META-ANALYSIS

An Individual Participant Data Meta-analysis: Behavioral

Treatments for Children and Adolescents With
Attention-Deficit/Hyperactivity Disorder
Annabeth P. Groenman, PhD, Rianne Hornstra, MSc, Pieter J. Hoekstra, MD, PhD,
Laura Steenhuis, PhD, Asma Aghebati, PhD, Bianca E. Boyer, PhD,
Jan K. Buitelaar, MD, PhD, Andrea Chronis-Tuscano, PhD, David Daley, PhD,
Parisa Dehkordian, MSc, Melissa Dvorsky, PhD, Nike Franke, PhD, George J. DuPaul, PhD,
Naama Gershy, PhD, Elizabeth Harvey, PhD, Timo Hennig, DrPhil, Sharonne Herbert, PhD,
Joshua Langberg, PhD, Jennifer A. Mautone, PhD, Amori Yee Mikami, PhD,
Linda J. Pfiffner, PhD, Thomas J. Power, PhD, Sijmen A. Reijneveld, MD, PhD,
Satyam Antonio Schramm, DrPhil, Julie B. Schweitzer, PhD, Margaret H. Sibley, PhD,
Edmund Sonuga-Barke, PhD, Catharine Thompson, MSc, BSc, MBBS,
Margaret Thompson, MBChB, MD, FRCPsy, FRCP, Carolyn Webster-Stratton, MSN, MPH, PhD,
Yuhuan Xie, MD, PhD, Marjolein Luman, PhD, Saskia van der Oord, PhD,
Barbara J. van den Hoofdakker, PhD
Dr. Luman, Prof. Dr. van der Oord, and Prof. Dr. van den Hoofdakker share last authorship and contributed equally to this work.

Objective: Behavioral interventions are well established treatments for children with attention-deficit/hyperactivity disorder (ADHD). However,
insight into moderators of treatment outcome is limited.
Method: We conducted an individual participant data meta-analysis (IPDMA), including data of randomized controlled behavioral intervention trials
for individuals with ADHD <18 years of age. Outcomes were symptoms of ADHD, oppositional defiant disorder (ODD), and conduct disorder (CD)
and impairment. Moderators investigated were symptoms and impairment severity, medication use, age, IQ, sex, socioeconomic status, and single
parenthood.
Results: For raters most proximal to treatment, small- to medium-sized effects of behavioral interventions were found for symptoms of ADHD, inat-
tention, hyperactivity/impulsivity (HI), ODD and CD, and impairment. Blinded outcomes were available only for small preschool subsamples and lim-
ited measures. CD symptoms and/or diagnosis moderated outcome on ADHD, HI, ODD, and CD symptoms. Single parenthood moderated ODD
outcome, and ADHD severity moderated impairment outcome. Higher baseline CD or ADHD symptoms, a CD diagnosis, and single parenthood
were related to worsening of symptoms in the untreated but not in the treated group, indicating a protective rather than an ameliorative effect of behav-
ioral interventions for these children.
Conclusion: Behavioral treatments are effective for reducing ADHD symptoms, behavioral problems, and impairment as reported by raters most prox-
imal to treatment. Those who have severe CD or ADHD symptoms, a CD diagnosis, or are single parents should be prioritized for treatment, as they
may evidence worsening of symptoms in the absence of intervention.
Key words: behavioral interventions, ADHD, individual participant data meta-analyses, moderator analyses
J Am Acad Child Adolesc Psychiatry 2022;61(2):144−158.

ttention-deficit/hyperactivity disorder (ADHD)1 comorbidities are oppositional defiant disorder (ODD),

A in children and adolescents is highly prevalent2

and impairing in multiple domains of function-
ing.3 It is a heterogeneous disorder regarding etiology,
conduct disorder (CD), anxiety disorder, and depression.4
Pharmacological and behavioral interventions are well-
established treatments for children and adolescents with
symptomatology, functional impairments, developmental ADHD,5−7 with behavioral interventions often targeting
expression, and comorbid psychopathology; common comorbid pathology and ADHD-related impairments.5

144 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry
Volume 61 / Number 2 / February 2022
 BEHAVIORAL INTERVENTIONS FOR ADHD

There is a clear need for behavioral treatments in clinic- if available. We examined whether several variables that are
based practice, as they may be preferred by patients and routinely identified in everyday clinic-based practice moder-
families8,9 and can reduce the need for medication.9−11 The ated treatment effects. Identification of such moderators
heterogeneity of the symptomatology and comorbidity of could yield a more personalized approach to intervention in
children with ADHD makes it unlikely that behavioral clinical settings. We used a hypothesis-generating approach,
treatments will work equally for all individuals, emphasizing as the increased power of IPDMA may generate significant
the need for more personalized treatment plans. moderators that have never emerged in aggregated meta-
Moderation analysis allows the identification of sub- analyses or individual studies.12,16 Candidate moderators
groups of children (or families) that are more or less likely included clinical characteristics of the child (ie, severity of
to respond to behavioral interventions.12 To date, meta- symptoms of ADHD, ODD, CD, and internalizing prob-
analyses in ADHD samples could not consistently identify lems, and impairment; comorbidity with ODD, CD, anxi-
moderators of behavioral treatment response, most possibly ety disorder, and depression; and medication use) and
due to lack of power or related to the diversity of sample demographic variables (ie, child’s age, IQ, and sex; social
compositions and study designs.13 For example, child age economic status of the family, and single parenthood).
was found to be a significant moderator in one meta-analy-
sis9 but not in two others.14,15
Individual treatment studies are mostly not designed or METHOD
powered for moderation analyses.12,16 An example of one of TagedPThis IPDMA has been registered in Prospero (https://www.
the few adequately powered studies for moderation analyses crd.york.ac.uk/prospero/display_record.php?RecordID=
in ADHD samples is the Multimodal Treatment Study of 69877). PRISMA IPDMA guidelines for reporting were fol-
ADHD (MTA),16,17 showed comorbid anxiety of the child lowed, and a checklist is available in Supplement 1, available
to be a moderator of (better) behavioral treatment online.
response.9,18 The MTA study, however, was limited in age
range (7-9 years) and in the comparability of the behavioral Identification and Selection of Studies
treatment arm to other behavioral treatments (ie, the MTA Inclusion and Exclusion Criteria. We included randomized
behavioral treatment was far more intense than most controlled trials (RCTs) of behavioral treatments of individ-
others). In sum, evidence regarding moderating factors is uals of individuals less than 18 years of age with ADHD
currently inconsistent or lacking, thereby hindering clini- (corroborated by clinical cutoffs on questionnaires or
cians from personalizing treatment plans. [semi]-structured interviews) and/or participants meeting
In contrast to individual randomized controlled trials, clinical cutoffs on questionnaires or (semi)-structured inter-
individual participant data meta-analysis (IPDMA) is partic- views. RCTs had to compare behavioral interventions with
ularly appropriate to run highly powered moderation analy- a control condition (including active control conditions,
ses, as it uses individual data from the original studies, except for medication; see below), or studies that compared
leading to uniform conclusions across studies.19 The 2 behavioral interventions. For the studies with head-to-
IPDMA approach has been shown to be of great advantage head comparisons of 2 behavioral treatments, we coded
in the general field of medicine; for example, an IPDMA of both active interventions as intervention, but for studies
Furukawa et al.20 on treatment of depression resulted in an with an active non-pharmacological control condition (eg,
interactive Web tool that shows the individual predicted weekly support groups), we coded the control condition as
disease course when taking the participants’ characteristics an active control condition. Behavioral interventions were
into account. However, this is the first IPDMA on the treat- defined as interventions directed at changing children/ado-
ment of children and adolescents with ADHD. In the cur- lescents’ behaviors (ie, increasing desirable behaviors and
rent study, we used IPDMA to identify behavioral decreasing undesirable behaviors), using (cognitive) behav-
intervention effects and moderators of outcomes for symp- ioral therapeutic techniques modeled to the definition used
toms of ADHD, ODD, and CD, and global impairment in by Daley et al.11 These include cognitive-behavioral inter-
children and adolescents with ADHD (<18 years of age). ventions, such as parent- and teacher-mediated treatments,
We focused primarily on outcome measures taken from as well as cognitive-behavioral interventions aimed directly
reporters most proximal to the delivery of the treatment: at the child/adolescent, such as behavioral skills training or
parent-rated outcomes for parent training and child-focused cognitive-behavioral therapy. We excluded studies or inter-
treatments, and teacher-rated outcomes for interventions vention arms that used optimized medication treatment
that were primarily school based. Furthermore, we aimed to next to a behavioral intervention as part of their study design
explore intervention effects on probably blinded measures, or as a control condition. English-, German-, and Dutch-
Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 145
Volume 61 / Number 2 / February 2022
GROENMAN et al.

language publications published in peer-reviewed journals within studies (an overview of instruments per study used
were included. No date restriction was applied. for each outcome domain can be found in Table S1, avail-
able online). Conversion into z scores was necessary because
Selection and Screening of Studies. A 2-step approach of the large heterogeneity in instruments used between stud-
was used to identify relevant articles. First, we searched ies. The analyses for ADHD, ODD, and CD focused pri-
Medline, CINAHL, PsycINFO, EMBASE+EMBASE marily on raters closest to the intervention: that is, parent
CLASSIC, ERIC, Web of Science (Science Citation Index ratings for parent training, multimodal interventions, and
Expanded) up to July 2017 for relevant papers using a com- child-focused interventions and teacher ratings for pure
bination of the following search terms and their synonyms, school-based interventions; for impairment parent ratings
as well as hierarchical family form (eg, MeSH terms): treat- were used or, if these were not available, a clinician or
ment specific terms (ie, behavioral treatment, psychosocial researcher ratings. Separate exploratory analyses were per-
treatment, parent and/or teacher training), ADHD, child formed, if possible, on blinded ratings (ie, observations by a
and adolescent, and randomized controlled trial (complete blinded rater) of ADHD, ODD, and CD symptoms. Cod-
search criteria available in Supplement 2, available online). ing of sex (men/women), age at baseline (in years), medica-
Two authors (APG, senior researcher, and RH, PhD stu- tion use at baseline (yes/no), IQ of the child, single
dent) performed the selection and screening of studies using parenthood (yes/no), socioeconomic status (low [<high
Rayyan, a web and mobile app for systematic reviews21; dis- school], medium [high school graduate or some college edu-
agreement was resolved by consensus with a third person cation], high [>college graduate], and diagnoses of ODD,
(either ML, SvdO or BvdH, senior researchers). Second, lit- CD, any anxiety disorder, and depression [including major
erature lists of relevant articles (identified studies and previ- depressive disorder and dysthymic disorder] [yes/no] were
ous systematic reviews and meta-analyses) were hand- equalized across datasets.
searched to identify possible missing articles.
Risk of Bias Assessment. Risk of bias assessment of the
Data Collection and Management included studies was done independently by 2 authors (a
Data Collection. We contacted the corresponding authors of combination of AG/ RH/LS) using the Cochrane risk of
all selected trials by e-mail to ask to share their data, with a bias tool. Random sequence generation, allocation conceal-
reminder after several weeks. If we failed to establish contact ment, blinding of outcome assessment, incomplete outcome
with the corresponding author, we emailed co-authors of the data, vested interest, and selective outcome reporting were
study. Furthermore, we contacted authors of selected papers rated on a 3-point scale (no risk of bias, unknown, or risk of
during conferences and through our personal network. Each bias). Any disagreement was resolved by consensus.
author signed an understanding agreeing that they were
responsible for ethical clearance in using the data. To assess Analyses
differences in results and baseline characteristics (age, sex, Analyses Between Studies Providing Data and Not Pro-
comorbidities) between the studies providing data and the viding Data. To assess possible differences between studies
studies not providing data, APG and RH also extracted the that provided data and those that did not, we examined
aggregated data from the published papers of the studies. (available) demographic differences (percentage of male par-
ticipants, ODD and CD, and mean age), region of origin
Data Checks and Harmonization. First, we checked the (Europe, Northern America, other), sample size, type of
data of all studies as obtained, and harmonized measures. intervention (parent, teacher, child, or multimodal treat-
Regarding data checks, the number of participants in each ment), and publication date based on data described in the
study was compared to the number of participants in the manuscript using independent t tests. Furthermore, we
published paper. We also checked the data by comparing examined whether there were differential intervention
intervention and control groups on sex, age and ADHD effects between studies that provided data and those that
severity (if available) with the published data. Authors were did not. For studies that provided data, we calculated ESs
contacted if any deviations from the reported data were based on the IPD; for studies in which IPD was not avail-
found, and inconsistencies were resolved. able, ESs were calculated based on aggregated data from the
Regarding harmonization, for each dataset continuous manuscripts. ESs were compared using random effects
measures (severity of total ADHD, inattentive, hyperactiv- meta-analysis with the “Metafor” package in R, for the
ity/impulsivity [HI], ODD, and CD symptoms; severity of main outcomes (ADHD, inattentive symptoms, HI, ODD,
internalizing problems, and global impairment) were con- CD, and global impairment). It should be noted that these
verted into z scores, using pre-intervention score SDs comparisons were possible only for those studies with a
146 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry
Volume 61 / Number 2 / February 2022
 BEHAVIORAL INTERVENTIONS FOR ADHD

control group, and not for head-to-head comparisons (ie, depended on the type of control condition (WL or
comparisons of 2 active treatments). In studies performing TAU/active control), we further tested whether there was a
head-to-head comparison, both interventions were coded as smaller intervention effect on the main outcome in studies
intervention. with a TAU/active control condition, compared to a WL
condition.
Analyses of Main and Moderator Effects. Analyses were
conducted using the “LME4” package in R (version
1.2.1335).22 We determined effects using a 1-stage RESULTS
IPDMA, in which data from participants across studies Collected Data
were analyzed in 1 stage, with a random intercept for We identified 17,897 studies; after removal of duplicates,
study. First, linear multilevel analysis was used to examine 10,351 studies were screened based on title and abstract.
effects of behavioral interventions on symptoms of total Authors AG and RH disagreed on 1.2% of decisions; these
ADHD, inattentive, HI, ODD, and CD symptoms and were resolved by consensus. From 284 full-text papers that
global impairment. Post-measurements of total ADHD, were screened, 62 were deemed eligible. Regarding these,
inattentive, HI, ODD, and CD symptoms and global we received 23 datasets. A summary of unrecovered stud-
impairment were used as outcomes in these models, and ies (n = 37) is available in Table S2, available online. In
pre-intervention measures and intervention group were addition, we received data from 2 studies that were pub-
added. Because all outcomes were transformed to z scores, lished at a later date than our initial search25,26 (also see
the regression coefficients can be interpreted as effect sizes Figure 1 for PRISMA flow chart). To assess whether the
(ESs) (a coefficient of 1 is a change of 1 SD in the outcome addition of these 2 studies influenced the results all analy-
measure) and can be interpreted as small d = 0.2; medium, ses were rerun excluding these studies (see Supplement 3
d = 0.5; and large, d = 0.8.23 Heterogeneity between stud- [Tables S3 S8], available online); results remained
ies with regard to the outcome measure was assessed using approximately similar). Data checks on the provided data
intraclass correlation (ICC). This measure varies between led to only minor deviations that could be resolved with
0 (low clustering within studies) and 1 (high clustering the corresponding author, and mostly considered the
between studies) and determines the proportion of vari- inclusion of more participants in the IPDMA compared
ance accounted for by clustering within studies. Of note, to the reported data.
this is not an appropriate indication of variance in effect With a recent systematic search (up to May 13, 2020),
sizes between studies, as we included both studies compar- using the same criteria and methods as for the current study,
ing 2 interventions and studies with a control group as we verified whether we had missed more recent studies. We
comparison. identified 8 new studies that met our inclusion criteria, of
Second, to test effects of the candidate moderators, we which 2 studies25,26 were already identified when inquiring
added the interaction between intervention group and can- about older studies. Data from these 2 studies were included
didate moderator to the models, with separate models for in the current IPDMA; the other 6 studies were not
each outcome and moderator. Benjamini Hochberg24 cor- included.
rection (based on 17 tests per outcome) was used to exert A total of 2,885 participants (1,936 intervention and
control over the false discovery rate. 949 control) with a mean age of 8.78 years (SD = 3.32
years; range 2 17.5 years) were included from a total of 25
Prespecified Additional Analyses. First, significant moder- studies (Table 127−48 for a summary of studies that pro-
ator effects were further explored by examining 3-way inter- vided data). In almost all studies, the informant was the par-
actions between the moderator and possible explanatory ent, except for 1 school-based treatment for which teacher
variables included in the IPDMA. Second, analyses were ratings were used.31 For the current paper, 21 studies
rerun separately for parenting interventions and child/ado- (n = 2,233) had most proximal ratings available. Only 3
lescent-focused treatments, to assess the effect of interven- studies of parenting interventions (all in preschool children)
tion type. Third, analyses were re-run on blinded measures. contained data on a probably blinded inattention measure
Fourth, in addition to dimensionally looking at age as a (directly observed time on task, n = 295),26,30,45 and only 4
moderator, a categorical variable of age was also used to studies (all in preschool children) contained data on blinded
assess differential treatment effects by comparing 3 mean- outcome measures for disruptive behaviors (child noncom-
ingful developmental stages: children <6 years; children pliance, n = 175).34,35,45,47 Given the small sample size, spe-
between 6 and 12 years; and children >12 years. Finally, cific age group, and variability in measures, analyses on the
to examine whether the deterioration of ADHD symptoms probably blinded outcomes can be found in Supplement 4,
Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 147
Volume 61 / Number 2 / February 2022
GROENMAN et al.

FIGURE 1 PRISMA Individual Participant Data (IPD) Flow Diagram

148 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry
Volume 61 / Number 2 / February 2022
Volume 61 / Number 2 / February 2022
Journal of the American Academy of Child & Adolescent Psychiatry

TABLE 1 Characteristics of Included Studies

Medication Single-
use at Sex, parent
Intervention Age baseline, female SES family, ODD CD Depression Anxiety Country
Total Intervention Control recipient mean (SD) yes n (%) n (%) n (%) yes n (%) n (%) n (%) n (%) n (%) of origin Ethnicityb
Low Medium High
Aghebati et al. 201427 27 14 Triple P 13 Parent 8.04 (1.40) 27 (100) 11 (40.7) 7 (25.9) 11 (40.7) 9 (33.3) 0 (0) Iran
Boyer et al. 201528 184 94 Plan My Life, 91 Solution- 0 Multimodal 14.45 (1.26) 141 (79.2) 53 (28.6) 31 (17.0) 84 (46.2) 67 (36.8) 25 (14.4) 57 (30.8) 0 (0) 32 (20.1) 14 (8.8) Netherlands
focused treatment,
Chronis-Tuscano et al. 98 51 Standard BPT, 47 0 Parent 8.78 (2.06) 58 (59.2) 33 (33.7) 1 (1.1) 30 (33.0) 60 (65.9) 51 (52) 17 (17.3) USA Caucasian (48.98%), African
201329 integrated parenting American (30.61), Other
intervention (IPI-A) (20.41)
Daley et al. 2013 30 43 19 New Forest Parenting 24 Parent 7.28 (1.53) 0 (0) 8 (18.6) UK
Programme
31
Dupaul et al. 2006 176 0 Teacher 8.58 (1.19) 51 (29) 43 (24.4) 18 (8.6) 136 (64.8) 56 (26.7) 0 (0) 0 (0) USA White (58.23)
Franke et al. 201632 53 27 Triple P 26 Parent 3.97 (0.59) 38 (71.7) 10 (18.9) 14 (26.4) 29 (54.7) New Zealand New Zealand European
ethnicity (79.2%)
Gershy et al. 2017 33 57 23 Nonviolent resistance 34 Parent 9.56 (2.49) 14 (17.7) 42 (28.0) 82 (54.7) 26 (17.3) Israel
(NVR)
Herbert et al. 201334 31 17 Parenting Your 14 Parent 4.58 (0.89) 8 (25.8) 8 (33.3) 4 (16.7) 12 (50.0) 15 (50) USA European American (83,87),
Hyperactive Preschooler African American (6,25)
Latino (3,13) Multietnic
(6,25)
Langberg et al. 201825 280 111 Homework, 52 Multimodal 11.98 (1.03) 149 (53.2) 75 (26.8) 117 (45.9) 89 (34.9) 49 (19.2) 127 (45.4%) 85 (30.4) 10 (3.6) 12 (4.3) 73 (26.1) USA White (54.74%), Black
organization, and (28.46%). Asian (1.46%),
planning skils American Indian (1.1 %),
111 Completing homework Multiracial (0.58%)
by improving efficiency
and focus
Mautone et al. 201235 61 29 Family school success 32 Multimodal 6.48 (0.60) 15 (24.6) 17 (27.9) 6 (4.9) 20 (16.4) 96 (78.7) 12 (19.7) 18 (29.5) 0 (0) 0 (0) 10 (16.4) USA Non-Hispanic (88%),
early elementary Hispanic (12%),White (75%)
Black/African American
(21%)Multiracial (3%),
Mikami et al. 201036 62 32 Parent friendship 30 Parent 8.26 (1.21) 40 (64.5) 20 (32.3) 1 (1.6) 26 (42.6) 34 (55.7) 5(8.2) 20(32.2) 3(4.8) USA White (85%),African
coaching American (5%), Asian
American (2%),Latino (1%),
More than 1 race (7%)
Mikami et al. 201337 24 12 Contingency 0 Teacher 9 (56.3) 11 (45.8) 0 (0) 12 (50) USA White (81%),Asian American
management training, 12 (6%),Latino (2%),More
making socially accepting than 1 ethnicity (8%)
inclusive classrooms
MTA group 199917 a 287 143 Parent training summer 144 Multimodal 7.76 (0.82) 57 (19.9) 121 (45.3) 41(14.3) 14(4.9) 96 (33.4) USA White (61%),African
treatment program American (20%),Hispanic
(8%),
Pfiffner et al. 200738 69 36 Child life and attention 33 Multimodal 8.67 (1.16) 2 (3.5) 23 (33.3) 1 (1.5) 17 (25.4) 49 (73.1) 16 (23.2) 0 (0) 1 (1.4) 8 (11.8) USA White (51%),Asian (16%),
skills program Hispanic (10%),African
American (6%),mixed

BEHAVIORAL INTERVENTIONS FOR ADHD

(17%)
Pfiffner et al. 201439 199 74 Child life and attention 51 MultimodalParent 8.64 (1.16) 7 (3.5) 83 (41.7) 0 (0) 37 (18.8) 160 (81.2) 25 (12.6) 31 (15.6) 5 (2.6) 4 (2) 24 (12.1) USA Caucasian (54%),Latino
skills program,74 parent- (17%),Asian (8%),African
www.jaacap.org

focused treatment American (5%),mixed race

(17%)
Pfiffner et al. 201640 135 72 Collaborative life skills 63 8.39 (1.13) 12 (8.9) 39 (28.9) 6 (4.5) 47 (35.1) 81 (60.4) 40 (29.6) 54 (40) 7 (5.2) USA White (26.8%),African
American (8.93%),Asian
(20.6%).Hispanic/Latino
(23.8%),Multiracial/
multiethnic (19.87%)
Power et al. 201241 199 100 Family school success 99 Multimodal 9.42 (1.29) 85 (42.7) 63 (31.7) 4 (2.0) 28 (14.1) 167 (83.9) 42 (21.1) 54 (27.1) 0 (0) 5 (2.5) 46 (23.1) USA Non-Hispanic (93%),Hispanic
(7%),White (72%),Black/
African American (22%),
Multiracial (4%),Asian (2%)

(continued)
149
 150

GROENMAN et al.
TABLE 1 Continued
Medication Single-
use at Sex, parent
Intervention Age baseline, female SES family, ODD CD Depression Anxiety Country
Total Intervention Control recipient mean (SD) yes n (%) n (%) n (%) yes n (%) n (%) n (%) n (%) n (%) of origin Ethnicityb
Schramm et al. 201642 113 40 Lerntraining f€
ur 73 Multimodal 13.99 (1.43) 56 (49.6) 16 (14.2) Germany
Jugendliche mit ADHS—
LeJA (Learning Skills
Training for
www.jaacap.org

Adolescentswith ADHD)
Sibley et al. 201343 36 18 STAND 18 Multimodal 12.39 (1.02) 20 (55.6) 10 (27.8) 2 (5.6) 12 (33.3) 22 (61.1) 7 (19.4) 10 (27.8) 5 (15.6) 21 (63.6) USA White non-Hispanic (25%),
Black non-Hispanic
(8.35%),Hispanic, any race
(61.15%),Mixed race
(5.55%)
Sibley et al. 201644 128 67 STAND 61 Multimodal 12.74 (0.86) 44 (34.3) 45 (35.2) 24 (19.4) 21 (16.9) 79 (63.7) 45 (35.2) 74 (57.8) 17 (13.3) USA Non-Hispanic white (7%),
African American (10.8%),
Hispanic (78.5%),other
(3%)
Sonuga-Barke et al. 306 133 New Forest Parenting 42 Parent 3.51 (0.58) 0 (0) 82 (26.8) 92 (33) UK
201826 Programme,131
Incredible Years
Thompson et al. 200945 41 21 New Forest Parenting 20 Parent 4.18 (1.05) 0 (0) 50 (100.0) 32 (76.2) 7 (16.7) 3 (7.1) UK
Programme
Van Den Hoofdakker et 94 47 Behavioral parent training 47 Parent 7.43 (1.95) 47 (50.5) 18 (19.1) 32 (34.4) 38 (40.9) 23 (24.7) 10 (10.6) 71 (75.5) 41(43.6) Netherlands White (94.7%),African (2.1),
al. 200746 Asian (2.1),unknown (1.1)
Webster-Stratton47 99 49 Incredible years, 50 Multimodal 5.36 (0.91) 0 (0) 24 (24.2) 20 (20.8) USA Minority (27.3%)
combined parent and
child
Xie et al. 201348 22 13 Parent Training face to 0 Parent 8.95 (1.89) 7 (31.8) USA
face,9 Parent training
video-conferencing

Note: CD = conduct disorder; n = number of participants; ODD = oppositional defiant disorder; SES = socioeconomic status; STAND = Supporting Teens' Academic Needs Daily. .
a
Some data were not available in the public access database of the Multimodal Treatment Study of ADHD (MTA) (ie, medication at baseline and single parenthood).
b
Data extracted and reported identically from the published manuscripts.
Journal of the American Academy of Child & Adolescent Psychiatry
Volume 61 / Number 2 / February 2022
 BEHAVIORAL INTERVENTIONS FOR ADHD

available online. In short, no effects of behavioral interven- not participating showed that in the studies not participat-
tions were found on these outcomes. ing, there were slightly more problems in allocation conceal-
ment, and more potential vested interests. Overall, 28% of
Risk of Bias studies had risk-of-bias violations (high risk of bias) on 2 or
Risk of bias agreement was high (k = 0.92) and is presented more items. Risk of bias (as measured by percentage of “no
in Figure 2 (separate panels for those studies providing data bias”) was not associated with any of the results on the out-
and those not; see Figure S1 and Figure S2, available come measures (ADHD p = .23, inattention p = .17, HI
online). Results (based on manuscripts) showed that for p = .31, ODD p = .92, CD p = .17, impairment p = .75).
many studies it was unclear how random sequences were
generated (56%) and how allocation concealment took Analyses Between Studies Providing Data and Not
place (68%). In almost all studies, selective reporting was Providing Data
unclear (84%), as few studies preregistered their trials. A Studies not providing data did not differ in age
visual comparison between studies participating and those (mean = 8.48 versus mean = 8.87, p = .70), percentage of

FIGURE 2 Moderating Effects on Behavioral Treatment for Attention-Deficit/Hyperactivity Disorder

Note: (A, B) Moderating effect of CD diagnosis on ADHD symptoms post-intervention. (C) Moderating effects of CD symptoms at baseline on ADHD outcome post-inter-
vention. (D) Moderating effects of CD symptoms at baseline on hyperactive impulsive outcome post-intervention. (E) Moderating effects of CD symptoms at baseline on
ODD outcome post-intervention. (F) Moderating effects of CD symptoms at baseline on CD outcome post-intervention. (G, H) Moderating effect of single parents. (I) Mod-
erating effects of ADHD symptoms at baseline on impairment post-intervention. The dashed black line was added for illustrative purposes; it represents the regression line
where baseline severity is similar to post-intervention severity (ie, x = y). A regression line of the intervention and/or control group below the dashed black line at x = 0 indi-
cates improvement at post-measurement compared to baseline, and a regression line above the dashed black line at x = 0 indicates deterioration at post-measurement.
Number of individuals with a CD diagnosis and with ADHD outcomes: n = 96. ADHD = attention-deficit/hyperactivity disorder; CD = conduct disorder; HI = hyperactivity/
impulsivity, imp = impairment; ODD = oppositional defiant disorder; symp = symptoms. Please note color figures are available online.

Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 151
Volume 61 / Number 2 / February 2022
GROENMAN et al.

TABLE 2 Main Effects of Behavioral Interventions

n k Estimate SE df t p
ADHD 2,109 20 −0.42 0.05 716.90 −9.17 <.001
Parent interventions 1,689 17 −0.42 0.05 782.38 −8.8 <.001
Child focused 423 3 −0.67 0.12 287.26 −5.68 <.001
Inattention 2,173 20 −0.46 0.05 662.48 −9.52 <.001
Parent interventions 1,635 16 −0.45 0.05 758.61 −8.61 <.001
Child focused 541 4 −0.64 0.10 290.02 −6.17 <.001
HI 2,233 21 −0.27 0.04 2230.00 −7.43 <.001
Parent interventions 1,692 17 −0.31 0.04 1689 −7.40 <.001
Child focused 544 4 −0.35 0.08 138.94 −4.52 <.001
ODD 1,798 16 −0.19 0.04 1413.62 −4.83 <.001
Parent interventions 1,568 14 −0.20 0.04 1460.05 −4.93 <.001
a
Child focused
CD 1,229 12 −0.20 0.06 1226.00 −3.38 <.001
Parent interventions 882 9 −0.23 0.05 879.00 −4.92 <.001
a
Child focused
Impairment 664 8 −0.59 0.10 661 −5.96 <.001
Parent interventions 567 7 −0.68 0.11 564.00 −6.19 <.001
a
Child focused

Note: Full analyses on parent- and child-focused treatments are provided in Supplements 6 and 7 (Tables S15−S26), available online. ADHD = atten-
tion-deficit/hyperactivity disorder; CD = conduct disorder; df = degrees of freedom; HI = hyperactivity/impulsivity; k = number of studies; n = number
of individuals; ODD = oppositional defiant disorder; SE = standard error.
a
There were not enough observations to run these analyses.

ODD (mean = 50.60% versus mean = 40.14%, p = .29), or .001) (Table 2, and Supplements 6 and 7 [Tables
CD (mean = 7.04% versus mean = 19.89%, p = .12), region S15 S26], available online). Symptoms of CD (b = 0.19,
of origin p = .53), type of intervention (p = .42), but studies p = .002) and a diagnosis of CD at baseline (b = 0.55, p =
not providing data did have a slightly larger percentage of .008) moderated the intervention effect. Figures 2 and 3
male participants (mean = 79.64% versus mean = 70.93%, show that this effect was driven by an increase in ADHD
p = .007), smaller samples sizes (mean = 64 versus symptoms in the control group for children with more base-
mean = 113, p = .017), and an earlier publication date line symptoms of CD Figure 2C, and/or a baseline diagnosis
(mean = 2008 versus mean = 2012, p = .044). No differen- of CD Figure 2A and B, compared to youths with fewer CD
ces in ESs were found for the intervention effects on symptoms and/or no CD diagnosis, or youths in the inter-
ADHD symptoms (difference in ESs [estimate] = 0.18, vention group. No other investigated candidate factor mod-
p = .39), inattentive symptoms (estimate = 0.04, p = .80), erated the intervention effect on ADHD symptom severity
hyperactive/impulsive symptoms (estimate = 0.11, p = .48) (Table 3).
or CD (estimate = 0.03, p = .89). However, studies that Additional analyses (see Supplement 8, available online)
provided data appeared to have a smaller effect on ODD showed that the moderating effect of CD symptom severity
symptoms (estimate = 0.29, p = .03) compared to studies and/or CD diagnosis did not vary as a function of any of
that did not provide data. the other variables, and remained significant in parenting
interventions only (b = 0.22, p = .002; b = 0.55, p =
ADHD Symptoms .008, respectively).
Behavioral interventions had a positive, small to medium ES
on ADHD symptoms (b = 0.42, p < .001) (Table 2, and Inattention Symptoms
Supplement 5 [Tables S9 S14], available online). The pro- Behavioral interventions had a positive, small- to medium-
portion of variance accounted for by clustering within stud- sized effect on inattention symptom severity (b = 0.46, p
ies was low (ICC = 0.05). This effect remained significant < .001), with low clustering within studies (ICC = 0.07).
in parenting interventions (b = 0.42, p < .001) and in This effect remained significant in parenting interventions
child/adolescent focused interventions (b = 0.67, p < only (b = 0.45, p < .001) and child-focused interventions
152 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry
Volume 61 / Number 2 / February 2022
 BEHAVIORAL INTERVENTIONS FOR ADHD

TABLE 3 Moderating Effects of Child/Adolescent and Family Factors on Outcomes of Behavioral Interventions for Attention-
Deficit/Hyperactivity Disorder (ADHD)

ADHD Inattention HI ODD CD Impairment

N k ES n k ES n k ES n k ES n k ES n k ES
ADHD symptoms 2,109 20 −0.09 2,035 18 −0.01 2,112 20 −0.03 1,786 16 −0.01 1,104 11 −0.12 659 8 0.50
Age 2,109 20 0.10 2,173 20 −0.03 2,233 21 0.03 1,798 16 −0.01 1,229 12 0.01 664 8 −0.03
CD symptoms 1,309 12 −0.19 1,405 12 −0.09 1,435 13 −0.14 1,092 10 −0.21 1,229 12 −0.21 563 6 −0.17
Diagnosis anxiety 1,199 9 −0.05 1,208 9 −0.25 1,208 9 −0.15 1,001 8 −0.13 647 5 −0.07 361 4 0.90
Diagnosis CD 1,515 12 −0.55 1,528 12 0.20 1,529 12 −0.37 1,320 11 −0.36 1,063 9 −0.44 604 6 −0.90
Diagnosis depression 1,105 8 0.01 1,114 8 −0.47 1,114 8 −0.46 907 7 −0.15 648 5 −0.19 360 4 −2.06
Diagnosis ODD 1,624 14 0.11 1,636 14 0.09 1,637 14 0.01 1,428 13 −0.09 1,051 9 −0.22 606 6 −0.31
Impairment 687 8 −0.17 692 8 −0.01 691 8 −0.05 519 6 −0.09 465 4 −0.16 664 8 0.1
Internalizing symptoms 1,203 11 −0.09 1,227 12 −0.07 1,215 11 −0.08 1,010 10 −0.11 620 6 −0.16 588 6 −0.25
IQ 949 8 0.19 953 8 0.01 951 8 0.00 942 8 0.00 589 5 0.00 387 4 0.00
Medication use at baseline 1,738 16 −0.01 1,827 16 0.05 1,861 17 0.16 1,466 13 −0.07 955 10 −0.16 645 7 −0.03
ODD symptoms 1,981 17 −0.06 1,987 17 −0.06 1,989 17 −0.03 1,798 16 −0.08 1,079 10 −0.05 615 7 −0.24
SES high vs low 1,386 15 −0.02 1,342 14 −0.01 1,399 15 0.03 1,134 12 0.28 817 9 0.56 595 6 0.10
Sex 2,109 20 −0.55 2,173 20 −0.42 2,233 21 0.09 1,798 16 0.01 1,229 12 0.17 664 8 0.42
Single parent (yes) 1,306 10 −0.20 1,309 10 −0.27 1,314 10 −0.12 1,195 9 −0.29 568 4 −0.13 539 5 0.11

Note: Boldface type indicates significant findings multiple testing correction. All linear multilevel analysis models have post-measurement score as
outcome and are corrected for baseline impairment score. Every moderator was tested in an individual model. Full information on all models can be
found in Supplement 5 (Tables S9−S14, available online); correlation matrix between all moderators can be found in Table S27, available online.
ADHD = attention-deficit/hyperactivity disorder, CD = conduct disorder; df = degrees of freedom; k = number of studies; n = number of individuals;
ODD = oppositional defiant disorder; SE = standard error; SES = socioeconomic status.

(b = .64, p < .001) (Table 2). None of the investigated ODD Symptoms
variables moderated the intervention effect (see Table 2 and Behavioral interventions had a significant positive, small ES
Table S10, available online). on symptoms of ODD (b = 0.19, p < .001), with low
clustering within studies (ICC = 0.02) (Table 3). This effect
Hyperactivity/Impulsivity Symptoms remained significant in parenting interventions (b = 0.20,
Behavioral interventions had a positive, small effect on HI p < .001); the number of studies was too low for analysis in
symptom severity (b = 0.27, p < .001), with low clustering child/adolescent-focused studies. Youths with more baseline
within studies (ICC = 0.01) (Table 3). This effect remained CD symptoms (b = 0.21, p < .001; also see Figure 2E)
significant in parenting interventions (b = 0.31, p<.001) and youths with single parents (Figure 2G and H) (b =
and child/adolescent focused interventions (b = 0.35, p < 0.29, p = .006) showed a larger intervention effect.
.001) (Table 2). CD symptoms moderated intervention effect Figures 3 shows that both moderating effects were driven
(b = 0.14, p = .002); Figure 2D shows that youths with by a deterioration of those with high CD symptoms or sin-
higher CD symptom severity in the control group increased gle parents in the control group.
more in HI symptom severity compared to youths with fewer Additional analyses (see Supplement 8, available
CD symptoms, or youths in the intervention group. online), showed that the moderating effect of CD symptom
Additional analyses (see Figure S3, available online), severity and/or single parenthood did not vary as a function
showed a significant 3-way interaction among age, CD of any of the other variables. The moderating effect of base-
severity, and intervention condition (p = .004); a median line CD severity and single parenthood remained significant
split (age 8.6 years) showed that younger children with in studies on parenting interventions only (b = 0.21, p <
higher CD symptom severity in the control group had a .001; b = 0.31, p = .004 respectively).
worse outcome compared to younger children in the inter-
vention group on HI symptom severity, but this effect was CD Symptoms
not significant in older children. The effect of CD symptom Behavioral interventions had a positive, small ES on CD
severity remained significant in parenting interventions only symptom severity (b = 0.20, p < .001). The proportion of
(b = 0.20, p < .001), but not in studies on child/adoles- variance accounted for by clustering within studies was low
cent-focused interventions (b = 0.00, p = .99). (ICC < 0.001). This effect remained significant in
Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 153
Volume 61 / Number 2 / February 2022
GROENMAN et al.

parenting interventions (b = 0.23, p < .001) (Table 2); the condition performed worse than those in a TAU/active con-
number of studies was too low for analysis in child-focused trol group (b = 0.27, p = .009).
studies. Youths with higher baseline CD symptom severity
showed a larger intervention effect (b = 0.21, p < .001),
although this seems to be driven by a larger deterioration of DISCUSSION
those with high CD symptoms in the control group The present study is the first IPDMA on behavioral inter-
(Figure 2F). ventions for children and adolescents with ADHD. Its large
Additional analyses showed a significant 3-way interac- database, including data from 21 randomized controlled
tion among baseline ADHD severity, baseline CD severity, studies and behavioral treatment outcomes of 2,233 chil-
and intervention condition (p = .02); those with high dren and adolescents with ADHD, enabled moderator anal-
ADHD and high CD symptom severity, but not those with yses of sufficient statistical power. Results showed robust
low ADHD symptoms and high CD symptoms, showed evidence that behavioral interventions reduced ADHD
worse outcome post-intervention in the control group (see symptoms, behavioral problems, and global impairment
Figure S3, available online) The effect of baseline CD symp- according to reports of raters most proximal to the delivery
tom severity remained significant in studies on parenting of the intervention, with small- to medium-sized effects.
interventions only (b = 0.29, p < .001). The intervention effect size in this IPDMA was highest for
impairment, which, for most interventions, is the primary
treatment target.5,49 Moreover, on total ADHD symptom
Global Impairment
severity reduction, we found a seemingly larger effect size
Behavioral interventions had a positive, medium-sized effect
compared to those of aggregated data meta-analyses (eg,
on global impairment (b = 0.59, p < .001). The propor-
Boyer et al50). Our analyses clearly showed the added value
tion of variance accounted for by clustering
of IPDMA compared to other designs, as we uncovered
within studies was low (ICC < 0.001) (Table 3). This effect
moderators that could not be identified in earlier meta-anal-
remained significant in parenting interventions (b = 0.68,
yses or randomized controlled trials. CD (baseline symp-
p < .001); the number of studies was too low for analysis in
toms as well as comorbid diagnosis) moderated treatment
child-focused studies. Individuals with higher baseline
effects. For all outcome measures except inattention and
ADHD symptom severity showed smaller intervention
global impairment, higher baseline CD symptomatology
effects (b = 0.50, p < .001) than those with low baseline
(and/or a diagnosis of CD) was associated with larger treat-
ADHD symptoms, although this effect seems to be driven
ment effects. For HI symptoms, this moderating effect was
by a larger deterioration in impairment in the control group
found only in younger individuals (<8.6 years of age, using
(Figure 2I). Additional analyses (see Supplement 8, available
a median split). For CD, the moderating effect was present
online), showed that the moderating effect of ADHD symp-
only in those with more severe (z scores >0) baseline
tom severity did not vary as a function of any of the other
ADHD symptom severity.
variables (including control condition). The effect of base-
The larger benefit of behavioral interventions in chil-
line CD symptom severity remained significant in studies
dren with more severe CD may seem consistent with some
on parenting interventions only (b = 0.51, p < .001).
of the existing literature on intervention effects for children
with conduct problems; however, most studies did not
Additional Analysis Regarding Age and Control report whether effects were driven by the intervention or
Condition control condition51 Our results clearly show that positive
Age as a dimensional variable did not moderate intervention intervention effects in children with elevated CD symptoms
outcome on any of the outcome variables (see Supplement 8, were driven by larger symptom deterioration in the control
available online). condition (mostly TAU or WL). These findings suggest
Compared to TAU/active controls, the effect of behav- that youths with more severe CD symptoms are more likely
ioral intervention on the outcome variables was smaller to suffer an increase in symptoms of ADHD and behavioral
compared to those in a WL condition on symptoms of problems over time, particularly when not treated, empha-
ADHD (b = 0.47, p < .001), inattention (b = 0.66, p < sizing the importance of direct access to care. A similar mod-
.001), HI (b = 0.25, p = .002), ODD (b = 0.28, p < .001), erating effect was found for ADHD symptoms on global
and CD (b = 0.35, p < .001), but not impairment impairment. In addition, this group did not improve (and
(b = 0.06, p = .79). Of the previously identified moderators, even seems to deteriorate for HI and CD outcomes) from
the control condition affected only the relation between pre- to post-assessment in the active treatment condition,
CD symptoms on CD symptoms; individuals in the WL suggesting a protective rather than an ameliorative effect of
154 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry
Volume 61 / Number 2 / February 2022
 BEHAVIORAL INTERVENTIONS FOR ADHD

behavioral interventions for those with more severe CD Major strengths of our IPDMA are that we rigorously
symptoms or ADHD symptoms. Perhaps treatments for analyzed a large database, thereby covering a range of out-
this subgroup of children with severe behavioral problems comes and potential moderators. Although we included just
should be more intensive than routine ADHD interventions over one-third of eligible studies, our database was represen-
and enhanced with CD-specific components.52 tative of all eligible behavioral intervention studies, given
With respect to the moderation effect of single parent- the similar ESs reported in included and not included stud-
hood, a similar pattern emerged. The data suggest that ies. Moreover, as in IPDMA the number of participants is
youth from single-parent families deteriorated with regard of importance, our IPDMA did include over one-half of the
to ODD symptoms in the control condition, indicating total number of participants of the eligible studies. Most of
that children and adolescents from single-parent families the missing studies were older studies for which the data
should also be treated immediately. Moreover, this particu- were no longer available.
lar subgroup did not improve substantially in terms of Limitations of our IPDMA include, first, in line with
ODD symptoms in the behavioral treatment condition, other meta-analyses11,50 that we examined, a heterogeneous
again indicating a protective rather than an ameliorative group of behavioral interventions (eg, parent training, class-
effect of behavioral interventions. Treatments that are tai- room interventions, or skills training) that targeted different
lored to the specific needs of single-parent families may outcomes, whereas intervention effects and moderators of
therefore be more suitable for this subgroup. For example, a outcome may differ between intervention types.5 Future
study by Chacko et al.53 in families of children with meta-analyses, preferably using IPD, should focus on differ-
ADHD showed that children’s ODD symptoms reduced ent intervention types. Nevertheless, robust effects of behav-
significantly only after a behavioral parent training program ioral interventions on ADHD, behavioral problems, and
that was specifically designed for single mothers, whereas impairment were obtained, and effects remained significant
this effect was not present in children whose mothers in parenting interventions only as well as child/adolescent-
received standard parent training. focused interventions only. Second, as cannot be avoided in
Other potential moderators were not significant, sug- an IPDMA, several assumptions had to be made in the pro-
gesting equal effectiveness of behavioral interventions across cess of data harmonization, for example, that different meas-
children’s age, IQ, sex, medication status, and child’s ures from different trials actually measured the same
comorbidity with ODD, anxiety, and depression, as well as construct. However, we did find low clustering of outcomes
socioeconomic status and impairment severity. For age, within studies (ICCs < 0.07), which is indicative of low
additional analyses showed no differential effect of behav- heterogeneity of constructs across studies. A third limitation
ioral interventions on any of the outcome measures in indi- concerns the blinded measures. The number of studies
viduals less than 6 years of age, between 6 and 12 years, and using objective measures of symptom change or impairment
more than 12 years. The lack of a moderating effect of age that met inclusion criteria was very limited. We only
or socioeconomic status is in line with the only currently included 4 parent training studies that used objective obser-
available IPDMA on effects of 1 specific parenting program vational measures. In addition, the measures used were
for behavioral problems (no specific ADHD sample) in diverse (from structured play observations to audio-tapes of
young children (<12 years),54 suggesting equal effectiveness problem situations), and not all were well validated, reflect-
of behavioral treatments in ADHD across age groups and ing the absence of appropriate objective measures for behav-
socioeconomic status. ioral interventions, which should be developed in future
In addition, we found evidence for a specific waitlist studies. These should not only include behavioral measures
effect,55 as we found that children in the waitlist control but impairment measures as well. Previous meta-analyses
group performed worse on almost all outcomes than those in have shown that effects of behavioral interventions on
the TAU/active control condition. 56,57 Given the negative ADHD symptoms assessed by blinded raters were not sig-
effects of waiting for behavioral intervention, efforts should nificant.50 Thus, it remains unclear whether intervention
be made to make behavioral interventions easily and quickly effects reflect actual changes in ADHD symptoms or are
accessible to all, for example, by integrating them into regular limited to the perceptions of parents or teachers. However,
services and daily lives of parents and children (eg, in schools, such changes in perceptions may, of course, be important
primary care practice, or community services) and to reduce psychological indicators of the rater’s beliefs about child
waitlists to a minimum, especially for those with severe behavior (even if not accompanied by actual changes in core
behavioral problems. Telehealth and digital approaches may symptoms). It could also be argued that intervention recipi-
also be of value in increasing access and uptake of behavioral ents themselves (parents or teachers in the instance of chil-
interventions, 58,59 and moreover could be cost-effective. dren) are the best raters of the child’s problems.60
Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 155
Volume 61 / Number 2 / February 2022
GROENMAN et al.

Furthermore, parent- and teacher-reported outcomes are

better translatable to clinic-based practices, as they are Accepted February 19, 2021.
nearly always the only available information source in clini- Drs. Groenman, Steenhuis, Ms. Hornstra, Prof. Drs. Hoekstra and van den
cal care. Another limitation is that we included only peer- Hoofdakker are with the Center for Child and Adolescent Psychiatry, Univer-
sity of Groningen, University Medical Center, Groningen, The Netherlands.
reviewed published work, which may have been biased by a Dr. Groenman is also with the Dutch Autism and ADHD Research Center, Uni-
possible exaggerated estimate of intervention effect, 61 versity of Amsterdam, The Netherlands. Prof. Dr. Reijneveld is with University
of Groningen, University Medical Center Groningen, Groningen, The Nether-
although moderator analyses on individual data may be less lands. Dr. Aghebati is with the School of Behavioral Sciences and Mental
influenced by this bias. Moreover, we did not have permis- Health, Tehran Institute of Psychiatry, Iran University of Medical Sciences, Teh-
ran, Iran. Dr. Boyer is with the University of Amsterdam, The Netherlands. Prof.
sion, which is required under the European General Data Dr. Buitelaar is with the Donders Institute for Brain, Cognition and Behaviour,
Radboud University Medical Centre, Nijmegen, The Netherlands, and Kar-
Protection Regulation (GDPR), to include ethnicity in our akter Child and Adolescent Psychiatry University Centre, Nijmegen, The Neth-
moderator analyses. Also, many studies (in fact, all of the erlands. Prof. Dr. Chronis-Tuscano is with the University of Maryland, College
Park. Prof. Dr. Daley is with the School of Medicine, Institute of Mental Health,
European studies except for 1 older study) did not collect University of Nottingham, United Kingdom. Ms. Dehkordian is with Shahid
data on ethnicity. The majority of the American studies Chamran University, Ahvaz, Iran. Drs. Dvorsky and Pfiffner are with the Univer-
sity of California San Francisco. Dr. Franke is with the Liggins Institute, Univer-
reported on ethnicity; it appeared that most of the included sity of Auckland, New Zealand. Dr. DuPaul is with Lehigh University,
populations were a reasonable representation of the United Bethlehem, Pennsylvania. Dr. Gershy is with the School of Education, Israel
and Schneider Children’s Medical Center, The Hebrew University of Jerusa-
States. Finally, the systematic search of the current IPDMA lem, Israel. Dr. Harvey is with the University of Massachusetts Amherst. Dr.
was 3 years old. Preparing and gathering the data for Hennig is with Universit€at Hamburg, Germany. Dr. Herbert is with the Child-
ren’s Hospital of Orange County, California. Prof. Dr. Langberg is with Virginia
IPDMA takes time (in the instance of this study, more than Commonwealth University, Richmond. Drs. Mautone and Power are with
2 years). However, with a recent systematic search, using Children’s Hospital of Philadelphia, Pennsylvania. Dr. Amori Yee Mikami is
with the University of British Columbia, Vancouver, Canada. Prof. Dr. Schramm
the same criteria and methods as the current study, we iden- is with Inclusive Education, University of Potsdam, Germany. Dr. Schweitzer is
tified 6 datasets that were not included in the current analy- with the University of California, Davis. Dr. Sibley is with the University of
Washington School of Medicine, Seattle Children’s Research Institute. Prof.
ses. Unfortunately, as collection, preparation, and Sonuga-Barke is with King’s College London, United Kingdom. Ms. C.
harmonizing of those data would have taken substantial Thompson and Prof. Dr. M. Thompson are with the School of Psychology, Uni-
versity of Southampton, United Kingdom. Prof. Emeritus Webster-Stratton is
time, we could not include these in the current study. with the University of Washington, Seattle. Dr. Xie is with the Specialty Mental
Future IPDMA studies should focus on parenting measures, Health Program of Asian Health Service, Oakland, California. Dr. Luman is
with Vrije Universiteit Amsterdam, Amsterdam, The Netherlands, and Bascule,
long-term outcomes, mediators of treatment effects (eg, Academic Centre for Child and Adolescent Psychiatry, Amsterdam, The Neth-
erlands. Prof. Dr. van der Oord is with Clinical Psychology, KU Leuven, Leuven,
increase in parenting skills), and moderators of behavioral Belgium; and the University of Amsterdam, Amsterdam, the Netherlands.
treatments in relation to other treatments for ADHD, such This research was funded by the Dutch Organization for Health Research and
as medication. Future studies also need to address how to Development (ZonMw) under grant number 729300013 to Barbara J. van den
Hoofdakker. The funder had no role in the design of this protocol, the collec-
sequence and combine behavioral and cognitive treatment tion of data, the data analysis, or the interpretation or publication of the study
to medication and other interventions, to reduce results.
impairment and improve functioning of children and ado- This work has been prospectively registered: https://www.crd.york.ac.uk/pros
lescents with ADHD.62 pero/display_record.php?RecordID=69877.

This IPDMA showed that behavioral interventions for Jos Twisk, PhD, of the Vrije Universteit Amsterdam, served as the statistical
expert for this research.
ADHD are effective treatments, significantly reducing core
Author Contributions
ADHD symptoms, associated behavioral problems, and Conceptualization: Groenman, Hoekstra, Steenhuis, Buitelaar, Reijneveld,
global impairment, as perceived by parents or teachers post- Luman, van der Oord, van den Hoofdakker
Data curation: Groenman, Hornstra, Aghebati, Boyer, Chronis-Tuscano,
treatment. Improvement in impairment is notable, as Daley, Dehkordian, Dvorsky, Franke, DuPaul, Gershy, Harvey, Hennig, Her-
bert, Langberg, Mautone, Mikami, Pfiffner, Power, Schramm, Sibley, Sonuga-
impairment in functioning is generally what prompts treat- Barke, C. Thompson, M. Thompson, Webster-Stratton, Xie, Luman, van der
ment seeking and is arguably the most salient outcome in Oord, van den Hoofdakker
Formal analysis: Groenman, van den Hoofdakker
behavioral treatment. We found significant moderators of Funding acquisition: Buitelaar, Reijneveld, Luman, van der Oord, van den
outcome that can be translated to a clear clinical message, Hoofdakker
Investigation: Groenman, Luman, van der Oord, van den Hoofdakker
although these moderators seemed to have the strongest Methodology: Groenman, Hoekstra, Luman, van der Oord, van den Hoofdak-
impact on the control condition rather than the active treat- ker
Project administration: Groenman
ment condition; in particular, children with high levels of Resources: Hoekstra
CD symptoms or diagnosis, children with single parents, Supervision: Luman, van der Oord, van den Hoofdakker
Visualization: Groenman, Hornstra
and children with high levels of ADHD symptoms, should Writing − original draft: Groenman, Hornstra, Luman, van der Oord, van den
be prioritized for behavioral treatment. For these children, Hoofdakker
Writing − review and editing: Groenman, Hornstra, Hoekstra, Steenhuis,
treatment seems to prevent them from further deterioration Aghebati, Boyer, Buitelaar, Chronis-Tuscano, Daley, Dehkordian, Dvorsky,
in terms of ADHD symptoms, behavioral problems, and/or Franke, DuPaul, Gershy, Harvey, Hennig, Herbert, Langberg, Mautone,
Mikami, Pfiffner, Power, Reijneveld, Schramm, Schweitzer, Sibley, Sonuga-
impairment.
156 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry
Volume 61 / Number 2 / February 2022
 BEHAVIORAL INTERVENTIONS FOR ADHD

Barke, C. Thompson, M. Thompson, Webster-Stratton, Xie, Luman, van der the NFPP, data from trials has been included in this paper. She has received
Oord, van den Hoofdakker. grants from the National Institute for Health Research (NIHR) to develop and
evaluate this work. She has received training fees from training therapists in
The authors would like to thank Jos Twisk, PhD, of the Vrije Universteit the program. Prof. Emer. Webster-Stratton has disclosed a potential conflict
Amsterdam, for his advice regarding the analyses. Furthermore, the authors of interest due to the fact she is developer of one of the intervention pro-
want to thank their research assistant Lieke Bruinsma, MSc, of the University grams (Incredible Years). Because she has provided training and instructional
of Groningen, University Medical Center Groningen, for her help in all types materials for these treatment programs, she stands to gain financially from a
of administrative tasks that arose during this IPDMA. positive review. This interest has been disclosed to the university and has
Disclosure: Prof. Dr. Hoekstra has attended a paid advisory board meeting of been managed consistent with federal and university policy regarding data
Shire (a Takeda Pharmaceutical Company). Dr. Boyer has co-developed Plan management. Dr. Luman was involved in developing Positivity and Rules, a
My Life and Solution Focused Treatment and has received royalties for both behavioral self-help teacher training, but has no financial interests. Prof. Dr.
treatments. Prof. Dr. Buitelaar has been a consultant to / member of advisory van der Oord has co-developed Plan My Life and Solution Focused Treat-
board of / and/or speaker for Takeda/Shire, Roche, Medice, Angelini, Jans- ments and other behavioral treatments but has reported no financial interest
sen, and Servier. He is not an employee of any of these companies and not a in any of these. Prof. Dr. van den Hoofdakker has received research grants
stock shareholder of any of these companies. He has no other financial or from ZonMw (The Netherlands Organisation for Health Research and Devel-
material support, including expert testimony, patents, royalties in the past opment), NWO (The Netherlands Organisation for Scientific Research), and
three years. Prof. Dr. Daley has reported grants, personal fees, and non-finan- UMCG (University Medical Centre Groningen) and royalties as one of the edi-
cial support from Shire/Takeda; personal fees and non-financial support from tors of “Sociaal Onhandig” (published by Van Gorcum), a Dutch book for
Medice and Eli Lilly and Co.; non-financial support from Qbtech; and book parents of children with ADHD or PDD-NOS that is being used in parent train-
royalties from Jessica Kingsley from the self-help version of the New Forest ing. She is and has been involved in the development and evaluation of sev-
Parenting Programme (NFPP). Drs. Harvey and Herbert have developed the eral Dutch parent training programs, without financial interests; she is and has
Parenting Hyperactive Preschoolers program and have received royalties been a member of Dutch ADHD guideline groups and an advisor of the
from the sale of the clinician workbook. Dr. Hennig has reported being one of Dutch Knowledge Centre for Child and Adolescent Psychiatry. Ms. Thompson
the developers of the Lerntraining f€ur Jugendliche mit ADHS (LeJA) and has has received funding from NIHR to develop and evaluate an online version of
received royalties from sales of the treatment manual. Prof. Dr. Langberg has the NFPP. She has received personal fees for helping with the training of the
developed the Homework, Organization, and Planning Skills (HOPS) interven- NFPP. Drs. Groenman, Steenhuis, Aghebati, Prof. Dr. Chronis-Tuscano, Drs.
tion and has received royalties from sales of the treatment manuals. Dr. Power Dvorsky, Franke, DuPaul, Gershy, Mautone, Mikami, Pfiffner, Prof. Dr. Reijne-
has received royalties from Guilford Press for a book, Homework Success for veld, Drs. Schweitzer and Xie, and Mss. Hornstra and Dehkordian have
Children with ADHD, that describes many components incorporated into the reported no biomedical financial interests or potential conflicts of interest.
Family-School Success investigations included in this study. Prof. Dr. Correspondence to Annabeth P. Groenman, PhD, Department of Child and
Schramm has reported being one of the developers of the Lerntraining f€ ur Adolescent Psychiatry, Freepost 176, 9700 VB, Groningen, the Netherlands;
Jugendliche mit ADHS (LeJA) and has received royalties from sales of the e-mail: a.groenman@gmail.com
treatment manual. Dr. Sibley has received royalties from Guilford Press and
Vimeo Inc. for intervention materials related to Supporting Teens’ Autonomy 0890-8567/$36.00/© 2021 The Authors. Published by Elsevier Inc. on behalf of
Daily (STAND). She has received consultancy fees from Takeda Pharmaceuti- American Academy of Child & Adolescent Psychiatry. This is an open access
cals. Prof. Sonuga-Barke was involved in the development of the NFPP for article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)
which he has received royalties. He has received consultancy, speaker fees,
https://doi.org/10.1016/j.jaac.2021.02.024
and conference attendance support variously from Shire, Neurotech Solu-
tions, and Qbtech. Prof. Dr. Thompson has reported being a co-developer of

REFERENCES 11. Daley D, van der Oord S, Ferrin M, et al. Behavioral interventions in attention-deficit/
hyperactivity disorder: a meta-analysis of randomized controlled trials across multiple out-
1. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders.
come domains. J Am Acad Child Adolesc Psychiatry. 2014;53:835-847. https://doi.org/
5th ed. Washington, DC: American Psychiatric Publishing; 2013.
10.1016/j.jaac.2014.05.013.
2. Polanczyk GV, Salum GA, Sugaya LS, Caye A, Rohde LA. Annual research review: a
12. Kraemer HC, Wilson GT, Fairburn CG, Agras WS. Mediators and moderators of treat-
meta-analysis of the worldwide prevalence of mental disorders in children and adoles-
ment effects in randomized clinical trials. Arch Gen Psychiatry. 2002;59:877-883.
cents. J Child Psychol Psychiatry. 2015;56:345-365. https://doi.org/10.1111/
https://doi.org/10.1001/archpsyc.59.10.877.
jcpp.12381.
13. Fabiano GA, Schatz NK, Aloe AM, Chacko A. Chronis-Tuscano A. A systematic review
3. Kuriyan AB, Pelham WE, Molina BS, et al. Young adult educational and vocational out-
of meta-analyses of psychosocial treatment for attention-deficit/hyperactivity disorder.
comes of children diagnosed with ADHD. J Abnorm Child Psychol. 2013;41:27-41.
Clin Child Family Psychol Rev. 2015;18:77-97. https://doi.org/10.1007/s10567-015-
https://doi.org/10.1007/s10802-012-9658-z.
0178-6.
4. Yoshimasu K, Barbaresi WJ, Colligan RC, et al. Childhood ADHD is strongly associated
14. Lee PC, Niew WI, Yang HJ, Chen VC, Lin KC. A meta-analysis of behavioral parent
with a broad range of psychiatric disorders during adolescence: a population-based birth
training for children with attention deficit hyperactivity disorder. Res Dev Disabil.
cohort study. J Child Psychol Psychiatry. 2012;53:1036-1043. https://doi.org/10.1111/
2012;33:2040-2049. https://doi.org/10.1016/j.ridd.2012.05.011.
j.1469-7610.2012.02567.x.
15. Mulqueen JM, Bartley CA, Bloch MH. Meta-analysis: parental interventions for pre-
5. Evans SW, Owens JS, Wymbs BT, Ray AR. Evidence-based psychosocial treatments for
school ADHD. J Atten Disord. 2015;19:118-124. https://doi.org/10.1177/
children and adolescents with attention deficit/hyperactivity disorder. J Clin Child Ado-
1087054713504135.
lesc Psychol. 2018;47:157-198. https://doi.org/10.1080/15374416.2017.1390757.
16. Hinshaw SP. Moderators and mediators of treatment outcome for youth with ADHD:
6. Evans SW, Owens JS, Bunford N. Evidence-based psychosocial treatments for children
understanding for whom and how interventions work. J Pediatr Psychology.
and adolescents with attention-deficit/hyperactivity disorder. J Clin Child Adolesc Psy-
2007;32:664-675. https://doi.org/10.1093/jpepsy/jsl055.
chol. 2014;43:527-551. https://doi.org/10.1080/15374416.2017.1390757.
17. Group MTA. A 14-month randomized clinical trial of treatment strategies for attention-
7. Faraone SV, Asherson P, Banaschewski T, et al. Attention-deficit/hyperactivity disorder.
deficit/hyperactivity disorder. The MTA Cooperative Group. Multimodal Treatment
Nat Rev Dis Primers. 2015;1:15020. https://doi.org/10.1038/nrdp.2015.20.
Study of Children with ADHD. Arch Gen Psychiatry. 1999;56:1073-1086. https://doi.
8. Schatz NK, Fabiano GA, Cunningham CE, et al. Systematic review of patients' and
org/10.1001/archpsyc.56.12.1073.
parents' preferences for adhd treatment options and processes of care. The Patient.
18. March JS, Swanson JM, Arnold LE, et al. Anxiety as a predictor and outcome variable in
2015;8:483-497. https://doi.org/10.1007/s40271-015-0112-5.
the Multimodal Treatment Study of Children with ADHD (MTA). J Abnorm Child
9. Daley D, Van Der Oord S, Ferrin M, et al. Practitioner review: current best practice in
Psychology. 2000;28:527-541. https://doi.org/10.1023/A:1005179014321.
the use of parent training and other behavioural interventions in the treatment of children
19. Riley RD, Lambert PC, Abo-Zaid G. Meta-analysis of individual participant data: ratio-
and adolescents with attention deficit hyperactivity disorder. J Child Psychol Psychiatry
nale, conduct, and reporting. BMJ (Clin Res Ed). 2010;340:c221. https://doi.org/
Allied Discip. 2018;59:932-947. https://doi.org/10.1111/jcpp.12825.
10.1136/bmj.c221.
10. Coles EK, Pelham Iii WE, Fabiano GA, et al. Randomized trial of first-line behav-
20. Furukawa TA, Efthimiou O, Weitz ES, et al. Cognitive-behavioral analysis system of psy-
ioral intervention to reduce need for medication in children with ADHD. J Clin
chotherapy, drug, or their combination for persistent depressive disorder: personalizing
Child Adolesc Psychol Aug 14 20191-15. https://doi.org/10.1080/15374416.2019.
the treatment choice using individual participant data network metaregression. Psy-
1630835.
chother Psychosom. 2018;87:140-153. https://doi.org/10.1159/000489227.

Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 157
Volume 61 / Number 2 / February 2022
GROENMAN et al.

21. Ouzzani M, Hammady H, Fedorowicz Z, Elmagarmid A. Rayyan−a Web and mobile 42. Schramm SA, Hennig T, Linderkamp F. Training problem solving and organizational skills
app for systematic reviews. Syst Rev. 2016;5:210. https://doi.org/10.1186/s13643-016- in adolescents with attention-deficit/hyperactivity disorder: a randomized controlled trial. J
0384-4. Cognit Educa Psychol. 2016;15:391-411. https://doi.org/10.1891/1945-8959.15.3.391.
22. Team RStudio. RStudio: integrated development for R. http://www.rstudio.com. 43. Sibley MH, Pelham WE, Derefinko KJ, Kuriyan AB, Sanchez F, Graziano PA. A pilot
23. Cohen J. Statistical Power Analysis for the Behavioral Sciences. London: Routledge; trial of Supporting Teens’ Academic Needs Daily (STAND): a parent-adolescent collabo-
2013. rative intervention for ADHD. J Psychopathol Behav Assess. 2013;35:436-449. https://
24. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful doi.org/10.1007/s10862-013-9353-6.
approach to multiple testing. J R Stat Soc Series B (Methodol). 1995;57:289-300. 44. Sibley MH, Graziano PA, Kuriyan AB, et al. Parent−teen behavior therapy + motiva-
25. Langberg JM, Dvorsky MR, Molitor SJ, et al. Overcoming the research-to-practice gap: a tional interviewing for adolescents with ADHD. J Consult Clin Psychol. 2016;84:699.
randomized trial with two brief homework and organization interventions for students https://doi.org/10.1037/ccp0000106.
with ADHD as implemented by school mental health providers. J Consult Clin Psychol. 45. Thompson MJ, Laver-Bradbury C, Ayres M, et al. A small-scale randomized controlled
2018;86:39. https://doi.org/10.1037/ccp0000265. trial of the revised New Forest Parenting Programme for preschoolers with attention defi-
26. Sonuga-Barke EJ, Barton J, Daley D, et al. A comparison of the clinical effectiveness and cost of cit hyperactivity disorder. Eur Child Adolesc Psychiatry. 2009;18:605-616. https://doi.
specialised individually delivered parent training for preschool attention-deficit/hyperactivity dis- org/10.1007/s00787-009-0020-0.
order and a generic, group-based programme: a multi-centre, randomised controlled trial of the 46. Van Den Hoofdakker BJ, Van der Veen-Mulders L, Sytema S, Emmelkamp PM, Minderaa
New Forest Parenting Programme versus Incredible Years. Eur Child Adolesc Psychiatry. RB, Nauta MH. Effectiveness of behavioral parent training for children with ADHD in rou-
2018;27:797-809. https://doi.org/10.1007/s00787-017-1054-3. tine clinical practice: a randomized controlled study. J Am Acad Child Adolesc Psychiatry.
27. Aghebati A, Gharraee B, Shoshtari MH, Gohari MR. Triple P-Positive Parenting Pro- 2007;46:1263-1271. https://doi.org/10.1097/chi.0b013e3181354bc2.
gram for Mothers of ADHD Children. Iran J Psychiatry Behav Sci. 2014;8:59. 47. Webster-Stratton CH, Reid MJ, Beauchaine T. Combining parent and child training for
28. Boyer BE, Geurts HM, Prins PJM. Van der Oord S. Two novel CBTs for adolescents young children with ADHD. J Clin Child Adolesc Psychol. 2011;40:191-203. https://
with ADHD: the value of planning skills. Eur Child Adolesc Psychiatry. 2015;24:1075- doi.org/10.1080/15374416.2011.546044.
1090. https://doi.org/10.1007/s00787-014-0661-5. 48. Xie Y, Dixon JF, Yee OM, et al. A study on the effectiveness of videoconferencing on
29. Chronis-Tuscano A, Clarke TL, O'Brien KA, et al. Development and preliminary evalua- teaching parent training skills to parents of children with ADHD. Telemed e-Health.
tion of an integrated treatment targeting parenting and depressive symptoms in mothers 2013;19:192-199. 0.1089/tmj.2012.0108.
of children with attention-deficit/hyperactivity disorder. J Consult Clin Psychol. 49. Chronis-Tuscano A, Chacko A, Barkley R. Key issues relevant to the efficacy of behav-
2013;81:918. https://doi.org/10.1037/a0032112. ioral treatment for ADHD. Am J Psychiatry. 2013;170. https://doi.org/10.1176/appi.
30. Daley D, O’Brien M. A small-scale randomized controlled trial of the self-help version of the ajp.2013.13030293. 799-799.
New Forest Parent Training Programme for children with ADHD symptoms. Eur Child Ado- 50. Sonuga-Barke EJ, Brandeis D, Cortese S, et al. Nonpharmacological interventions for
lesc Psychiatry. 2013;22:543-552. https://doi.org/10.1007/s00787-013-0396-8. ADHD: systematic review and meta-analyses of randomized controlled trials of dietary
31. DuPaul GJ, Jitendra AK, Volpe RJ, et al. Consultation-based academic interventions and psychological treatments. Am J Psychiatry. 2013;170:275-289. https://doi.org/
for children with ADHD: effects on reading and mathematics achievement. J Abnorm 10.1176/appi.ajp.2012.12070991.
Child Psychology. 2006;34:635-648. https://doi.org/10.1007/s10802-006-9046-7. 51. Shelleby EC, Shaw DS. Outcomes of parenting interventions for child conduct problems:
2006-1-1. a review of differential effectiveness. Child Psychiatry Hum Dev. 2014;45:628-645.
32. Franke N, Keown LJ, Sanders MR. An RCT of an online parenting program for parents https://doi.org/10.1007/s10578-013-0431-5.
of preschool-aged children with ADHD symptoms. J Atten Disord 20161716-1726. 52. Fairchild G, Hawes DJ, Frick PJ, et al. Conduct disorder. Nat Rev Dis Primers.
https://doi.org/10.1177/1087054716667598. 2019;5:1-25. https://doi.org/10.1038/s41572-019-0095-y.
33. Gershy N, Meehan KB, Omer H, Papouchis N, Sapir IS. Randomized Clinical Trial of 53. Chacko A, Wymbs BT, Wymbs FA, et al. Enhancing traditional behavioral parent train-
Mindfulness Skills Augmentation in Parent Training. New York: Springer; 2017:783-803. ing for single mothers of children with ADHD. J Clin Child Adolesc Psychol.
34. Herbert SD, Harvey EA, Roberts JL, Wichowski K, Lugo-Candelas CI. A randomized 2009;38:206-218. https://doi.org/10.1080/15374410802698388.
controlled trial of a parent training and emotion socialization program for families of 54. Gardner F, Leijten P, Harris V, et al. Equity effects of parenting interventions for child
hyperactive preschool-aged children. Behav Ther. 2013;44:302-316. https://doi.org/ conduct problems: a pan-European individual participant data meta-analysis. Lancet Psy-
10.1016/j.beth.2012.10.004. chiatry. 2019;6:518-527. https://doi.org/10.1016/S2215-0366(19)30162-2.
35. Mautone JA, Marshall SA, Sharman J, Eiraldi RB, Jawad AF, Power TJ. Development of 55. Furukawa T, Noma H, Caldwell D, et al. Waiting list may be a nocebo condition in psy-
a family school intervention for young children with attention deficit hyperactivity disor- chotherapy trials: a contribution from network meta-analysis. Acta Psychiatr Scand.
der. Sch Psychol Rev. 2012;41:447-466. 2014;130:181-192. https://doi.org/10.1159/000489227.
36. Mikami AY, Lerner MD, Griggs MS, McGrath A, Calhoun CD. Parental influence on 56. Dadds MR, Cauchi AJ, Wimalaweera S, Hawes DJ, Brennan J. Outcomes, moderators,
children with attention-deficit/hyperactivity disorder: II. Results of a pilot intervention and mediators of empathic-emotion recognition training for complex conduct problems
training parents as friendship coaches for children. J Abnorm Child Psychol. in childhood. Psychiatry Res. 2012;199:201-207. https://doi.org/10.1016/j.psy-
2010;38:737-749. https://doi.org/10.1007/s10802-010-9403-4. chres.2012.04.033.
37. Mikami AY, Griggs MS, Lerner MD, et al. A randomized trial of a classroom inter- 57. NICE. Antisocial behaviour and conduct disorders in children and young people: recog-
vention to increase peers' social inclusion of children with attention-deficit/hyperac- nition and management. https://www.nice.org.uk/guidance/cg158.
tivity disorder. J Consult Clin Psychol. 2013;81:100. https://doi.org/10.1037/ 58. Myers K, Vander Stoep A, Zhou C, McCarty CA, Katon W. Effectiveness of a
a0029654. telehealth service delivery model for treating attention-deficit/hyperactivity
38. Pfiffner LJ, Mikami AY, Huang-Pollock C, Easterlin B, Zalecki C, McBurnett K. A ran- disorder: a community-based randomized controlled trial. J Am Acad
domized, controlled trial of integrated home-school behavioral treatment for ADHD, Child Adolesc Psychiatry. 2015;54:263-274. https://doi.org/10.1016/j.
predominantly inattentive type. J Am Acad Child Adolesc Psychiatry. 2007;46:1041- jaac.2015.01.009.
1050. https://doi.org/10.1097/chi.0b013e318064675f. 59. Tse YJ, McCarty CA, Stoep AV, Myers KM. Teletherapy delivery of caregiver behavior
39. Pfiffner LJ, Hinshaw SP, Owens E, et al. A two-site randomized clinical trial of integrated training for children with attention-deficit hyperactivity disorder. Telemed e-Health..
psychosocial treatment for ADHD-inattentive type. J Consult Clin Psychology. 2015;21:451-458. https://doi.org/10.1089/tmj.2014.0132.
2014;82:1115. https://doi.org/10.1037/a0036887. 60. Cuijpers P. Targets and outcomes of psychotherapies for mental disorders: an overview.
40. Pfiffner LJ, Rooney M, Haack L, Villodas M, Delucchi K, McBurnett K. A randomized World Psychiatry. 2019;18:276-285. https://doi.org/10.1002/wps.20661.
controlled trial of a school-implemented school−home intervention for attention-deficit/ 61. McAuley L, Pham B, Tugwell P, Moher D. Does the inclusion of grey literature influ-
hyperactivity disorder symptoms and impairment. J Am Acad Child Adolesc Psychiatry. ence estimates of intervention effectiveness reported in meta-analyses? Lancet.
2016;55:762-770. https://doi.org/10.1016/j.jaac.2016.05.023. 2000;356:1228-1231. https://doi.org/10.1016/s0140-6736(00)02786-0.
41. Power TJ, Mautone JA, Soffer SL, et al. A family−school intervention for children with 62. Nahum-Shani I, Qian M, Almirall D, et al. Experimental design and primary data analy-
ADHD: results of a randomized clinical trial. J Consult Clin Psychology. 2012;80:611. sis methods for comparing adaptive interventions. Psychol Methods. 2012;17:457.
https://doi.org/10.1037/a0028188. https://doi.org/10.1037/a0029372.

158 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry
Volume 61 / Number 2 / February 2022