Group-based trajectory modelling of multiple health outcomes in a cost-consequence framework: a randomized controlled trial of a remote person-centred care intervention for people with common mental disorders in Sweden.

backgroundEvaluations of complex interventions are increasingly used within the health and social sciences and form an integral part of the ongoing research into person-centred care (PCC). The longitudinal nature of these interventions, measuring multiple outcomes over time, is often challenged by traditional economic frameworks when informing decision-makers. Therefore, the aim of this study is to explore how group-based trajectory modelling (GBTM) can be used to evaluate multiple outcomes within economic evaluations, specifically when outcomes are measured longitudinally and lack an established summary metric.

methodsGBTM and a cost-consequence analysis (CCA) were performed on two-year data from the PROMISE randomised controlled trial, a remote PCC add-on intervention combining a web-based platform and telephone support for people on sick leave due to common mental disorders. Group trajectories were modelled combining General Self-Efficacy (GSE), EQ-5D values, the Perceived Stress Scale (PSS), and the Shirom-Melamed Burnout Questionnaire (SMBQ). Costs were reported from a societal perspective for trajectory groups, and within a CCA summarising healthcare use, prescription drugs, productivity loss, and quality-adjusted life years (QALY).

resultsThe intervention group had lower total and mean costs for primary care, prescription drugs, and productivity loss compared to the control group at two years. GBTM identified four groups: High, Moderate, Low, and No Improvement. The High Improvement group, of whom 53% belonged to the intervention group, had the greatest improvements across all outcome measures and the lowest mean costs. The No Improvement group experienced the worst baseline health as well as significant differences in education level and primary diagnosis. No statistically significant differences were found between the intervention and control groups in relation to trajectory group allocation. GBTM and QALY differences were sensitive to imputation and value sets.

conclusionThis study demonstrates that GBTM coupled with CCA, offers a practical framework when evaluating complex interventions where multiple outcomes evolve over time. Understanding the importance of sociodemographic factors and heterogeneous response patterns meaningfully enriches economic evaluation and can help guide the development and evaluation of person-centred, complex interventions. CLINICAL TRIAL NUMBER: NCT03404583 - ClinicalTrials.gov.