of Australian Children; SDQ = Strengths and Difficulties Questionnaire.

All proportions are weighted and adjusted for LSAC sample design.

SDQ Emotional Problems subscale.

SDQ Conduct Problems subscale.

MBS Cost Differences by ADHD Status

The 2-year MBS costs by ADHD status are outlined in Table 2. Patterns of cost differentials were similar irrespective of the approach used to define ADHD status, with mean MBS costs being significantly higher for children with ADHD and children with clinical ADHD symptoms at each age examined. MBS cost differences remained significant, when adjusted for demographic characteristics and child internalizing and externalizing problems. MBS costs appeared to increase with age for children with ADHD. For example, the mean difference in MBS costs at 4 to 5 years was Aus$313 (unadjusted: approximately 68% higher than children without ADHD—calculated by dividing the costs for children with ADHD by the costs for children without ADHD), with this mean difference increasing to Aus$666 at 8 to 9 years (unadjusted: approximately 156% higher than children without ADHD). A very similar pattern was observed for children with clinical ADHD symptoms. The mean difference in MBS costs for children with clinical ADHD symptoms at 4 to 5 years was Aus$171 (unadjusted: approximately 38% higher than for children without clinical ADHD symptoms) with this mean difference increasing to Aus$338 at 8 to 9 years (unadjusted: approximately 80% higher than for children without clinical ADHD symptoms). In contrast, the MBS costs for children without ADHD (by either case definition) remained stable from 4 to 9 years.

Table 2.

Unadjusted and Adjusted 2-Year MBS Cost Estimates by ADHD Group.

		MBS cost (Aus$) per child by ADHD status^a
		M (SE)		Mean difference [95% CI]
Age	n (yes)	Yes	No	Unadjusted	p value	Adjusted^b	p value^b
Children with ADHD
4-5 years	52	771 (96)	458 (8)	313 [122, 504]	.001	398 [137, 660]	.003
6-7 years	61	924 (86)	393 (8)	531 [360, 702]	<.001	494 [293, 695]	<.001
8-9 years	95	1,092 (97)	426 (10)	666 [474, 858]	<.001	734 [500, 967]	<.001
Children with clinical ADHD symptoms
4-5 years	262	622 (38)	450 (8)	171 [95, 248]	<.001	184 [87, 282]	<.001
6-7 years	238	548 (38)	393 (9)	156 [79, 232]	<.001	145 [59, 232]	.001
8-9 years	205	760 (61)	422 (11)	338 [217, 459]	<.001	284 [167, 402]	<.001

Note. MBS: Medicare Benefit Schedule; CI = confidence interval.

Costs over 2 years, in 2012 Aus$, rounded to the nearest dollar.

Adjusted for child age in months and gender, single-parent family, mother’s main language spoken at home, maternal high school completion, internalizing problems, externalizing problems.

Further examination of the MBS costs incurred indicated that cost of specialist services (e.g., pediatricians, child psychiatrists) were significantly higher for children with ADHD at each wave (mean difference: Aus$97-Aus$380; p < .001), irrespective of the method used to define ADHD. Costs of general practitioner services were significantly higher at 6 to 7 (mean difference: Aus$104; p = .02) and 8 to 9 (mean difference: Aus$117; p = .005) years for children with ADHD, and at 8 to 9 years for children with clinical ADHD symptoms (mean difference: Aus$81; p = .001). Costs associated with allied health services were significantly higher at 8 to 9 years only, for both children with ADHD (mean difference: Aus$131; p = .005) and children with clinical ADHD symptoms (mean difference: Aus$85; p = .02). All other comparisons (except allied health costs for children with clinical ADHD symptoms at 4-5 years) showed larger costs for children with ADHD, but cost differences were not statistically significant.

PBS Cost Differences by ADHD Status

The 2-year PBS costs by ADHD status are outlined in Table 3. Mean-adjusted PBS costs were higher for children with ADHD at every age although this difference was not significant at 4 to 5 years. Children with clinical ADHD symptoms also had higher adjusted mean PBS costs at each age although this difference was not significant at 6 to 7 years. Similar to the pattern observed for MBS costs, PBS costs tended to increase with age for children with ADHD. The largest differences in PBS costs were observed at 8 to 9 years for children with ADHD (unadjusted: 1071% higher than for children without ADHD). Similarly, the largest differences in PBS costs were observed at 8 to 9 years for children with clinical ADHD symptoms (unadjusted: 321% higher than for children without clinical ADHD symptoms). The PBS costs for children without ADHD remained stable from 4 to 9 years.

Table 3.

Unadjusted and Adjusted 2-Year PBS Cost Estimates by ADHD Group.

		PBS cost (Aus$) per child by ADHD status^a
		M (SE)		Mean difference [95% CI]
Age	n (yes)	Yes	No	Unadjusted	p value	Adjusted^b	p value^b
Children with ADHD
4-5 years	52	102 (31)	37 (2)	65 [4, 125]	.04	87 [−12, 186]	.09
6-7 years	61	206 (57)	39 (4)	167 [54, 280]	.004	100 [41, 159]	.001
8-9 years	95	492 (67)	42 (5)	450 [317, 583]	<.001	432 [285, 578]	<.001
Children with clinical ADHD symptoms
4-5 years	262	70 (12)	35 (2)	35 [12, 59]	.004	28 [6, 50]	.01
6-7 years	238	83 (19)	38 (4)	45 [7, 83]	.02	25 [−7, 56]	.12
8-9 years	205	177 (29)	42 (5)	135 [76, 195]	<.001	88 [30, 145]	.003

Note. PBS = Pharmaceutical Benefit Schedule; CI = confidence interval.

Costs over 2 years, in 2012 Aus$, rounded to the nearest dollar.

Adjusted for child age in months and gender, single-parent family, mother’s main language spoken at home, maternal high school completion, internalizing problems, externalizing problems.

Further examination of the PBS costs revealed that the cost differences for children with ADHD were largely driven by cost differences associated with core ADHD medications and other centrally acting medications. Core ADHD medication costs were significantly higher at each age, for children with ADHD compared with children without ADHD, and for children with clinical ADHD symptoms compared with children without (all p < .05). Across waves, core ADHD medications comprised 16% (4-5 years), 47% (6-7 years), and 28% (8-9 years) of total PBS costs for children with ADHD, and 3% (4-5 years) and 20% (6-7 and 8-9 years) of total PBS costs for children with clinical ADHD symptoms. Costs for other centrally acting medications were significantly higher at 8 to 9 years for children with ADHD (p = .008), and at 6 to 7 (p = .04) and 8 to 9 years (p = .02) for children with clinical ADHD symptoms. Across waves, other centrally acting medications comprised 34% (4-5 years), 33% (6-7 years), and 59% (6-7 years) of total PBS costs for children with ADHD, and 12% (4-5 years), 39% (6-7 years), and 23% (8-9 years) of total PBS costs for children with clinical ADHD symptoms. Costs for other medications were significantly higher only at 4 to 5 years for children with clinical ADHD symptoms compared with those without clinical ADHD symptoms (p = .01); no other differences in costs for other medications were statistically significant.

Total Medicare and Population-Based Cost Differences Between Children With and Without ADHD

Table 4 shows the unadjusted and adjusted mean differences in total Medicare costs per child (MBS and PBS combined), and estimated excess population costs by ADHD status at each age. Per child and population-level costs were higher among children with ADHD at each age, both in the unadjusted analyses and after adjusting for demographic characteristics and child internalizing and externalizing problems. At the population level, total excess health care costs for children with ADHD were estimated to fall between Aus$18.1 and Aus$31.2 million dollars from 4 to 9 years of age.

Table 4.

Australian Population Estimates of 2-Year Total Medicare Costs Associated With ADHD.

	Total Medicare costs (Aus$) per child by ADHD status^a
	M (SE)		Mean difference [95% CI]
Age	Yes	No	Unadjusted	p value	Adjusted^b	p value^b	Estimated prevalence (%)	Population n^c	Estimated excess Medicare cost (Aus$ millions)^d M [95% CI]
Children with ADHD
4-5 years	873 (116)	495 (9)	378 [148, 608]	.001	486 [162, 809]	.003	1.3	5,960	2.3 [0.9, 3.6]
6-7 years	1,131 (111)	433 (10)	698 [478, 919]	<.001	594 [384, 803]	<.001	1.7	9,137	6.4 [4.4, 8.4]
8-9 years	1,585 (114)	469 (12)	1,116 [896, 1,343]	<.001	1,165 [926, 1,405]	<.001	2.6	14,320	16.0 [12.8, 19.2]
							Total excess costs		24.7 [18.1, 31.2]
Children with clinical ADHD symptoms
4-5 years	692 (44)	485 (9)	207 [118, 296]	<.001	212 [104, 321]	<.001	6.0	32,509	6.7 [3.8, 9.6]
6-7 years	632 (46)	431 (10)	201 [108, 294]	<.001	169 [73, 267]	.001	6.1	32,786	6.6 [3.5, 9.6]
8-9 years	938 (75)	464 (13)	473 [324, 623]	<.001	372 [235, 509]	<.001	6.4	35,248	16.7 [11.4, 22.0]
							Total excess costs		30.0 [18.7, 41.2]

Note. CI = confidence interval.

Costs over 2 years, in 2012 Aus$, rounded to the nearest dollar.

Adjusted for child age in months and gender, single-parent family, mother’s main language spoken at home, maternal high school completion, internalizing problems, externalizing problems.

Estimated population size of same-age Australian children meeting the ADHD case definition.

Estimated whole population cost over 2 years (calculated as unadjusted mean differences multiplied by population size), in millions and rounded to the nearest million.

Similar patterns emerged for children with clinical ADHD symptoms. Per child costs were higher among children with clinical ADHD symptoms at each age (Table 4), and these differences remained when accounting for demographic characteristics and child internalizing and externalizing problems. At the population level, total excess health care costs for children with clinical ADHD symptoms were estimated to fall between Aus$18.7 and Aus$41.2 million dollars from 4 to 9 years of age.

Costs associated with persistent clinical ADHD symptoms

We also examined the costs associated with persistent clinical ADHD symptoms for those who had parent-reported SDQ data available across the three waves (n = 3,604). Eighty-nine percent (n = 3,206) did not have clinical ADHD symptoms at any wave, 7% (n = 255) had clinical ADHD symptoms at one wave, and 4% (n = 141) had clinical ADHD symptoms at two or more waves.

Figure 1 shows that MBS (coefficient [mean increased costs]: Aus$390; 95% CI = [Aus$262, Aus$518]; p < .001), PBS (coefficient: Aus$151; 95% CI = [Aus$73, Aus$230]; p < .001), and total Medicare (coefficient: Aus$541; 95% CI = [Aus$364, Aus$719]; p < .001) costs increased significantly with the persistence of clinical ADHD symptoms over 6 years. This finding remained significant when adjusting for demographic characteristics and child internalizing and externalizing problems. Total Medicare costs for children with clinical ADHD symptoms at two or more waves were 79% higher than for children who did not have clinical ADHD symptoms at any wave, and 25% higher than for children who had clinical symptoms at one wave. Total Medicare costs for children with clinical ADHD symptoms at one wave were 44% higher than for those who did not have clinical ADHD symptoms at any wave.

Figure 1.

Six-year PBS and MBS costs by persistence of clinical ADHD symptoms across Waves 1, 2, and 3.

Discussion

This study examined the health care costs associated with ADHD in children aged 4 to 9 years, using two different definitions of ADHD and using a nationally representative Australian sample. We found that ADHD is associated with additional health care costs from 4 to 9 years of age. At a population level, these excess health care costs amounted to Aus$25 to Aus$30 million from 4 to 9 years, depending on the definition of ADHD used. Costs associated with ADHD were higher for older children with ADHD irrespective of the ADHD case definition used, and increased linearly with the persistence of ADHD symptoms over the middle childhood years.

We found that both MBS (health care attendance) and PBS (medication) costs were higher for children with ADHD and for children with clinical ADHD symptoms at each age examined. This finding is consistent with previous studies examining the health care costs associated with ADHD (Doshi et al., 2012; Pelham et al., 2007). We found that the higher PBS costs for children with ADHD were largely driven by the costs associated with core ADHD and other centrally acting medications. Higher MBS costs appeared to be particularly driven by costs associated with specialist services at all ages (e.g., pediatricians, child psychiatrists), with differentials in general practitioner and allied health costs emerging at 8 to 9 years. This may be in part due to the introduction of the Better Access to Mental Health Care Scheme that was introduced throughout Australia in 2006. This scheme provides Medicare rebates for children/adults, referred by a medical practitioner and meeting criteria for a DSM-IV disorder, for up to 10 sessions with a psychologist per year.

Excess population health care costs were similar despite the ADHD definition used. This was interesting given that the prevalence of clinical ADHD symptoms was much higher (6.0%-6.4%) than the prevalence of an actual ADHD diagnosis (1.3%-2.6%). It is quite likely that a proportion of the children presenting with clinical ADHD symptoms will be diagnosed with ADHD as they continue progress through primary school (Kessler et al., 2005). Similarities in excess population costs likely reflect that the children with a diagnosis of ADHD are a more severe group, given their early diagnosis of ADHD. Also given that the parents of children with an ADHD diagnosis are aware of their child’s difficulties, this is likely to drive health service and medication costs, which is consistent with our data, that is, children with ADHD incurred higher health care costs than the group with clinical ADHD symptoms. Some parents of children with clinical ADHD symptoms may not see their child’s behavior as a “problem” per se and may therefore not attend clinical services, and even if the parents do recognize a problem, there may be barriers to health service (e.g., socioeconomic factors or cultural; Bussing, Zima, Gary, & Garvan, 2003). It is also likely that the children presenting with clinical ADHD symptoms are a heterogeneous group. From our data, we are unable to determine whether ADHD symptoms are the primary presenting problem or secondary to other diagnoses, including learning, autism spectrum, and internalizing or externalizing disorders.

This study adds to the existing literature by highlighting that the increased health care costs associated with both ADHD and clinical ADHD symptoms emerge from an early age (i.e., 4-5 years) and continue to rise as children grow (at least up until their 10th birthday). For example, ADHD is associated with concurrent population costs in excess of Aus$2.3 million at age 4 to 5 years, and this increases to Aus$16.0 million for children with ADHD at 8 to 9 years. This pattern, of increased health care costs incurred with age, is likely to reflect both the increased prevalence of ADHD and increased use of pharmacological interventions as children grow older. In addition, increased costs associated with ADHD at 8 to 9 years, may also reflect the increasing impairments experienced by children with ADHD as they progress through primary school (Loe & Feldman, 2007). Many previous studies examining the health care costs associated with ADHD have not controlled for both demographic factors and internalizing and externalizing comorbidities, making it difficult to conclude whether increased health care costs are associated with ADHD per se, or other factors (Guevara et al., 2000). We found that differences in health care costs remained even after adjusting for sociodemographic factors, and internalizing and externalizing comorbidities.

To our knowledge, this is the first study to examine the relationship between health care costs and the persistence of ADHD symptoms. We found that children presenting with transient ADHD symptoms (i.e., children presenting with clinical ADHD symptoms on one occasion only) had significantly higher health care costs compared with children who never presented with clinical ADHD symptoms. We also found that health care costs (both MBS and PBS) increased with the persistence of clinical ADHD symptoms from the ages of 4 to 9 years. For example, the PBS costs for children presenting with clinical ADHD symptoms at two or more waves were nearly triple that of children who never presented with clinical ADHD symptoms. Similarly, MBS costs also increased as a function of ADHD persistence. These findings may provide some support for the early detection and treatment of childhood ADHD symptoms. However, it is important to note that the decreased persistence of ADHD symptoms for some children could be due to the effective management of ADHD symptoms via stimulant medication use or other psychological therapies, both of which would contribute to increased health care costs. Therefore, the costs reported in this study are likely to be an underrepresentation of the costs to be saved by the remission of ADHD symptoms.

We have demonstrated that ADHD is associated with significant health care costs within the Australian health care system. The costs identified in this study are likely to be only a fraction of the total health care and social costs associated with ADHD, given that Medicare costs do not reflect hospital or emergency department admissions, out-of-pocket expenses, the time costs incurred by caregivers in the use of health care services, or the impact on educational and other services. Children with ADHD are known to be at increased risk of severe injuries with associated need for medical attendance, and while this may account for some of the increased health care costs shown here, it suggests further cost differentials may be found in hospital-based health care (DiScala, Lescohier, Barthel, & Li, 1998). Higher use of Medicare services is likely to be associated with higher burden on family time and out-of-pocket expenses to receive these services. Furthermore, children with ADHD may also make greater use of health care contexts which are not Medicare rebateable, for example, some allied health services. We also know that over-the-counter and complementary medicines are commonly used for children with ADHD in Australia, and again these costs are not reflected in Medicare data (Srivastava & Efron, 2005). This study did not focus on costs incurred to other service systems, including the education or justice systems, as well as wider societal costs incurred as a result of ADHD, including school dropout, lack of productivity, and impact on families. Previous studies examining costs to systems aside from the health care system (e.g., education system) have demonstrated significant increased costs associated with ADHD (Marks et al., 2009).

The strengths of this study lie in the linkage to reliable, administrative data on health care utilization, as well as the population-based design of the study. Several limitations to this study should be noted. ADHD was defined by parent-reported diagnosis. However, when examining the health care costs for children with clinical ADHD symptoms, similar trends were observed. Furthermore, we have previously reported on the use of our ADHD case definition within the LSAC data set (Sciberras et al., 2011). It is important to recognize that we are unable to determine whether the increased health care costs reported in this article are directly related to ADHD, given that we do not have information about the reasons children were attending health care services or taking medications. We did, however, take into account a broad range of demographic factors, as well as internalizing and externalizing problems when examining health care costs at each age. Children with ADHD are likely to present with other comorbid conditions, such as learning/language disorders, autism spectrum disorders, and epilepsy, which were not taken into account in the present study (Gillberg et al., 2004; Mayes, Calhoun, & Crowell, 2000; Reiersen, Constantino, Volk, & Todd, 2007; Tirosh & Cohen, 1998). The excess costs observed in this study could be due to the assessment and/or treatment of ADHD specifically and/or due to the comorbidities associated with ADHD. Future research should examine how these other comorbidities influence the health care costs of children with ADHD.

This study demonstrates that ADHD is a nontrivial problem, associated with significant health care costs from early in life, with higher costs for older children with ADHD. The health care costs associated with ADHD comprise the costs of both health care attendances and medications. Of note, both transient and persistent ADHD symptoms are associated with increased health care costs. Understanding the costs associated with ADHD is an important first step in helping to plan for service system changes, including early detection and intervention approaches. We plan to continue to monitor and report on the health care costs associated with ADHD as children progress through late primary school and high school.

Footnotes

Acknowledgements

We thank all the families participating in the LSAC study. We acknowledge the peer review provided by the LSAC analysis group comprising staff from Murdoch Childrens Research Institute and Parenting Research Centre.

Authors’ Note

This article uses confidential unit record files from the Longitudinal Study of Australian Children (LSAC) survey. The LSAC project was initiated and funded by the Commonwealth Department of Families, Housing, Community Services, and Indigenous Affairs and was managed by the Australian Institute of Family Studies. The findings and views reported in this article are those of the authors and should not be attributed to either the Commonwealth Department of Families, Housing, Community Services, and Indigenous Affairs or the Australian Institute of Family Studies.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Drs Sciberras and Gold were funded by Australian National Health and Medical Research Council (NHMRC) Population Health Capacity Building Grant (436914 and 425855) and an NHMRC Early Career Research Fellowship (1037159 and 1035100) for the duration of this manuscript’s preparation. This research was supported by the Victorian Government’s Operational Infrastructure Support Program to Murdoch Childrens Research Institute (MCRI). The Parenting Research Centre receives funding from the Victorian Government Department of Education and Early Child Development.

Author Biographies

Emma Sciberras is a postdoctoral research fellow and clinical psychologist at the Murdoch Childrens Research Institute and The Royal Children’s Hospital, Melbourne, Australia.

Nina Lucas is a research officer at the Parenting Research Centre in Melbourne and the Centre for Ageing, Health, and Wellbeing in Canberra; visiting fellow at the National Centre for Epidemiology and Population Health, Canberra; and an honorary research fellow at the Murdoch Childrens Research Institute, Melbourne.

Daryl Efron is a developmental-behavioral pediatrician at The Royal Children’s Hospital, Melbourne, Australia; a senior lecturer in the Department of Pediatrics, at the University of Melbourne; and a senior research fellow at the Murdoch Childrens Research Institute.

Lisa Gold is a health economist at Deakin Health Economics, Deakin University, Australia, and an honorary research fellow at the Murdoch Childrens Research Institute, Melbourne.

Harriet Hiscock is a pediatrician and senior research fellow at the Murdoch Childrens Research Institute and The Royal Children’s Hospital, Melbourne, Australia. She has a keen interest in designing and evaluating interventions to improve outcomes for children with behavioral problems.

Jan M. Nicholson is the director of research for the Parenting Research Centre in Melbourne, an honorary principal research fellow at the Murdoch Childrens Research Institute in Melbourne, and an adjunct professor in the Centre for Learning Innovation at Queensland University of Technology.

References

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author.

Australian Bureau of Statistics. (2010). 3201.0-Population by age and sex, Australian States and Territories. Retrieved from http://www.abs.gov.au/ausstats/abs@.nsf/mf/3201.0

Australian Institute of Health and Welfare. (2012). Australia’s health 2012: 12th Biennial health report. Canberra: Australian Institute of Health and Welfare.

Barkley

R. A.

Fischer

Edelbrock

C. S.

Smallish

(1990). The adolescent outcome of hyperactive children diagnosed by research criteria: I. An 8-year prospective follow-up study. Journal of the American Academy of Child & Adolescent Psychiatry, 29, 546-557.

Biederman

Monuteaux

M. C.

Mick

Spencer

Wilens

T. E.

Silva

J. M.

. . .Faraone

S. V.

(2006). Young adult outcome of attention deficit hyperactivity disorder: A controlled 10-year follow-up study. Psychological Medicine, 36, 167-179.

Bussing

Zima

B. T.

Gary

F. A.

Garvan

C. W.

(2003). Barriers to detection, help-seeking, and service use for children with ADHD symptoms. Journal of Behavioral Health Services & Research, 30, 176-189.

Chan

Zhan

Homer

C. J.

(2002). Health care use and costs for children with attention-deficit/hyperactivity disorder: National estimates from the medical expenditure panel survey. Archives of Pediatrics & Adolescent Medicine, 156, 504-511.

Commonwealth Department of Health and Aged Care. (2000). The Australian health care system: An outline. Canberra, Australia: Australian Institute of Health and Welfare.

De Ridder

De Graeve

(2006). Healthcare use, social burden and costs of children with and without ADHD in Flanders, Belgium. Clinical Drug Investigation, 26, 75-90.

10.

DiScala

Lescohier

Barthel

(1998). Injuries to children with attention deficit hyperactivity disorder. Pediatrics, 102, 1415-1421.

11.

Doshi

J. A.

Hodgkins

Kahle

Sikirica

Cangelosi

M. J.

Setyawan

Neumann

P. J.

(2012). Economic impact of childhood and adult attention-deficit/hyperactivity disorder in the United States. Journal of the American Academy of Child & Adolescent Psychiatry, 51, 990-1002.

12.

Faraone

S. V.

Biederman

Mick

(2006). The age-dependent decline of attention deficit hyperactivity disorder: A meta-analysis of follow-up studies. Psychological Medicine, 36, 159-165.

13.

Gillberg

I. C.

Rasmussen

Kadesjö

Söderström

Råstam

Niklasson

(2004). Co-existing disorders in ADHD—Implications for diagnosis and intervention. European Child & Adolescent Psychiatry, 13(Suppl. 1), i80-i92.

14.

Goodman

(2001). Psychometric properties of the Strengths and Difficulties Questionnaire. Journal of the American Academy of Child & Adolescent Psychiatry, 40, 1337-1345.

15.

Gray

Sanson

(2005). Growing up in Australia: The Longitudinal Study of Australian Children. Retrieved from http://search.informit.com.au/documentSummary;dn=052395322391224;res=IELHSS

16.

Guevara

Lozano

Wickizer

Mell

Gephart

(2000). Utilization and cost of health care services for children with attention-deficit/hyperactivity disorder. Pediatrics, 108, 71-78.

17.

Hakkaart-van Roijen

Zwirs

B. W. C.

Bouwmans

Tan

S. S.

Schulpen

T. W. J.

Vlasveld

Buitelaar

J. K.

(2007). Societal costs and quality of life of children suffering from attention deficient hyperactivity disorder (ADHD). European Child & Adolescent Psychiatry, 16, 316-326.

18.

Hawes

D. J.

Dadds

M. R.

(2004). Australian data and psychometric properties of the Strengths and Difficulties Questionnaire. Australian and New Zealand Journal of Psychiatry, 38, 644-651.

19.

Hiscock

Roberts

Efron

Sewell

J. R.

Bryson

H. E.

Price

A. M.

Wake

M.A.

(2011). Children attending paediatricians study: A national prospective audit of outpatient practice from the Australian paediatric research network. Medical Journal of Australia, 194, 392-397.

20.

Kessler

R. C.

Berglund

Demler

Jin

Merikangas

K. R.

Walters

E. E.

(2005). Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the national comorbidity survey replication. Archives of General Psychiatry, 62, 593-768.

21.

Loe

I. M.

Feldman

H. M.

(2007). Academic and educational outcomes of children with ADHD. Journal of Pediatric Psychology, 32, 643-654.

22.

Lumley

Diehr

Emerson

Chen

(2002). The importance of the normality assumption in large public health data sets. Annual Review of Public Health, 23, 151-169.

23.

Marks

D. J.

Mlodnicka

Bernstein

Chacko

Rose

Halperin

J. M.

(2009). Profiles of service utilization and the resultant economic impact in preschoolers with attention deficit/hyperactivity disorder. Journal of Pediatric Psychology, 34, 681-689.

24.

Mayes

S. D.

Calhoun

S. L.

Crowell

E. W.

(2000). Learning disabilities and ADHD overlapping spectrum disorders. Journal of Learning Disabilities, 33, 417-424.

25.

Misson

Sipthorp

(2007). Wave 2 weighting and non-response (LSAC Technical Paper No. 5). Melbourne, Victoria, Australia: Australian Institute of Family Studies.

26.

Pelham

W. E.

Foster

E. M.

Robb

J. A.

(2007). The economic impact of attention-deficit/hyperactivity disorder in children and adolescents. Journal of Pediatric Psychology, 32, 711-727.

27.

Polanczyk

de Lima

M. S.

Horta

B. L.

Biederman

Rohde

L. A.

(2007). The worldwide prevalence of ADHD: A systematic review and metaregression analysis. American Journal of Psychiatry, 164, 942-948.

28.

Reiersen

A. M.

Constantino

J. N.

Volk

H. E.

Todd

R. D.

(2007). Autistic traits in a population-based ADHD twin sample. Journal of Child Psychology and Psychiatry, 48, 464-472.

29.

Schlander

Trott

G. E.

Schwarz

(2010). The health economics of attention deficit hyperactivity disorder in Germany. Part 1: Health care utilization and cost of illness. Nervenarzt, 81, 289-300.

30.

Sciberras

Ukoumunne

O. C.

Efron

(2011). Predictors of parent-reported attention-deficit/hyperactivity disorder in children aged 6-7 years: A national longitudinal study. Journal of Abnormal Child Psychology, 39, 1025-1034.

31.

Soloff

Lawrence

Johnstone

(2005). Sample design (LSAC Technical Paper No. 1). Melbourne, Victoria, Australia: Australian Institute of Family Studies.

32.

Soloff

Lawrence

Misson

Johnstone

(2006). Wave 1 weighting and non-response (LSAC Technical Paper No. 3). Melbourne, Victoria, Australia: Australian Institute of Family Studies.

33.

Srivastava

Efron

(2005). Complementary and alternative medicine use in children with attention deficit hyperactivity disorder. Journal of Paediatrics and Child Health, 41, 23-26.

34.

Tirosh

Cohen

(1998). Language deficit with attention-deficit disorder: A prevalent comorbidity. Journal of Child Neurology, 13, 493-497.

35.

Willis

Reynolds

(2008). Understanding the Australian health care system. Chatswood, New South Wales, Australia: Elsevier.

Health Care Costs Associated With Parent-Reported ADHD: A Longitudinal Australian Population–Based Study

Abstract

Keywords

Method

Study Design and Participants

Measures

ADHD status

Internalizing and externalizing problems

Health care costs

Demographic characteristics

Analyses

Results

Sample Characteristics

MBS Cost Differences by ADHD Status

PBS Cost Differences by ADHD Status

Total Medicare and Population-Based Cost Differences Between Children With and Without ADHD

Costs associated with persistent clinical ADHD symptoms

Discussion

Footnotes

Acknowledgements

Authors’ Note

Declaration of Conflicting Interests

Funding

Author Biographies

References