Abstract
OBJECTIVE: To identify patient- and treatment-level factors that predict intervention engagement and outcome for adolescents with attention-deficit/hyperactivity disorder (ADHD), guiding efforts to enhance care. METHOD: Integrative data analysis was used to pool data from 4 randomized controlled trials of adolescent ADHD treatment with participants (N = 854) receiving various evidence-based behavioral therapy packages in 5 treatment arms (standard [STANDARD], comprehensive [COMP], engagement-focused [ENGAGE]), community-based usual care (UC), or no treatment (NOTX). Participants also displayed varying medication use patterns (negligible, inconsistent, consistent) during the trial. Regression and latent growth curve analyses examined treatment- and patient-level predictors of engagement and outcome. RESULTS: Compared with COMP, ENGAGE was associated with higher parent engagement in behavioral therapy (d = 1.35-1.73) when delivered in university, but not community, clinics. Under some conditions, ENGAGE also predicted youth engagement in behavioral therapy (d = 1.21) and lower likelihood of negligible medication use (odds ratio = 0.49 compared with NOTX). UC was associated with poorer parent engagement compared with COMP (d = -0.59) and negligible medication use (odds ratio = 2.29) compared with NOTX. Compared with COMP, ENGAGE (in university settings) was consistently associated with larger ADHD symptom improvements (d = 0.41-0.83) at 6-month follow-up and sometimes associated with larger grade point average (d = 0.68) and parent-teen conflict (d = 0.41) improvements. Consistent medication use during behavioral therapy was associated with larger improvements in ADHD symptoms (d = 0.28) and parent-teen conflict (d = 0.25-0.36). An ADHD+internalizing clinical profile predicted larger improvements in grade point average (d = 0.45). Family adversity predicted poorer parent and youth engagement (rate ratio = 0.90-0.95), negligible medication use (odds ratio = 1.22), and smaller improvements in grade point average (d = -0.23). African American race predicted smaller improvements in parent-teen conflict (d = -0.49). CONCLUSION: Engagement-focused behavioral therapy and consistent medication use most frequently predicted stronger clinical engagement and outcomes for adolescents with ADHD. Youths who are African American or who experience family adversity may demonstrate treatment-related disparities for certain outcomes; youths with ADHD+internalizing symptoms may demonstrate excellent academic outcomes following behavioral therapy. DATA SHARING: The full ADHD TIDAL dataset is publicly available through the National Data Archive (https://nda.nih.gov), including a data dictionary. The study protocol is also publicly available: https://doi.org/10.1186/s12888-020-02734-6.