Predicting Dropout from Children's Mental Health Services: Using a Need-Based Definition of Dropout.

Abstract

Dropout from children's mental health services has negative impacts on children, families and service providers. To target interventions aimed to reduce dropout, it is essential to predict individuals who dropout. This study compares predictors of dropout using a novel need-based definition, to existing definitions of dropout. Children (N = 650; 61% male) aged 5-13 attended five children's mental health agencies in Ontario. A mixed effects logistic regression was used to model binary outcome variables (i.e., dropout or treatment completion), for each definition of dropout. Using the need-based definition, older child age, lower child problem presentation, higher child risk behaviors, higher caregiver needs, and more child strengths predicted an increased likelihood of dropout. The need-based definition results in different predictors of dropout than existing definitions in the literature. High caregiver needs was the only predictor of dropout across all definitions. Caregiver needs represent a prospective target when distributing interventions aimed to reduce dropout.