IMPORTANCE Social deficits are a common and disabling feature of many pediatric disorders; however, whether behavioral interventions are associated with benefits for children and adolescents with social deficits is poorly understood. OBJECTIVE To assess whether behavioral interventions in children and adolescents with neurodevelopmental or mental health disorders are associated with improvements in social function and social cognition, and whether patient, intervention, and methodological characteristics moderate the association.DATA SOURCES For this systematic review and meta-analysis, the PsycINFO, MEDLINE, and PubMed electronic databases were searched in December 2020 for randomized clinical trials published from database inception to December 1, 2020, including terms related to neurodevelopmental or mental health disorders, social behavior, randomized clinical trials, and children and adolescents. Data were analyzed in January 2021.STUDY SELECTION Randomized clinical trials that enrolled participants aged 4 to 17 years with social deficits and examined the efficacy of a clinician-administered behavioral intervention targeting social functioning or social cognition were included. A total of 9314 records were identified, 78 full texts were assessed for eligibility, and 33 articles were included in the study; 31 of these reported social function outcomes and 12 reported social cognition outcomes.DATA EXTRACTION AND SYNTHESIS Articles were reviewed using the Cochrane Risk of Bias Assessment for randomized clinical trials. Data were independently extracted and pooled using a weighted random-effects model.
MAIN OUTCOMES AND MEASURESThe main outcome was the association of behavioral intervention with social function and social cognition. Hedges g was used to measure the standardized mean difference between intervention and control groups. Standardized effect sizes were calculated for the intervention group vs the comparison group for each trial.RESULTS A total of 31 trials including 2131 participants (1711 [80%] male; 420 [20%] female; mean [SD] age, 10.8 [2.2] years) with neurodevelopmental or mental health disorders (autism spectrum disorder [ASD] [n = 23], attention-deficit/hyperactivity disorder [n = 4], other conditions associated with social deficits [n = 4]) were analyzed to examine differences in social function between the intervention and control groups. Significantly greater gains in social function were found among participants who received an intervention than among the control groups (Hedges g, 0.61; 95% CI, 0.40-0.83; P < .001). The type of control condition (wait list vs active control vs treatment as usual) was a significant moderator of effect size (Q 2 , 7.11; P = .03). Twelve studies including 487 individuals with ASD (48 [10%] female; 439 [90%] male; mean [SD] age, 10.4 [1.7] years) were analyzed to examine differences in social cognition between intervention and control groups. The overall mean weighted effect was significant (Hedges g, 0.67; 95% CI, 0.39-0.96; P < .001), indicating the treatme...