youthu

youthu maps psychological measures collected in youth mental health services to Assessment of Quality of Life – Six Dimensions (AQoL-6D) utility scores to enable estimation and longitudinal prediction of health-related quality of life.

Key Features:

• Mapping Models Development: Develops models to convert six psychological measures of distress, depression, and anxiety into AQoL-6D utility scores for adolescents in primary mental health services.
• Longitudinal Prediction Capability: Assesses the ability of mapping models to predict longitudinal changes in AQoL-6D utility scores over time.
• Dataset: Uses data from 1,107 young people attending Australian primary mental health services with two time points collected three months apart.
• Model Types Evaluated: Explores five linear and three generalized linear models for mapping psychological measures to utility scores.
• Mixed-Effects Modeling: Employs linear and generalized linear mixed-effect models for longitudinal predictive modeling of AQoL-6D change.
• Model Evaluation Metrics: Evaluates models using ten-fold cross-validation and reports R², root mean square error (RMSE), and mean absolute error (MAE).
• Psychological Measures Integration: Integrates six candidate measures and identifies the Patient Health Questionnaire-9 (PHQ-9) as a strong independent predictor of AQoL-6D utility.
• Transformation and Performance: Identifies linear regression models with complementary log-log transformation of utility scores as best-performing, capturing both between-person and within-person associations with slight bias toward between-person effects.

Scientific Applications:

• Health Utility Estimation: Enables estimation of AQoL-6D utilities from routine psychological assessments to quantify health-related quality of life in youth mental health services.
• Longitudinal Health Monitoring: Supports monitoring of changes in health utility over time to inform clinical assessment of patient trajectories.
• Research and Policy Development: Provides empirical mapping models and performance metrics to inform research on adolescent mental health outcomes and policy decisions.

Methodology:

Evaluates five linear and three generalized linear models, applies complementary log-log transformation to utility scores, uses ten-fold cross-validation with R², RMSE, and MAE for model assessment, and employs linear and generalized linear mixed-effect models for longitudinal prediction of AQoL-6D change.