IntroductionBehavioral parent training is recommended as evidence-based intervention for children with attention-deficit/hyperactivity disorder (ADHD) in clinical practice guidelines (e.g.,1American Academy of Child and Adolescent Psychiatry
Practice parameters for the assessment and treatment of children and adolescents with Attention Deficit/Hyperactivity Disorder.,2

NICE. Attention Deficit Hyperactivity Disorder: Diagnosis and Management.; 2018.

). The presumed working mechanism of behavioral parent training is to alter the child’s behavior via parenting behaviors3

Barkley RA. Attention-Deficit Hyperactivity Disorder: A Handbook for Diagnosis and Treatment. Guilford Publications; 2014.

. Several meta-analyses demonstrated effects of parent training on several parental outcomes (see 4Fabiano G.A. Schatz N.K. Aloe A.M. Chacko A. Chronis-Tuscano A. A Systematic Review of Meta-Analyses of Psychosocial Treatment for Attention-Deficit/Hyperactivity Disorder. for a review of these meta-analyses), with largest and most robust effects on positive and negative parenting, smaller improvements in parenting sense of competence, but no effects on parental mental health5Daley 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 outcome domains.. Behavioral parent training consists of many different behavioral techniques that are being taught to parents, and it is yet unknown which of these techniques are most effective. In the current study, we investigated which behavioral techniques contribute to the effects of parent training on parental outcomes.Several studies show that improvements in parenting behaviors mediate improvements in the behavior of children with ADHD. There is consistent evidence that parent training for children with ADHD is less likely to improve the child’s behavior when negative parenting practices such as harsh and inconsistent discipline do not decrease6Chronis-Tuscano A. O’Brien K.A. Johnston C. et al.The relation between maternal ADHD symptoms & improvement in child behavior following brief behavioral parent training is mediated by change in negative parenting., 7Dose C. Hautmann C. Bürger M. Schürmann S. Döpfner M. Negative parenting behaviour as a mediator of the effects of telephone-assisted self-help for parents of pharmacologically treated children with attention-deficit/hyperactivity disorder., 8Hinshaw S.P. Owens E.B. Wells K.C. et al.Family processes and treatment outcome in the MTA: Negative/ineffective parenting practices in relation to multimodal treatment.. Furthermore, increased parenting sense of competence and decreased parenting stress after parent training has repeatedly been found to be associated with behavioral improvements of children with ADHD9Larsen L.B. Daley D. Lange A.-M. Sonuga-Barke E.J.S. Thomsen P.H. Rask C.U. Effect of Parent Training on Health-Related Quality of Life in Preschool Children With Attention-Deficit/Hyperactivity Disorder: A Secondary Analysis of Data From a Randomized Controlled Trial.,10Heath C.L. Curtis D.F. Fan W. McPherson R. The Association Between Parenting Stress, Parenting Self-Efficacy, and the Clinical Significance of Child ADHD Symptom Change Following Behavior Therapy..Although the effectiveness of parent training on parenting outcomes is recognized by all major clinical practice guidelines on ADHD1American Academy of Child and Adolescent Psychiatry
Practice parameters for the assessment and treatment of children and adolescents with Attention Deficit/Hyperactivity Disorder.,2

NICE. Attention Deficit Hyperactivity Disorder: Diagnosis and Management.; 2018.

,11Taylor E. Döpfner M. Sergeant J. et al.European clinical guidelines for hyperkinetic disorder - First upgrade., there is surprisingly little research about which specific behavioral techniques that are being used in parent training contribute (most) to changes in parenting behaviors. Some examples of behavioral techniques are restructuring situations, offering rewards, and time-out procedures12How to Improve Behavioral Parent and Teacher Training for Children with ADHD: Integrating Empirical Research on Learning and Motivation into Treatment.,13Fabiano G.A. Pelham W.E. Manos M.J. et al.An evaluation of three time-out procedures for children with attention-deficit/hyperactivity disorder.. Most behavioral parent training programs are rooted in social learning theories14Antshel K.M. Barkley R.A. Psychosocial Interventions in Attention Deficit Hyperactivity Disorder., and the overarching aim of these programs is teaching parents to prevent and manage their child’s behavior15Evans S.W. Owens J.S. Wymbs B.T. Ray A.R. Evidence-Based Psychosocial Treatments for Children and Adolescents With Attention Deficit/Hyperactivity Disorder.. The content of existing behavioral parent training programs, however, differs substantially: some mainly consist of psycho-education (e.g., Sunshine Program for ADHD16Ferrin M. Perez-Ayala V. El-Abd S. et al.A Randomized Controlled Trial Evaluating the Efficacy of a Psychoeducation Program for Families of Children and Adolescents With ADHD in the United Kingdom: Results After a 6-Month Follow-Up.,17Ferrin M. Moreno-Granados J.M. Salcedo-Marin M.D. Ruiz-Veguilla M. Perez-Ayala V. Taylor E. Evaluation of a psychoeducation programme for parents of children and adolescents with ADHD: Immediate and long-term effects using a blind randomized controlled trial.), whereas others emphasize disciplinary communication (e.g., Helping the Noncompliant Child18Abikoff H.B. Thompson M. Laver-Bradbury C. et al.Parent training for preschool ADHD: A randomized controlled trial of specialized and generic programs.), observation and monitoring (e.g., Family-School Success-Early Elementary19Mautone J.A. Marshall S.A. Sharman J. Eiraldi R.B. Jawad A.F. Power T.J. Development of a family-School intervention for young children with attention deficit hyperactivity disorder.), or positive reinforcement (e.g., Incredible Years20Sonuga-Barke E.J.S. Barton J. Daley D. et al.A comparison of the clinical effectiveness and cost of specialised individually delivered parent training for preschool attention-deficit/hyperactivity disorder and a generic, group-based programme: a multi-centre, randomised controlled trial of the New F.).Knowing which specific behavioral techniques are effective components of parent training is crucial as it could guide the development, improvement, and tailoring of interventions, and thereby benefits the well-being of children with ADHD and their families21DuPaul G.J. Evans S.W. Mautone J.A. Owens J.S. Power T.J. Future Directions for Psychosocial Interventions for Children and Adolescents with ADHD.. The current study is the first meta-analysis that disentangled the effects of these specific behavioral techniques of parent training for ADHD on parenting outcomes.Two earlier endeavors examining the effectiveness of specific behavioral techniques of parent training are worth mentioning. First, for children with disruptive behavior disorders, the use of positive reinforcement (particularly praise) and natural/logical consequences was associated with stronger effects on disruptive child behavior22Leijten P. Gardner F. Melendez-Torres G.J. et al.Meta-Analyses: Key Parenting Program Components for Disruptive Child Behavior.. Second, a meta-analysis including a broad range of parenting programs23Kaminski J.W. Valle L.A. Filene J.H. Boyle C.L. A meta-analytic review of components associated with parent training program effectiveness. suggested that techniques focusing on positive parent–child interactions and emotional communication skills, time-out, parenting consistency, and practicing skills during the sessions were associated with larger effectiveness, as measured by child and parenting outcomes. Despite the valuable contribution of these studies, their conclusions were limited for several reasons: (1) in Leijten et al.22Leijten P. Gardner F. Melendez-Torres G.J. et al.Meta-Analyses: Key Parenting Program Components for Disruptive Child Behavior., only child outcomes were assessed, whereas the presumed working mechanism of parent training is to change parenting behaviors; (2) both studies only scored the presence or absence of specific behavioral techniques, which ignores crucial and more nuanced information about the dosage of these techniques; (3) both studies only used information about the components of parent training that was provided in papers and supplements, which may be less detailed relative to information derived from treatment manuals; (4) both studies did not focus specifically on children with ADHD, and therefore the implications of these studies for children with ADHD are yet unclear, particularly as children with ADHD are suggested to have specific underlying instrumental learning deficits which potentially should be targeted with specific behavioral techniques12How to Improve Behavioral Parent and Teacher Training for Children with ADHD: Integrating Empirical Research on Learning and Motivation into Treatment.. Unique problems in intra-individual variability in responding (i.e., increased number of lapses of attention24Kofler M.J. Rapport M.D. Sarver D.E. et al.Reaction time variability in ADHD: A meta-analytic review of 319 studies.) as well as aberrant motivational processes are documented for children with ADHD relative to children with disruptive behavior disorders25Luman M. Sergeant J.A. Knol D.L. Oosterlaan J. Impaired decision making in oppositional defiant disorder related to altered psychophysiological responses to reinforcement., 26Salum G.A. Sato J.R. Manfro A.G. et al.Reaction time variability and attention-deficit/hyperactivity disorder: is increased reaction time variability specific to attention-deficit/hyperactivity disorder? Testing predictions from the default-mode interference hypothesis., 27Hobson C.W. Scott S. Rubia K. Investigation of cool and hot executive function in ODD/CD independently of ADHD.. Specifically, alterations in reward and punishment sensitivity may provide targets for behavioral parent training in this population (see 12How to Improve Behavioral Parent and Teacher Training for Children with ADHD: Integrating Empirical Research on Learning and Motivation into Treatment. for a review on specific instrumental learning deficits in ADHD and the link with behavioral parent training).

The current meta-analysis took these limitations into account, by including only randomized controlled trials with samples in which all children were diagnosed with ADHD, and by scoring the dosage of behavioral techniques (i.e., the number and percentage of sessions in which a specific technique is being taught to parents). More specifically, we scored the behavioral techniques of all behavioral parent training manuals that were used in the included trials using a taxonomy consisting of 39 different behavioral techniques, grouped in 8 main categories (see Supplement 1, available online, for details).

In addition to the effects of the dosage of different techniques, we exploratively investigated whether other characteristics of the intervention (setting, delivery method, format, duration, collaboration with school, allowance of concurrent medication, and checking treatment integrity) or characteristics of the child (age, sex, comorbid disorders) were associated with the effectiveness of the intervention. So far, evidence for the possible moderating effect of these characteristics is mixed. One meta-analysis showed that individual parent training was more effective than group training on parenting outcomes28Lundahl B. Risser H.J. Lovejoy M.C. A meta-analysis of parent training: Moderators and follow-up effects., whereas a meta-analysis on parent training in preschoolers with ADHD did not observe this effect29Rimestad M.L. Lambek R. Zacher Christiansen H. Hougaard E. Short- and Long-Term Effects of Parent Training for Preschool Children With or at Risk of ADHD: A Systematic Review and Meta-Analysis.. Another meta-analysis on parent training in preschoolers with ADHD reported no moderation of age (within preschoolers) and duration of the intervention30Mulqueen J.M. Bartley C.A. Bloch M.H. Meta-Analysis: Parental Interventions for Preschool ADHD., but a meta-analytic review on behavioral interventions for children with a wider age range demonstrated larger effects on positive parenting for younger children5Daley 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 outcome domains.. Meta-analytic evidence for other moderator effects is scarce, but highly needed to establish who benefits most from behavioral parent training31Daley D. Van Der Oord S. Ferrin M. et al.Practitioner review: Current best practice in the use of parent training and other behavioural interventions in the treatment of children and adolescents with attention deficit hyperactivity disorder..

The primary aim of the current study was to investigate whether the behavioral techniques that were included in our taxonomy were associated with the effects of behavioral parent training for parents of children with ADHD on five different parental domains, i.e., positive parenting, negative parenting, parenting sense of competence, quality of the parent-child relationship, and parental mental health. A secondary aim was to investigate whether intervention and child characteristics influenced the training effects.

Results Study selectionIn total, 29 studies with 35 interventions and 138 relevant effect sizes were included (see Figure 1). Study characteristics can be found in Table 1. The mean and the range of the dosage of all scored techniques are provided in Supplement 6, available online. Risk of bias

Risk-of-bias analyses can be found in Supplement 7, available online (interrater reliability κ = .96). Few studies included blinded measures, most studies reported complete outcome data. Overall, the information in the published manuscripts of many studies was not sufficient to determine the risk of bias for the categories random sequence generation, allocation concealment, vested interest, and selective outcome reporting.

 Main effectsSignificant small-to-medium-sized main effects (ranging from .41 to .60, all p’s Table 2 and Figure 2).

Table 2Main Effects per Outcome Domain

Note: Table presents overall main effects, main effects on probably blinded and unblinded measures separately, main effects after trim-fill correction. k represents the number of studies, with the number of effect sizes (n_es) in parentheses. The trim-fill correction calculates the effect size after correcting for funnel plot asymmetry. CI = confidence interval; I2

NICE. Attention Deficit Hyperactivity Disorder: Diagnosis and Management.; 2018.

represents “the proportion of variation in study estimates that is due to heterogeneity” 51Higgins J.P.T. Thompson S.G. Quantifying heterogeneity in a meta-analysis., separately for heterogeneity between variables (level 2) and between studies (level 3); s/of = sense of; SMD = standardized mean difference in terms of Hedges’ g.

* p < .05, ** p < .01, *** p < .001.

Figure 2Forest Plots With all Effect Sizes Sorted per Outcome Domain

Effect sizes were not significantly different for probably blinded and unblinded measures on positive parenting (βˆ1 = .05, p = .82), negative parenting (βˆ1 = -.15, p = .52) and parent-child relation (βˆ1 = .02, p = .95) (see Table 2 for effect sizes). None or too few blinded measures were available for parenting sense of competence and parental mental health.Visual inspection of the forest plot (Figure 2) depicted one outlying study with particularly high effect sizes on the outcome domain of parental mental health 50Aghebati A. Gharraee B. Shoshtari M.H. Gohari M.R. Triple P-positive parenting program for mothers of ADHD children.. As noted in Table 2, the effect size decreased and the heterogeneity between studies dropped considerably after removing that study. For the forthcoming analyses on the parental mental health domain, we therefore decided to exclude that study from further analyses (analyses including this study were also performed and are described in Supplement 8, available online). A sensitivity check excluding one study measuring parental ADHD revealed highly similar results (see Supplement 9, available online).

A sensitivity check comparing studies with active control conditions and studies with waitlist control conditions demonstrated that—as expected—effect sizes for most of the outcome domains were smaller when parent training was compared to an active control condition than when it was compared to a waitlist. However, this difference in effect sizes was only significant for negative parenting (see Supplement 10, available online, for detailed results).

 Robustness analysesEgger’s test46Egger M. Smith D. Schneider M. Minder C. Bias in meta-analysis detected by a simple, graphical test. indicated signs of funnel plot asymmetry (which potentially indicates publication bias) on negative parenting (t = 2.57, p = .02), parenting sense of competence (t = 2.17, p = .04), and parent-child relation (t = 2.46, p = .03), but not on positive parenting (t = 1.83, p = .08) and parental mental health (t = .55, p = .59). Note that Egger’s test and trim-and-fill analyses were performed without modeling for dependency between effect sizes, as this is currently not possible.Trim-and-fill analyses52

Duval S, Tweedie R. Trim and Fill : A Simple Funnel-Plot-Based Method of Testing and Adjusting for Publication Bias in Meta-Analysis. 2014;56(2):455-463.

demonstrated that, on the left side of the funnel plot, 9 studies were missing for negative parenting, 4 studies were missing for parental mental health, 4 for parent-child relation, 3 for positive parenting, and 1 for parenting sense of competence (see Supplement 11, Figure S1, available online, for funnel plots with estimated missing effect sizes). After taking into account these missing studies, the effect size slightly decreased (but remained significant) for positive parenting, negative parenting, parent-child relation, and parental mental health and remained similar for parenting sense of competence (effect sizes ranged between .29 and .54, all p’s Table 2).P-curves48Simonsohn U. Nelson L.D. Simmons J.P. P-curve: A key to the file-drawer. were created for all outcome categories. In all cases, p-curves were right-skewed and indicated evidential value (all p-values for the half p-curve p-curves.

Sensitivity analyses (i.e., leave-one-out) indicated that effect sizes were not heavily dependent on single studies, as the minimum and maximum effect size calculated by the leave-one-out analyses did not differ substantially for all outcome domains, and all were within the range of the confidence intervals of the main effects: for positive parenting (minimum SMD = .53, p < .001; maximum SMD = .64, p < .001), for negative parenting (minimum SMD = .53, p < .001; maximum SMD = .63, p < .001), for parenting sense of competence (minimum SMD = .49, p < .001; maximum SMD = .57, p < .001), for parent-child relation quality (minimum SMD = .45, p < .001; maximum SMD = .58, p < .001), and for parental mental health (minimum SMD = .32, p < .001; maximum SMD = .37, p < .001).

Additional sensitivity analyses were performed to assess the influence of dependency in studies that compared two intervention arms with one control condition. The influence was minimal (see Supplement 5, available online, for details).

In sum, although some analyses indicate the presence of publication bias (Egger’s test, trim-fill analyses), the main effects are robust given the results of the trim-fill corrected analyses, p-curves and leave-one-out analyses.

Primary analyses: Does the dosage of behavioral techniques influence the effects of parent training?

All results regarding the association between the overall effect size and the dosage of technique categories or specific techniques (measured by both the percentage and the number of sessions in which a technique from a certain category/a specific technique occurs) can be found in Supplement 12, available online.

In Table 3, we summarized those categories/techniques that were significantly associated with outcomes. We found that (1) the dosage of psycho-education was negatively associated with treatment effects on positive parenting and the quality of the parent-child relation; (2) the dosage of teaching parents to manipulate antecedents, and in particular to anticipate misbehaviors, was positively associated with treatment effects on parenting sense of competence and parental mental health; (3) the dosage of teaching parents to work with positive consequences was associated with larger treatment effects on negative parenting; (4) the dosage of practicing/generalization techniques was negatively associated with the effect of treatment on the parent-child relation. Dosage was established by either the number or the percentage of sessions in which the techniques occur (see Table 3 for specific results). Most of the analyses, however, were not significant (see Supplement 12, available online).

Table 3Significant Effects of Primary and Secondary Analyses

Note: Significant effects of dosage of categories of behavioral techniques and specific techniques (primary analyses; upper panel) and significant effects of intervention and participant characteristics (secondary analyses; lower panels). k represents the number of studies, with the number of effect sizes (n_es) in parentheses. βˆ1 denotes the change in standardized mean difference (in terms of Hedges’ g) when the moderator increases with one unit (95% confidence intervals are depicted in parentheses). I2

NICE. Attention Deficit Hyperactivity Disorder: Diagnosis and Management.; 2018.

represents “the proportion of variation in study estimates that is due to heterogeneity” 51Higgins J.P.T. Thompson S.G. Quantifying heterogeneity in a meta-analysis., separately for heterogeneity between variables (level 2) and between studies (level 3). Category numbers (and specific techniques within categories with numbers and letters) correspond with the taxonomy (see Supplement 1, available online).

∗p < .05; ∗∗p < .01; ∗∗∗p < .001; **** p < .10.

 Secondary analyses: Moderation by intervention and participant characteristics Intervention characteristicsSecondary analyses of intervention characteristics on all outcomes were all non-significant except for two findings. First, on negative parenting, parent training only was associated with a larger effect than parent training as part of multimodal treatment. Second, interventions without integrity checks (although only 2 studies containing 6 effect sizes) were more effective than interventions with integrity checks on positive and negative parenting. See Table 3 for significant findings, see Supplement 13, Table S1-3, available online, for all results. Child characteristicsEffects of parent training on negative parenting were higher when the child was younger, and effects on positive parenting were higher for preschoolers than for school-age children (the latter were more pronounced when two studies on adolescents were excluded, see Supplement 14, available online, for details on these sensitivity analyses). Sex of the child moderated the effect on negative parenting: the higher the percentage of boys in the sample, the higher the effects. Finally, the percentage of comorbid externalizing disorders in the sample did not moderate any outcomes. All analyses containing moderation by participant characteristics can be found in Supplement 13, Table S3, significant findings are also in Table 3.Discussion

Behavioral parent training for children with ADHD consists of many different techniques that are being taught to parents. To enhance the development of more effective future parent training interventions, the primary aim of the current study was to investigate which behavioral techniques were associated with better or worse parental outcomes.

We found that higher dosages of behavioral techniques teaching parents to manipulate the antecedents of behavior (i.e., stimulus control techniques), in particular by explicitly anticipating potential misbehavior of the child (e.g., thinking ahead, prepare a plan before entering a problematic situation), were associated with positive effects on parenting sense of competence and parental mental health. Stimulus control techniques may be particularly relevant for children with ADHD, with their noted disorganization and executive functioning deficits12How to Improve Behavioral Parent and Teacher Training for Children with ADHD: Integrating Empirical Research on Learning and Motivation into Treatment.,53Kasper L.J. Alderson R.M. Hudec K.L. Moderators of working memory deficits in children with attention-deficit/hyperactivity disorder (ADHD): A meta-analytic review., 54Willcutt E.G. Doyle A.E. Nigg J.T. Faraone S.V. Pennington B.F. Validity of the executive function theory of attention-deficit/hyperactivity disorder: a meta-analytic review., 55Martinussen R. Hayden J. Hogg-Johnson S. Tannock R. A meta-analysis of working memory impairments in children with Attention-Deficit/Hyperactivity Disorder.. These intervention components are (relative to for example techniques focusing on relationship building and teaching communication skills) relatively straightforward for parents and focus on the prevention of non-adaptive behavior, which may provide a sense of control. This may give parents immediate mastery of these skills. Also, it may well be that children respond quickly to these techniques and there