Background. Despite an increasing prevalence of diagnosed attention-deficit/hyperactivity disorder (ADHD) among children, the impact of ADHD on utilization and costs of health care services is largely unknown.
Objective. To examine differences in health care utilization and costs between children with and without ADHD.
Design. Retrospective matched cohort study conducted from January 1 to December 31, 1997.
Setting. Health maintenance organization in western Washington State.
Participants. Children aged 3 through 17 years who were continuously enrolled in the health maintenance organization and used services during 1997 were eligible. Children were identified with ADHD if they had a diagnosis of ADHD or a prescription for a stimulant medication using automated patient files. Children without ADHD were randomly selected as controls and matched 4:1 to children with ADHD on age and sex.
Outcome Measure. Utilization and costs of specific categories of health care services.
Results. A total of 2992 children (5.2%) were identified with ADHD. Children with ADHD incurred significantly greater per capita total costs ($1465 vs $690) than children without ADHD. Children with ADHD had 9.9 times more outpatient mental health visits (1.35/year vs 0.14/year), 3.4 times more pharmacy fills (11.25/year vs 3.30/year), and 1.6 times more primary care visits (3.84/year vs 2.36/year) than children without ADHD. The adjusted incremental costs were estimated to be $375 (95% confidence interval: $336–$416) for children with ADHD alone and $812 (95% confidence interval: $671–$973) for children with ADHD plus coexisting mental health disorders.
Conclusions. Children with ADHD use significantly more health care resources and incur significantly higher costs than children without ADHD. Coexisting mental health disorders substantially increase the cost of treating ADHD. Resource allocation decisions should consider the contributions of primary care, outpatient mental health, and pharmacy costs to the overall costs of care for children with ADHD.
Attention-deficit/hyperactivity disorder (ADHD) is the most common mental health disorder in children1 and is diagnosed in 3% to 6% of school-aged children in the United States2,,3 ADHD is a chronic disorder, with a majority of children continuing to exhibit symptoms into adolescence and early adulthood.4–6 The impact of ADHD on children is considerable. Children with ADHD are more likely to experience academic failure, delinquency, strained peer and family relations, and poor self-esteem.2,,7,8 Children with ADHD are also diagnosed with other mental health disorders, which may contribute to their difficulties.9,,10
Although ADHD may be increasing in prevalence,11 the impact of this disorder on health care utilization and costs is unknown. Although most children with ADHD are prescribed stimulant medications,12,,13 little is known about the frequency of health care visits they make. Evidence suggests that children with ADHD underuse mental health services for behavioral treatments14 and overuse emergency department (ED) services for accidental injuries and poisonings.15 Trends in service use indicate that the proportion of physician visits attributable to ADHD has doubled, and the percentage of these visits involving a stimulant prescription has risen dramatically.16,,17 It is not known how other mental health disorders in association with ADHD affect health care utilization and costs. A National Institutes of Health Consensus Panel recently recommended that future research on the subject of ADHD address the “financial costs related to the diagnosis and care of children with ADHD.”18
Our objective was to determine differences in per capita utilization and costs of health care services between children with and without ADHD from a health care system perspective. In addition, we sought to estimate the incremental costs of care for children with ADHD. We postulated that children with ADHD would have greater utilization and incur higher costs for primary care, outpatient mental health, pharmacy, and ED care. Knowledge from this study can be used to guide decisions about the allocation of health care resources and assist in planning interventions to reduce inappropriate utilization for children and adolescents with ADHD.
Group Health Cooperative of Puget Sound (GHC) is a large not-for-profit staff model health maintenance organization (HMO) located in western Washington State. It serves approximately 500 000 enrollees and is considered the country's largest consumer-governed health care organization. Primary care for children is provided by family practitioners (50%) and by pediatricians (50%), >95% of whom are board certified. Outpatient mental health care is provided at 6 geographically dispersed clinics. Treatment emphasizes brief psychotherapy, group therapy, and pharmacotherapy. GHC patients are permitted to self-refer directly for specialty mental health care. Inpatient, partial hospitalization, and day treatment are provided by contracted providers throughout the GHC service area. The Center for Attention Deficit Disorders is a referral clinic that conducts initial evaluations, provides follow-up services, and sponsors various educational activities and behavioral interventions. At the time of this study, no systemwide quality improvement initiatives or clinical guidelines were in place.
Data on patient demographics, hospital length of stay, all diagnostic and procedural codes for ambulatory and inpatient care, pharmacy fills, ambulatory visits, laboratory and radiology studies, and health care costs were obtained from GHC's computerized patient information system. In this system, information contained in multiple databases is linked through a unique identifier to individual enrollees. Verification of computerized data occurs through online edits, automatic prompts for incomplete or missing information, and monthly comparison of data from all data input sources. Data on the costs to GHC of individual health care services provided to enrollees were available through the Decision Information Support Center (DISC). DISC uses a cost accounting system to estimate the true cost of health care services at the unit-of-service level. DISC allocates overhead costs from the general ledger to individual clinical departments. A cost per unit of service is estimated that reflects costs for medical personnel, supplies, and overhead. Contractual services to external providers were included based on GHC's payment to these providers. Medical costs to enrollees in the form of copayments or out-of-pocket expenses were not recorded. DISC updates all health care service costs based on monthly utilization data. Additional details of DISC are provided elsewhere.19
All children ages 3 through 17 years who were continuously enrolled in the HMO from January 1, 1997, to December 31, 1997 and who made at least 1 ambulatory visit or hospitalization during this period were eligible for inclusion. Eligibility was restricted to users of services to remove the potential bias of a greater proportion of controls with nonutilization than cases and to allow comparisons to reflect differences in per capita rates of use.
The cohort was divided into children with and without ADHD. Children with ADHD were defined as all children who made at least 1 ambulatory visit or had a hospitalization during the study period that included a diagnosis of ADHD by International Classification of Diseases, Ninth Revision (ICD-9) code 314 or who had a pharmacy fill for a stimulant medication. Providers at GHC were instructed to code only diagnoses for which treatment was rendered. The following stimulant drugs were categorized as ADHD medications because of their negligible use in other conditions3: methylphenidate, dextroamphetamine, mixed amphetamine salts, and pemoline. Children without ADHD were randomly selected as controls and matched on age and sex to children with ADHD by a 4:1 ratio. We chose to match on age and sex to control for the effect of these covariates on unadjusted utilization and cost differences and to improve the precision of our adjusted cost estimates. Children with a diagnosis of cancer (ICD-9 codes 140–209) were excluded because they were regarded as high users of care in this HMO.19
Children in the cohort were identified with additional mental health disorders from automated files. The following mental health disorders were selected using ICD-9 codes: depression (ICD-9 codes 296.2, 296.3, 300.4, and 311), anxiety (ICD-9 codes 300.0–300.2 and 313.0), bipolar disorder (ICD-9 code 296), obsessive-compulsive disorder (ICD-9 code 300.3), oppositional-defiant disorder (ICD-9 code 313.81), conduct disorder (ICD-9 code 312), learning disorders (ICD-9 codes 315.0–315.3), substance abuse and dependence disorders (ICD-9 codes 303–305), and tic disorders (ICD-9 code 307.2). Children were classified for the purposes of analysis as having any coexisting mental health disorders (1 or more of the 9 disorders), internalizing conditions (depression, anxiety, or obsessive-compulsive disorders), externalizing conditions (oppositional-defiant or conduct disorders), or substance abuse or dependence disorders. The study was approved by the Human Subjects and Research Committees at GHC and the University of Washington.
The outcomes of interest were utilization and costs of total and specific categories of health care services. Patient encounters were classified into 6 categories of service: primary care visits, ED visits, outpatient mental health visits, pharmacy fills, hospitalizations, and other visits. Primary care visits were defined as all visits to family practice, pediatrics, or urgent care clinics. ED visits were defined as visits to an ED. Outpatient mental health visits were defined as visits to outpatient mental health clinics. Pharmacy fills were defined as all GHC medication fills, both initial and refill along with any over-the-counter medications included in the pharmacy benefit. Hospitalizations were defined as inpatient admissions at any GHC-affiliated institution. Other visits encompassed all other patient encounters including radiology, laboratory, and specialty care. Health care encounters were reported as per capita visits, pharmacy fills, and inpatient days per year. Health care costs were reported as per capita costs in 1997 US dollars.
Differences in utilization and costs between children with and without ADHD were assessed for statistical significance using the Wilcoxon rank sum test, which is commonly applied to data that exhibit skewness. Differences in the proportion of children with and without ADHD who had utilization or incurred costs were assessed for statistical significance by the chi-square test. A Pvalue < .05 was considered statistically significant. Utilization and cost ratios were calculated for specific categories of health care services and represent the relative utilization (or cost) for children with ADHD compared with children without ADHD.
Incremental costs represent differences in costs between children with and without ADHD that can be attributed to the diagnosis of ADHD. A single model was used to estimate total adjusted incremental costs, and 2-step models were used to estimate adjusted incremental costs for each category of health care service.20 In the first step, conditional logistic regression was used to estimate the probability of any costs in each category of service. The probability of any costs for children with ADHD relative to children without ADHD was adjusted explicitly for coexisting mental health disorders and other comorbidities and implicitly for age and sex. In the second step, generalized least squares was used to estimate incremental costs between children with and without ADHD for total costs and for each category of service only among those with costs in a particular category.21 All dependent cost variables were log transformed to reduce skewness. Incremental costs were adjusted for age, sex, coexisting mental health disorders, and other comorbid conditions in all models. To account for a priori matching, the matching variable designating age and sex was incorporated as a random-effects variable. Covariates were interpreted according to standard econometric convention to obtain cost estimates.22 To control for comorbidity, the Pediatric Chronic Disease Score (PCDS), a measure of chronic disease based on automated pharmacy data, was incorporated into the models as a categorical covariate.23,,24 PCDS classifies children into 1 or more of 42 disease categories based on specific prescription medication usage. We removed the PCDS classification for ADHD because it was already explicitly included in the model.
A sensitivity analysis of 70 randomly selected children with ADHD was performed to assess the validity of ADHD diagnoses and its impact on cost estimates. Charts were randomly selected from the entire cohort of children with ADHD using computer-generated random numbers. Individual patient encounters and behavioral rating scales included in the outpatient record were compared with diagnostic criteria for ADHD published in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV).25Criteria for ADHD were met if an individual encounter or an associated rating scale met 6 of 9 criteria for inattentive ADHD, hyperactive or impulsive ADHD, or combined ADHD. Functional impairment and symptoms across settings were not assessed in the outpatient record. Individual diagnoses were categorized as probable (chart encounter listed diagnosis of ADHD or treatment with stimulant medication, and criteria for ADHD were met), possible (chart encounter listed diagnosis of ADHD or treatment with stimulant medication, but criteria for ADHD were not met), or doubtful (chart encounter did not list diagnosis of ADHD or treatment with stimulant medication, and criteria for ADHD were not met). Differences in costs of total and specific categories of service between children with probable ADHD and children with either possible or doubtful ADHD were assessed for statistical significance using the Wilcoxon rank sum test. All statistical analyses were performed using Stata Statistical Software, Release 6.0 (College Station, TX).
We identified 57 216 children between ages 3 and 17 years who were continuously enrolled in the HMO during 1997 and used services. Of these, 2992 children (5.2%) were identified with ADHD, and 11 968 children (20.9%) without ADHD were selected as controls to yield a study cohort of 14 960. Children with ADHD in this HMO were predominantly male and school-age. The mean age was 11.7 years (standard deviation: 3.2 years), and the majority (52.9%) were between 7 and 12 years old (Table 1). Boys outnumbered girls by more than a 3:1 ratio. This ratio did not vary significantly among the different age groups. More than one fourth of children with ADHD were diagnosed with a coexisting mental health disorder (28.7%), whereas few children without ADHD were diagnosed with a mental health disorder (3.8%). Externalizing conditions, internalizing conditions, and substance abuse or dependence disorders accounted for the majority of comorbid mental health disorders; in each instance, these conditions were significantly more prevalent in children with ADHD (P < .001) than children without ADHD. Thirty-five percent of children with ADHD and 31% of children without ADHD were identified with other medical comorbidities by PCDS classification.
Children with ADHD were more likely to have utilization and incur costs in most categories of health care service than children without ADHD (Table 2). Children with ADHD were 1.5 times more likely to use primary care (adjusted odds ratio [OR] 1.54; 95% confidence interval [CI], 1.23–1.91), 9.0 times more likely to use outpatient mental health (adjusted OR 9.05; 95% CI, 7.53–10.86), 8.8 times more likely to use pharmacy services (adjusted OR 8.75; 95% CI, 7.27–10.53), and 2.1 times more likely to be hospitalized (adjusted OR 2.08; 95% CI, 1.78–2.42) than children without ADHD. The probability of ED use did not differ between children with and without ADHD.
Children with ADHD used more health care services per capita than children without ADHD for most categories of service (Table 3). Children with ADHD made 9.6 times as many outpatient mental health visits (1.35 visits/year vs 0.14 visits/year), filled 3.4 times as many medications (11.25 fills/year vs 3.30 fills/year), and made 1.6 times as many primary care visits (3.84 visits/year vs 2.36 visits/year) as children without ADHD. Children with ADHD made 1.6 times as many ED visits (0.08 visits/year vs 0.05 visits/year) as children without ADHD; however the ED utilization in both groups was low, and the difference was clinically insignificant. These utilization differences were all statistically significant (P < .001). There were no significant differences in hospital inpatient days between the 2 groups (P = .23).
Most hospitalizations (81.5%) were for short stays of <24 hours. These short hospitalizations represented admissions to observational units or same-day surgery units. Few children in either group were admitted to the hospital for stays of >24 hours: 45 children (1.50%) with ADHD and 147 children (1.26%) without ADHD. Still fewer had multiple hospital admissions: 3 children (6.7% of those hospitalized) with ADHD and 18 children (12.2% of those hospitalized) without ADHD. Children with ADHD had a higher proportion with psychiatric admissions (.6% vs .08%; P < .001), but mean length of stay was not different in the 2 groups among those who were hospitalized (5.9 days/year vs 6.7 days/year; P = .75). Children with ADHD also had a higher proportion of admissions for drug and alcohol dependence (0.7% vs 0.07%; P < .001), but mean length of stay among those hospitalized was not different (14.4 days/year vs 15.9 days/year; P = .84).
Children with ADHD incurred higher costs per capita for most categories of service than children without ADHD (Table 4). Overall, children with ADHD incurred 2.1 times the total costs ($1465 vs $690) of children without ADHD. For specific categories of service, children with ADHD had 1.7 times the primary care costs ($427 vs $245), 11.0 times the outpatient mental health costs ($222 vs $20), and 5.1 times the pharmacy costs ($335 vs $66) of children without ADHD. In addition, children with ADHD had ED costs 1.3 times as high ($38 vs $29) and 8% lower hospital inpatient costs ($115 vs $125), but these differences were small and clinically insignificant. All cost differences between the 2 groups were statistically significant (P < .001).
When costs were examined only among users in specific categories of services, children with ADHD generally had higher mean costs. Children with ADHD had 1.7 times as high mean primary care costs ($444 vs $261;P < .001), 1.4 times as high mean outpatient mental health costs ($669 vs $494; P < .001), and 3.7 times as high pharmacy costs ($351 vs $94; P < .001) as children without ADHD. There were no differences in mean ED costs between users in the 2 groups ($462 vs $493; P = .47). Hospitalized children with ADHD had 60% lower mean inpatient costs ($907 vs $2254; P < .001) than hospitalized children without ADHD, however these costs included short hospital stays of <24 hours.
Among children with ADHD, those with comorbid mental health disorders were a higher-cost subset. Figure 1 shows 2 groups of children (those with ADHD alone and those with ADHD and coexisting mental health disorders) divided into deciles according to total per capita cost. The maximum cost for each decile is shown. More than 70% percent of children with ADHD and comorbid mental health disorders had total costs above the median for all children with ADHD (eg, the 3rd decile for that subset exceeds $884, the overall group median), whereas only 30% of children with ADHD alone had total costs above the median for all children with ADHD.
The distribution of health care costs differed between children with and without ADHD (Fig 2). The majority of total costs for children with ADHD (67%) were attributable to primary care, outpatient mental health, and pharmacy costs. In all 3 of these service categories, costs were significantly higher for children with ADHD than for children without ADHD. The proportion of total costs for children with ADHD that was accounted for by ED and hospital inpatient care represented a fairly small portion of the total costs. The majority of total costs for children without ADHD (53%) were attributable to primary care and inpatient costs. The proportion of total costs attributable to inpatient care was substantially lower among children with ADHD than among children without ADHD.
The adjusted incremental costs for children with ADHD controlling for age, sex, and other comorbidities differed depending on the presence of coexisting mental health disorders for most categories of service (Table 5). The adjusted incremental costs for total health care services were estimated to be $375 (95% CI: $336–$416) for children with ADHD alone and $812 (95% CI: $671–$973) for children with ADHD who had other coexisting mental health disorders. The model explained 24% of the variation in total health care costs. For specific categories of service, the adjusted incremental costs for children with ADHD were estimated to be significantly higher for primary care, outpatient mental health, and pharmacy services. In each of these 3 categories of service, costs estimates were significantly higher among children with ADHD who had other coexisting mental health disorders than among children with ADHD alone. The adjusted incremental costs of ED services and inpatient services were not significant except that children with ADHD alone had lower inpatient costs than children without ADHD.
Categories of mental health disorders were examined for their impact on overall incremental cost estimates for children with ADHD (Table 6). Children with ADHD who had internalizing conditions had substantially higher total incremental costs ($1012 vs $898 or $847) than children with ADHD who had either externalizing conditions or substance abuse and dependence disorders. In addition, they incurred higher incremental costs for outpatient mental health ($301 vs $197 or $148) and pharmacy services ($150 vs $125 or $122). Differences in the incremental costs for primary care were not substantially different among children with ADHD who had any one of the 3 categories of mental health disorders.
Seventy randomly selected charts of children with ADHD were abstracted to determine the impact of diagnostic validity on costs (Table 7). Fifty (71.4%) of these children were determined to have a probable diagnosis of ADHD, 15 (21.4%) had a possible diagnosis of ADHD, and 5 (7.2%) had a doubtful diagnosis of ADHD. It should be noted that 8 of the children with possible ADHD were evaluated and diagnosed by mental health clinics or the Center for Attention Deficit Disorders at GHC. Information pertaining to diagnostic evaluations from these clinics was stored in confidential convenience files at the respective clinics and was not routinely available in the outpatient medical record for abstraction. The remainder of the children underwent diagnostic evaluations through their primary care providers. Costs for children with probable ADHD were not statistically significantly different from those for children with either possible or doubtful ADHD for total and individual categories of service (Table 7).
To our knowledge, this is the first population-based analysis of the impact of ADHD on the utilization and costs of health care services for children. Children with ADHD had higher rates of utilization and costs for most categories of health care services. In addition, children with ADHD incurred total costs that were >2 times as high as those for children without ADHD ($1465 vs $690). These incremental costs represent >$2.3 million for the nearly 3000 children with ADHD in the HMO. Most of the total incremental costs were attributable to differences in primary care, outpatient mental health, and pharmacy costs. ED and inpatient hospital costs were low and had little impact on incremental costs. The adjusted incremental cost estimates for children with ADHD depended on the presence of other mental health disorders. Children with coexisting mental health disorders, particularly internalizing conditions, had significantly higher incremental costs than children with ADHD alone.
Previous research has suggested that children with ADHD have higher urgent care utilization than children without ADHD. Szatmari et al15 in a study from Ontario, Canada, reported that children with ADHD had higher urgent care utilization, mainly for accidental injuries and poisonings, than children without ADHD. DiScala et al26 found that children with ADHD sustained more severe injuries than other children. This was mainly the result of unintentional pedestrian, bicycle, motor vehicle, and fall injuries. We found that children with ADHD in this study did not have higher ED utilization than children without ADHD after adjustment for other comorbidities. In addition, the adjusted incremental costs for ED services were not significant. However, because urgent care visits were subsumed under primary care visits in the dataset, we could not disentangle urgent care visits from overall primary care utilization. In this regard, we may have underestimated the true difference in urgent care visits between children with and without ADHD by examining ED utilization alone. Future studies that examine specific urgent care utilization using E-codes between children with and without ADHD would help clarify injury risk.
Previous research has also shown that children with ADHD have higher rates of mental health utilization than children without ADHD.15 However, it has been suggested that children with ADHD may still underuse mental health services for behavioral problems,14 but it is unclear what proportion of children with ADHD should optimally use mental health treatments. We found that mental health utilization depended on the diagnosis of coexisting mental health disorders. Children with ADHD alone had higher adjusted incremental costs for mental health care than children without ADHD, confirming previous studies. However, children with ADHD and diagnosed coexisting mental health disorders had higher adjusted incremental costs for mental health care than children with ADHD alone. This suggests that the addition of each mental health disorder substantially drives utilization in mental health care.
The data from this study suggest that ADHD may have a unique pattern of utilization and costs. Asthma, another common chronic disorder that is similar in prevalence to ADHD, has been regarded as a model for chronic disease in children. Although total incremental costs for the 2 conditions are similar at this HMO, the pattern of utilization and costs of health care services is different.19 In both conditions, affected children incurred about twice the health care costs as unaffected children. However, inpatient costs represented 26% of total costs for children with asthma but only 8% of total costs for children with ADHD. Because of differences in study design, urgent care and mental health utilization could not be compared.
Our study has particular strengths that merit attention. First, we used a population-based approach to identify all treated children with ADHD, both mild and severe cases. Studies that include only referral-based populations or primary care clinic populations may miss the entire spectrum of ADHD and report biased utilization and cost estimates. Second, we examined utilization and costs in individual categories of health care service. This allowed assessment of the contribution of individual categories of service to the total distribution of costs. Third, our study reflected the true costs of health care services at the HMO rather than charges. GHC's database contains costs of service that reflect actual resource consumption. Charges may not be an adequate proxy for costs for a number of reasons. Charges cover costs of capital expenses, indigent and unreimbursed care, and community service, and are adjusted based on community norms and negotiated contracts.27 Fourth, our utilization and cost estimates captured almost all service use, both within the HMO and from external providers to whom GHC provided payment. Services that were paid for out of pocket or by third-party (non-GHC) reimbursement were not included in the analysis.
There are limitations to our study. First, the use of automated patient files may result in misclassification of participants. Chart reviews suggested probable or possible ADHD in 93% of children identified with ADHD through automated data, and a sensitivity analysis suggested that cost estimates were not significantly different between those with a more definite diagnosis of ADHD and those with a less definite diagnosis of ADHD. Any misclassification probably would bias cost estimates toward the null. Second, the demographics of our study population at GHC may not be representative of children with ADHD in all geographic areas. The GHC population has been reported as predominantly white and middle income, similar to the population of Puget Sound but with a slightly higher educational level.28 It is unclear what effect this may have on patterns of utilization and associated costs. Third, the utilization of health care services at this HMO may not be typical of patterns of utilization in other geographic locations or in other local health plans. GHC is a primary care HMO with a reputation for high-quality health care. It is possible that differences in utilization and costs may be different at other health plans. Fourth, GHC cost data do not capture out-of-pocket expenses, uncovered treatment costs, indirect costs of care to families, and educational costs. These nonmedical expenses for ADHD care may be substantial. Fifth, we do not have cost data for ADHD by subtype (inattentive versus hyperactive and impulsive versus combined). It is possible that incremental costs differ widely between subtypes as a result of differences in medication treatments and mental health utilization. Sixth, we do not have cost data related to specific prescription medications. It is unclear to what extent differences in pharmacy costs between children with and without ADHD reflect differences in costs of specific psychotropic medications. Finally, state regulations governing controlled substances may artificially inflate pharmacy utilization; however, pharmacy costs should not be affected substantially by these regulations.
This study has important implications for children with ADHD. First, resource allocation decisions should consider the large contribution of primary care, outpatient mental health, and pharmacy costs to the overall costs of health care services for children with ADHD. In each of these service categories, children with ADHD had substantially greater utilization and costs than children without ADHD. Second, children with ADHD may not have higher ED utilization than children without ADHD. Interventions targeted to children with ADHD to reduce urgent care use may not be warranted. Third, coexisting mental health disorders substantially increase the cost of treating ADHD. Populations of children that have higher proportions with coexisting mental health disorders, particularly internalizing conditions, may incur greater overall costs for ADHD than those reported in this study. It is unknown whether the timely provision of mental health services will result in a reduction in long-term costs of care for children with ADHD. Future research on the potential cost offset of providing timely mental health services to children with ADHD is needed.
Presented in part at the Ambulatory Pediatrics Association National Meeting, May 1, 1999, San Francisco, California. This project was supported by funds from the Center for Health Studies at Group Health Cooperative of Puget Sound and by a generous contribution from Robert Davis, MD, MPH.
We gratefully acknowledge Frederick Rivara, MD, MPH; Robert Davis, MD, MPH; and Diane Martin, PhD for their critical reviews of the manuscript.
- Received July 13, 2000.
- Accepted December 19, 2000.
Reprint requests to (J.G.) Division of General Pediatrics, Children's Hospital of Philadelphia, 34th and Civic Center Blvd, Philadelphia, PA 19104. E-mail:
- ADHD =
- attention-deficit/hyperactivity disorder •
- ED =
- emergency department •
- GHC =
- Group Health Cooperative of Puget Sound •
- HMO =
- health maintenance organization •
- DISC =
- Decision Information Support Center •
- ICD-9 =
- International Classification of Diseases, Ninth Revision •
- PCDS =
- Pediatric Chronic Disease Score •
- DSM-IV =
- Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition •
- OR =
- odds ratio •
- CI =
- 95% confidence interval
- Reiff MI,
- Banez GA,
- Culbert TP
- Diller L
- Wolraich ML,
- Lindren S,
- Stromquist A,
- Milich R,
- Davis C,
- Watson D
- ↵Diagnosis and treatment of attention deficit hyperactivity disorder (ADHD). NIH Consens Statement. 1998;Nov 16–18:1–37
- Lozano P,
- Fishman P,
- VonKorff M,
- Hecht J
- ↵Stata Corporation. Stata Statistical Software. Release 6.0.College Station, TX: Stata Corporation; 1999
- ↵American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th ed. Washington, DC: American Psychiatric Association; 1994
- DiScala C,
- Lescohier I,
- Barthel M,
- Li G
- Copyright © 2001 American Academy of Pediatrics