Therapist-guided and self-guided internet-based behavioural activation versus treatment as usual for adolescents with depression: randomised controlled trial with economic evaluation.

Abstract

Background: Adolescent depression is a major public health concern with limited access to effective treatments. Behavioural activation (BA) is a suitable intervention for digital delivery, but definitive adolescent trials are scarce.

Objective: To evaluate whether therapist-guided and self-guided internet-based BA (I-BA) are more efficacious and cost-effective than treatment as usual (TAU) for adolescents with mild to moderate major depressive disorder (MDD).

Methods: Single-blinded, parallel-group randomised controlled trial with economic evaluation. A total of 219 adolescents (13-17 years) with mild-moderate MDD were randomised (1:1:1) to 10 weeks of therapist-guided I-BA, self-guided I-BA or TAU. Both I-BA interventions included adolescent and parent modules.

Primary outcome: change in depression severity (Children's Depression Rating Scale-Revised, CDRS-R, range 17-113) from baseline to 3-month follow-up (primary endpoint), assessed by blinded evaluators. Analyses included all randomised participants, with statisticians blinded to allocation. A health economic evaluation was performed at the primary endpoint.

Findings: Baseline CDRS-R indicated clinically significant depression (mean 57.1, threshold ≥40). Retention at the primary endpoint was 82.6%. Mean reductions were 17.0 (therapist-guided I-BA), 16.0 (self-guided I-BA) and 11.6 (TAU), bringing both I-BA groups below the clinical cut-off. Therapist-guided I-BA showed greater reductions than TAU (estimated mean difference -4.68; 95% CI -0.05 to -9.30; p=0.048, d=-0.47; 95% CI 0.00 to -0.93), while self-guided I-BA was not statistically superior (estimated mean difference -3.44; 95% CI -8.05 to 1.17, p=0.14, d=-0.37; 95% CI -0.86 to 0.12). The predefined six-point clinical important difference was not met. Results were sensitive to modelling: significance disappeared with random slopes but effect estimates were similar; baseline-adjusted analyses favoured both I-BA arms. Both I-BA options had lower costs than TAU (p=0.03, p<0.001), with self-guided I-BA being the most economical.

Conclusions: Therapist-guided I-BA reduced depressive symptoms at lower cost than TAU, but clinical importance and robustness are uncertain. Self-guided I-BA showed no clear superiority but was cost-efficient.

Clinical implications: Therapist-guided I-BA may increase access to evidence-based care for adolescents with MDD, though findings require cautious interpretation. Self-guided I-BA may be useful where therapist access is limited, but more research is needed.