Disruptive child behavior severity and parenting program session attendance Individual participant data meta-analysis

Objective: We tested if baseline disruptive child behavior problem severity predicts parental attendance at sessions of a parenting group program. Method: We used a database of randomized trials of the Incredible Years parenting program in Europe and restricted the sample to participants randomized to the intervention arm. Using baseline Eyberg Child Behavior Inventory scores, we distinguished between trial-level problem severity and child-level problem severity, compared linear and quadratic functional forms at both levels, and considered cross-level interactions, all in a multilevel Poisson regression framework. Results: Drawing on 918 participants in 12 trials, we found that within trials, parents of children with the least and most severe problems attended fewer sessions. Between trials, each additional 10-point increase in the Eyberg Child Behavior Inventory trial mean predicted an 11% increase in attendance. Models including child sex, age, or family low-income did not change coefficients or their interpretation. Conclusions: Our findings suggest that although generally attendance is higher in parents of children with more challenging behavior, it seems difficult for group programs to keep families with the least or most severe problems engaged. Our findings call for the need to better understand the conditions under which lower attendance translates into equivalent or lesser program benefits.