Context A shortage of data exists on medical care use by persons with attention-deficit/hyperactivity
disorder (ADHD).
Objective To compare medical care use and costs among persons with and without
ADHD.
Design and Setting Population-based cohort study conducted in Rochester, Minn.
Subjects All children born in 1976-1982 were followed up through 1995, using
school and medical records to identify those with ADHD. The 4880 birth cohort
members (mean age, 7.3 years) still residing in Rochester in 1987 were followed
up in medical facility–linked billing databases until death, emigration,
or December 31, 1995.
Main Outcome Measures Clinical diagnoses, likelihood and frequency of inpatient and outpatient
hospitalizations, emergency department (ED) visits, and total medical costs
(including ambulatory care), compared among individuals with and without ADHD.
Results Among the 4119 birth cohort members who remained in the area through
1995 (mean age, 15.3 years), 7.5% (n = 309) had met criteria for ADHD. Compared
with persons without ADHD, those with ADHD were more likely to have diagnoses
in multiple categories, including major injuries (59% vs 49%; P<.001) and asthma (22% vs 13%; P<.001).
The proportion with any hospital inpatient, hospital outpatient, or ED admission
was higher for persons with ADHD vs those without ADHD (26% vs 18% [P<.001], 41% vs 33% [P = .006],
and 81% vs 74% [P = .005], respectively). The 9-year
median costs for persons with ADHD compared with those without ADHD were more
than double ($4306 vs $1944; P<.001), even for
the subset with no hospital or ED admissions (eg, median 1987 costs, $128
vs $65; P<.001). The differences between individuals
with and without ADHD were similar for males and females and across all age
groups.
Conclusion In our cohort, compared with persons without ADHD, those with ADHD exhibited
substantially greater use of medical care in multiple care delivery settings.
Attention-deficit/hyperactivity disorder (ADHD) is a relatively common
behavioral disorder of childhood, with important consequences for affected
individuals, their families, and society.1,2
The financial burden of ADHD, however, has not been well described.1 Individuals with ADHD have been shown to exhibit increased
use of mental health, social, and special education services,3,4
but there is a paucity of information on the use and costs of medical care.
The present study took advantage of the previous application of standardized
research criteria for ADHD among members of a large population-based birth
cohort.5,6 Members were followed
up in a medical facility–linked billing data system over a 9-year period
from a minimum of age 5 years to a maximum of age 19 years. Members with and
without ADHD were compared for comorbid clinical diagnoses, the likelihood
and frequency of emergency department (ED) visits, inpatient and outpatient
hospitalizations, and total medical (including ambulatory care) costs.
All children born January 1, 1976, through December 31, 1982, to mothers
residing in Rochester, Minn, townships that comprise Independent School District
535 (N = 8548) were previously identified using Minnesota Department of Health
computerized birth certificate information.6
Each birth cohort member was retrospectively followed up using his/her medical
and school records for evidence of ADHD until the earliest event of emigration,
death, or December 31, 1995.5 Because formal
diagnostic criteria for ADHD are typically met after children enter the school
system, the review was limited to the 5718 cohort members who had not died
or emigrated before age 5 years. The identification of ADHD among birth cohort
members was facilitated because Rochester is geographically isolated and home
to the Mayo Clinic, one of the world's largest referral centers. Therefore,
local residents receive medical services from a limited number of providers,
primarily Mayo Clinic and Olmsted Medical Center, another group practice,
and their affiliated hospitals. Since 1907, information from every Mayo encounter
has been contained within a patient-based medical record. Under the auspices
of the Rochester Epidemiology Project,7 access
to the medical records was expanded to include other medical facilities in
and around Rochester, including Olmsted Medical Center and the area's few
private medical practitioners. The medical records contain complete (hospital
inpatient, hospital outpatient, ED, physician's office, and specialty clinic
encounters) and detailed information (all clinicians' notes; psychiatric and
neurological examinations, surveys, and questionnaires; all other diagnostic
and laboratory results; and correspondence) from essentially all providers
of care to local residents.
Permission was also obtained to access the resources of Independent
School District 535, including the complete school records of all birth cohort
members ever registered at any of the district's public, parochial, or private
schools, plus those cohort members who were home schooled. The school records
include medical reports, medication records, private tutoring or evaluation
reports, individual and group-administered ability and achievement tests,
and notations from teacher, parent, or other person related to any type of
school performance difficulty. The district uses a standardized protocol for
referrals for any type of difficulty in school performance, learning, or other
potentially handicapping condition. A referral form is filled out and depending
on the type and extent of the difficulty, an individual educational assessment
report is filed, a meeting is held that includes both parents and teachers,
and an individual education program is developed. Copies of all reports, meeting
minutes, assessments, and periodic reassessments and reviews are maintained
in the record and were included in the review for ADHD.
The medical and school records were reviewed by a trained abstractor,
under the supervision of a developmental behavioral pediatrician (W.J.B.).
The records were reviewed for (1) clinical diagnoses of ADHD or ADHD-like
conditions, (2) parent or teacher questionnaires that assess ADHD symptoms,
with scores that were at least 1 SD above the mean (t
score ≥60) defined as positive, and (3) diagnostic criteria for ADHD as
defined by the Diagnostic and Statistical Manual of Mental
Disorders, Fourth Edition (DSM-IV).8
Review for the latter was facilitated with an extensive dictionary of words/terms
consistent with ADHD symptoms as specified in DSM-IV
that was developed and pilot tested as part of the ADHD incidence study.5 Data on date, setting (home or school), and informant
(teacher, parent, physician, or psychologist) were also abstracted.
Information obtained from the medical and school records was used to
categorize each of the 5718 individuals as either having (1) definite ADHD,
which consisted of a clinical diagnosis plus supporting documentation (ie,
positive questionnaire results or problems consistent with DSM-IV criteria or both); (2) probable ADHD, which consisted of either
a clinical diagnosis, but no supporting documentation or no clinical diagnosis,
but both types of supporting documentation; or (3) neither definite nor probable
ADHD.5,9 In this report, ADHD
case status included both definite and probable cases.
Studies of medical care use by Rochester residents are facilitated by
the geographical isolation and small number of providers. More than 95% of
all hospitalizations among residents occur at the 3 area hospitals affiliated
with the 2 practice groups, Mayo Clinic and Olmsted Medical Center (1988 MEDPAR
file, Health Care Financing Administration). Since 1987, use and billing data
from these institutions are electronically linked; individuals are identified
across institutions and over time. Therefore, the capacity exists to capture
information on all hospital and ambulatory care delivered by these providers
to area residents from January 1, 1987, through the present.10-13
The files serve as a major source of financial information and include line-item
detail on date, type, frequency, and billed charge for every good or service
provided.
This study assigned a standardized, inflation-adjusted cost to each
line item using a recently developed unit-costing algorithm.14
The algorithm differs depending on whether an item is covered under Medicare
Part A or Part B. The distinction is methodological and does not imply that
the database covers only Medicare patients. Costs for Medicare Part A items
(ie, primarily hospital-billed items provided to inpatients [eg, room and
board, radiology, physical therapy, supplies, etc]) are assigned using Medicare
cost reports. These reports include the total costs allocated to each revenue-generating
cost center within the hospital and the total charges billed by each cost
center, thus affording calculation of a cost-to-charge ratio for each cost
center at each of the 3 local hospitals for each year. Because the line-item
detail includes a designated cost center and a billed charge for each item,
a cost for each item can be calculated by multiplying the billed charge by
the cost-center–specific cost-to-charge ratio for the year in which
the service was delivered. The algorithm applies an inflation adjuster,15 and adjusts for geographical wage differences16 to express the costs for each year in 1995 national
average dollars.
Medicare Part B items consist primarily of those billed by physicians
(eg, examinations and consultations, diagnostic and therapeutic procedures),
irrespective of the site of care. Part B also covers services (eg, laboratory,
radiology, physical therapy, etc) provided to persons other than hospital
inpatients. Part B items are identified using the Health Care Financing Administration
Common Procedure Coding system. The costing algorithm uses Medicare national
fee schedules that provide a published allowed fee for each item coded by
the Health Care Financing Administration Common Procedure Coding system. A
cost is assigned to each item by applying the code-specific 1995 Medicare
national average allowed fee, therefore no inflation or geographical adjustment
is required.
Because 1987 was the first year complete data were available, the analyses
of cost and use were limited to the 5151 members of the birth cohort who had
not died or emigrated prior to 1987. Following identification of these individuals,
a Minnesota State statute was instituted that required patient authorization
for use of medical records for research.17
The 271 individuals who refused authorization were excluded from review. After
receiving institutional review board approvals, we reviewed the records of
each of the remaining 4880 individuals from January 1, 1987, until the earliest
event of emigration, death, or December 31, 1995, for data on total costs,
inpatient and outpatient hospitalizations, inpatient days, and ED visits.
The billing data also afforded characterization of birth cohort members
with respect to comorbid clinical diagnoses. The Johns Hopkins Ambulatory
Care Group case mix system software18 was used
to categorize every International Classification of Diseases,
Ninth Revision diagnosis code assigned to each person into 1 of 32
mutually exclusive diagnostic morbidity clusters known as an ambulatory diagnostic
group (ADG). The assignment is based on clinical similarity; the likelihood
of persistence or recurrence of the diagnosis, disability, and/or mortality;
the expected need for return visits, continued treatment, specialist services,
and/or hospitalization; and the expected need for and cost of diagnostic and
therapeutic procedures.18,19 The
ADG categorization scheme was particularly appropriate for the present study
because it was derived from a sample that included children and adolescents
and was developed for the purpose of examining differences in use.
Analyses of use and cost were performed for each calendar year and during
all 9 years; the latter were limited to persons residing in the area in 1995
(ie, who remained during the full period of follow-up). Between-group differences
in the likelihood of admission were tested using the χ2 statistic.
For persons with at least 1 admission, between-group differences in the number
of admissions were analyzed using the Wilcoxon rank sum test. To test whether
differences between persons with and without ADHD varied by age, the dollar
costs (log transformation) in each year of age were regressed against age
for each individual. Regression coefficients for persons with and without
ADHD were compared using the t test. To test whether
ADHD-associated costs varied by sex, the predicted (log-transformed) costs
for each individual at ages 7, 12, and 17 years served as dependent variables
in 3 separate regression analyses that included ADHD, sex, and an interaction
term as predictor variables. The sample sizes afforded more than 90% power
to detect a 22% increase in costs between individuals with and without ADHD.
There were 4880 members of the original 1976-1982 birth cohort residing
in the area in 1987 (mean age, 7.3 years); 350 met criteria for definite (n
= 284) or probable (n = 66) ADHD between the age of 5 years and the earliest
date of emigration, death, or December 31, 1995. In this study, birth cohort
members who met research criteria for an ADHD incidence case were defined
as having the condition throughout the study period (ie, both before and after
the date they met the criteria). The 350 ADHD individuals constituted 7.2%
(5.8% definite, 1.4% probable) of the cohort in 1987; the mean (SD) age for
individuals with and without ADHD was 7.2 (1.9) and 7.3 (2.1) years, respectively;
males constituted 75% of individuals with ADHD and 50% of individuals without
ADHD.
By 1995, 4119 members of the birth cohort were residing in the area;
individuals with ADHD constituted 7.5% (6.1% definite, 1.4% probable); the
mean (SD) age for individuals with and without ADHD was 15.2 (2.0) and 15.3
(2.0) years, respectively. The diagnoses assigned to these individuals since
1987 were categorized into ADG clusters. For each of the 32 ADG clusters,
the proportions of individuals assigned at least 1 diagnosis in that cluster
are provided in Table 1. Males
and females with ADHD were more likely than their non-ADHD counterparts to
have been assigned a diagnosis in almost every cluster. The differences reached
significance for both males and females in the psychosocial clusters. Significant
differences between ADHD and non-ADHD males were observed in a number of other
clusters, including signs/symptoms, asthma, and major injuries. There were
fewer clusters for which the differences reached significance for females,
probably due in part to the smaller numbers of female ADHD cases. Females
with ADHD were significantly more likely to have received a diagnosis of signs/symptoms
and asthma than their non-ADHD counterparts.
Data on the proportion of individuals with 1 or more medical encounters
are provided by encounter type for individuals with and without ADHD for each
of the years 1987-1995 in Figure 1.
The annual likelihood of a hospital inpatient admission was small and was
similar for individuals with and without ADHD in every year but 1992; among
those admitted, the number of inpatient days was similar for individuals with
and without ADHD in every year but 1994 (median, 8 vs 3 for ADHD and non-ADHD,
respectively; P = .04). The likelihood of a hospital
outpatient admission was greater for individuals with than without ADHD in
4 of the 9 years. Among those admitted, the median number of admissions was
1 for both groups in each year (P>.99 in 1987; P = .93 in 1988; P = .33 in 1989; P = .79 in 1990; P = .66 in 1991; P = .74 in 1992; P = .13 in 1993; P = .91 in 1994; and P = .60 in
1995). The likelihood of an ED admission was significantly increased for individuals
with ADHD compared with individuals without ADHD in 8 of the 9 years; among
those admitted, individuals with ADHD experienced significantly more admissions
than individuals without ADHD in 5 of the 9 years (P
= .02 in 1988; P = .002 in 1989 and 1990; and P = .04 in 1994 and 1995). An apparent decline in the proportion
admitted to the ED after 1989 was likely due to a shift from ED to urgent
care use following the opening of an urgent care facility in 1990.
In each year, the proportions of birth cohort members who were billed
for any medical care services or procedures ranged from 84% to 91% for individuals
with ADHD and 76% to 81% for individuals without ADHD. Medical care costs
were higher for individuals with ADHD compared with those without ADHD in
every year; comparisons of Part A (hospital-billed) and Part B (physician-billed)
dollars revealed that the differences were greatest for Part B costs (Figure 2).
In addition to comparisons between individuals with and without ADHD
for each year of follow-up, analyses were performed at the level of the individual
during the full 9 years of follow-up. The analyses were limited to the 309
individuals with ADHD and 3810 without ADHD who were residing in the area
in 1995. Whether between-group comparisons for these 4119 individuals were
representative of the entire cohort was addressed by testing for a significant
effect of ADHD, adjusted for age and sex, on the likelihood of remaining in
the area to 1995. Individuals with ADHD were more likely to remain than were
individuals without ADHD (odds ratio [OR], 1.13; 95% confidence interval [CI],
1.05-2.07; P = .02). The likelihood did not vary
as a function of sex (P = .20); individuals who were
older in 1987 were more likely to remain than those who were younger (P<.001). There were no interactions between ADHD and
age or sex; thus the effects of age and sex on loss to follow-up were similar
for individuals with and without ADHD.
Over the full 9 years, the likelihood of at least 1 admission was significantly
increased for individuals with ADHD compared with individuals without ADHD
for each encounter type: hospital inpatient (26% vs 18%; P<.001), hospital outpatient (41% vs 33%; P
= .006), and ED visit (81% vs 74%; P = .005). Compared
with their non-ADHD counterparts, individuals with ADHD who were admitted
experienced similar numbers of inpatient days (median, 3 vs 2; P = .30) and outpatient admissions (median, 2 vs 2; P = .10), but more frequent ED visits (median, 4 vs 3; P<.001). Median costs for all episodes of care during the 9 years
of follow-up for individuals with ADHD were more than double those for individuals
without ADHD ($4306 vs $1944; P<.001). To assess
the effect of age on costs, the log-transformed dollar costs in each year
of age were regressed against year of age for each of the 4119 individuals.
The regression coefficients differed from zero for both groups (ADHD, mean
[SD], 0.06 [0.35]; P = .001; non-ADHD, 0.07 [0.32]; P<.001), and did not differ between groups (P = .70 using the t test). These findings
suggest that costs increased with increasing age and that the age-associated
increases were similar for individuals with and without ADHD. Three regression
analyses, with predicted log-transformed costs at ages 7, 12, and 17 years
as dependent variables and ADHD and sex as predictor variables, revealed costs
were unaffected by sex at age 7 years (P = .70),
but were higher for females than males at ages 12 (P<.001)
and 17 years (P<.001). There was a significant
association between ADHD and costs (P<.001) in
all 3 models. Tests for interactions between sex and ADHD were insignificant,
suggesting that the contribution of ADHD to medical costs was similar for
males and females.
This population-based historical cohort study followed up 4880 individuals
from 1987 (age range, 5-11 years) through 1995 (age range, 13-19 years) to
compare individuals with and without ADHD for medical care use and costs.
During the 9-year period, individuals with ADHD compared with those without
ADHD exhibited a significantly increased likelihood of hospital inpatient,
hospital outpatient, and ED admissions. Among those admitted, the numbers
of inpatient days and outpatient admissions were similar for individuals with
and without ADHD; individuals with ADHD experienced more ED admissions than
did individuals without ADHD.
Median costs (including ambulatory costs) for individuals with ADHD
were greater than double the costs for individuals without ADHD. The differences
between individuals with and without ADHD were similar for males and females
and were consistent over all ages. The differences were greatest for Medicare
Part B (physician-billed) costs. The increases remained significant in analyses
limited to individuals who experienced some costs but did not experience any
hospital inpatient, hospital outpatient, or ED admissions (eg, median 1987
costs were $128 vs $65 for individuals with and without ADHD, respectively; P<.001). These findings suggest that non-ED, nonhospital
care may account for a substantial proportion of the excess costs associated
with ADHD.
These findings and those from Table
1 are consistent with multiple reports that individuals with ADHD
exhibit more psychosocial comorbidity, chronic health conditions, and adverse
medical outcomes (eg, substance abuse, automobile collisions, poisoning, and
fractures).4,20-22
There are few studies comparing individuals with and without ADHD for medical
care use. An Ontario Child Health Study survey found no difference between
children with and without ADHD in the likelihood of any ambulatory care encounter
in the prior 6 months.4,23 The
power to detect a difference in the Ontario Child Health Study was limited
by the relatively few cases (n = 147) and short observation period. A large
national health interview survey asked parents whether their child had a disabling
chronic mental health condition, and if yes, to identify the condition.24 Parents were asked to recall the number of physician
contacts and hospitalizations by their child in the prior year; compared with
children with no condition, those with ADHD experienced significantly more
physician contacts but no increased risk of hospitalization.24
Interpretation of these findings is problematic because the parent-reported
prevalence of disabling ADHD was 0.5%,24,25
a rate substantially lower than other estimates based on more generally accepted
criteria.2,26-29
The present study consisted of 350 ADHD cases followed up for up to
9 years. The presence of ADHD was established by applying comprehensive, standardized
research criteria with retrospective review of complete medical and school
records from age 5 to a maximum of 19 years among members of a population-based
birth cohort.5 The present study was also advantaged
in that the types and frequency of encounters were specified, and it afforded
comparison between individuals with and without ADHD for medical care costs.
The costs were adjusted for inflation and normalized to average national estimates.
To our knowledge, no other such comparisons exist.
The present study has a number of limitations. The compilation of costs
did not include those for services provided by the few private psychologists
and psychiatrists practicing in the area. The impact of these missing costs
is likely to be small; a list of birth cohort members ever seen at the largest
of these practices revealed less than 3% were seen for any condition. Costs
for outpatient drugs were also not included. Because individuals with ADHD
are frequently treated with medication on an outpatient basis over extended
periods, the cost differences observed here are likely an underestimate of
the differences that would be observed if outpatient drug costs were included.
The generalizability of study findings to the US population is limited by
the fact that more than 95% of Rochester residents are white and both the
proportion of the working population employed in the health care industry
and the education level are higher compared with the entire US white population.7 The generalizability of study findings is also limited
by marked differences in ADHD case ascertainment and detection over time and
across geographic regions.30,31
In the present study, the determination of case status was based on retrospective
record review by a single trained abstractor in close consultation with a
developmental behavioral pediatrician (W.J.B.). The cumulative incidence of
ADHD in this birth cohort from age 5 up to age 19 years is consistent with
point prevalence estimates from a majority of population-based studies that
used DSM criteria.2,26-29
It is important to recognize, however, that evidence of DSM-IV criteria was not necessary to qualify an individual as a case
in this study. Individuals could qualify as a definite case on the basis of
both a clinical diagnosis and positive results from either a parent or teacher
questionnaire, and they could qualify as a probable case on the basis of a
clinical diagnosis alone. Neither was evidence of DSM-IV criteria sufficient. Subjects with information in the records that
was consistent with DSM-IV criteria also had to have
either a clinical diagnosis of ADHD or positive results from standardized
parent and/or teacher questionnaires. To investigate the robustness of our
study findings, we reanalyzed the data limiting the definition of ADHD cases
to individuals with information in the records consistent with DSM-IV criteria (n = 221). Sixteen of these individuals did not qualify
as having either definite or probable ADHD using our research definition.
Median costs during the full 9 years of follow-up were $4829 (interquartile
range, $2535-$8230) for members who met DSM-IV criteria
compared with $4306 (interquartile range, $2222-$7150) using our definition.
Both were significantly higher (P<.001) than the
respective median costs for the remaining birth cohort members ($1945 vs $1907).
The data were also analyzed with definite cases categorized as ADHD (median
costs, $4317; interquartile range, $2331-$7092) and probable cases categorized
as non-ADHD (median costs, $1926; interquartile range, $847-$4124; P<.001). Thus, despite the fact that alternative criteria identified
different sets of individuals, the results were similar.
In conclusion, this study provides unique population-based longitudinal
estimates of health care use and costs for individuals with and without ADHD.
The findings have important clinical and public policy implications. They
suggest that the burden of ADHD extends beyond the recognized social, behavioral,
and academic outcomes3,4 to include
markedly increased use of medical care. Dramatic differences in cost were
observed for both sexes and at every age; the differences were not attributable
to a few high-cost individuals but were broadly based.
1.National Institutes of Health, Centers for Disease Control and Prevention. Diagnosis and Treatment of Attention Deficit Hyperactivity
Disorder: NIH Consensus Statement. Bethesda, Md: National Institutes of Health, Centers for Disease
Control and Prevention; 1998:1-37.
2.Anderson J, Werry JS. Emotional and behavioral problems. In: Pless IB, ed. The Epidemiology of Childhood
Disorders. New York, NY: Oxford University Press; 1994:304-338.
3.Hansen C, Weiss D, Last CG. ADHD boys in young adulthood: psychosocial adjustment.
J Am Acad Child Adolesc Psychiatry.1999;38:165-171.Google Scholar 4.Szatmari P, Offord DR, Boyle MH. Correlates, associated impairments and patterns of service utilization
of children with attention deficit disorder.
J Child Psychol Psychiatry.1989;30:205-217.Google Scholar 5.Barbaresi WJ, Katusic SK, Colligan RC.
et al. The cumulative incidence of diagnosed and documented AD/HD in a population-based
birth cohort.
Pediatr Res.2000;47:22A.Google Scholar 6.Katusic SK, Colligan RC, Barbaresi WJ, Schaid DJ, Jacobsen SJ. Potential influence of migration bias in birth cohort studies.
Mayo Clin Proc.1998;73:1053-1061.Google Scholar 7.Melton III LJ. History of the Rochester Epidemiology Project.
Mayo Clin Proc.1996;71:266-274.Google Scholar 8.American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders,
Fourth Edition. Washington, DC: American Psychiatric Association; 1994.
9.Barbaresi WJ. Primary care approach to the diagnosis and management of attention-deficit
hyperactivity disorder.
Mayo Clin Proc.1996;71:463-471.Google Scholar 10.Campion ME, Naessens JM, Leibson CL.
et al. The Olmsted County Benchmark Project.
Mayo Clin Proc.1992;67:5-14.Google Scholar 11.Leibson C, Naessens JM, Krishan I.
et al. Disposition at discharge and 60-day mortality among elderly people
following shorter hospital stays.
Gerontologist.1990;30:316-322.Google Scholar 12.Leibson CL, Naessens JM, Campion ME.
et al. Trends in elderly hospitalization and readmission rates for a geographically-defined
population.
J Am Geriatr Soc.1991;39:895-904.Google Scholar 13.Leibson CL, Hu T, Brown RD.
et al. Utilization of acute care services in the year before and after first
stroke.
Neurology.1996;46:861-869.Google Scholar 14.Wagner JL, Alberts SR, Sloan JA.
et al. Incremental costs of enrolling cancer patients in clinical trials.
J Natl Cancer Inst.1999;91:847-853.Google Scholar 15.Prospective Payment Assessment Commission. Medicare and the American Health Care System: Report
to Congress. Washington, DC: Prospective Payment Assessment Commission; 1997.
16. Health Care Financing Administration medical program: changes to the
hospital inpatient prospective payment systems and fiscal year 1998 rates.
62 Federal Register.460 (1997).Google Scholar 17.Melton III LJ. The threat to medical-records research.
N Engl J Med.1997;337:1466-1470.Google Scholar 18.ACG Case-Mix Development Team. Clinician's Guide: The Johns Hopkins University ACG
Case-Mix Adjustment System. Version 4.1. Baltimore, Md: Johns Hopkins University; 1998:E.1-E.61.
19.Starfield B, Weiner J, Mumford L, Steinwachs D. Ambulatory care groups.
Health Serv Res.1991;26:53-74.Google Scholar 20.Barkley RA, Murphy KR, Kwasnik D. Motor vehicle driving competencies and risks in teens and young adults
with attention deficit hyperactivity disorder.
Pediatrics.1996;98:1089-1095.Google Scholar 21.Barkley RA, Guevremont DC, Anastopoulos AD.
et al. Driving-related risks and outcomes of attention deficit hyperactivity
disorder in adolescents and young adults.
Pediatrics.1993;92:212-218.Google Scholar 22.Barkley RA, Fischer M, Edelbrock CS, Smallish L. The adolescent outcome of hyperactive children diagnosed by research
criteria.
J Am Acad Child Adolesc Psychiatry.1990;29:546-547.Google Scholar 23.Offord DR, Boyle MH, Fleming JE, Blum HM, Grant NIR. Ontario Child Health Study: summary of selected results.
Can J Psychiatry.1989;34:483-491.Google Scholar 24.Halfon N, Newacheck PW. Prevalence and impact of parent-reported disabling mental health conditions
among US children.
J Am Acad Child Adolesc Psychiatry.1999;38:600-609.Google Scholar 25.Costello EJ. Commentary on prevalence and impact of parent-reported disabling mental
health conditions among US children.
J Am Acad Child Adolesc Psychiatry.1999;38:610-613.Google Scholar 26.Wolraich ML, Hannah JN, Pinnock TY.
et al. Comparison of diagnostic criteria for attention-deficit hyperactivity
disorder in a county-wide sample.
J Am Acad Child Adolesc Psychiatry.1996;35:319-324.Google Scholar 27.Baumgartel A, Wolraich ML, Dietrich M. Comparison of diagnostic criteria for attention deficit disorders in
a German elementary school sample.
J Am Acad Child Adolesc Psychiatry.1995;34:629-638.Google Scholar 28.Pelham WE, Gnagy EM, Greenslade KE, Milich R. Teacher ratings of
DSM-III-R symptoms for
the disruptive behavior disorders.
J Am Acad Child Adolesc Psychiatry.1992;31:210-218.Google Scholar 29.Szatmari P, Offord DR, Boyle MH. Ontario Child Health Study.
J Child Psychol Psychiatry.1989;30:219-230.Google Scholar 30.Committee on Quality Improvement SoA-DHD. Clinical practice guideline.
Pediatrics.2000;105:1158-1170.Google Scholar 31.Kelleher KJ, McInerny TK, Gardner WP, Childs GE, Wasserman RC. Increasing identification of psychosocial problems.
Pediatrics.2000;105:1313-1321.Google Scholar