Cost-effectiveness of behavioral activation compared to treatment as usual for depressed older adults in primary care: A cluster randomized controlled trial

Depression in older adults is associated with decreased quality of life and increased utilization of healthcare services. Behavioral activation (BA) is an effective treatment for late-life depression, but the cost-effectiveness compared to treatment as usual (TAU) is unknown. An economic evaluation was performed alongside a cluster randomized controlled multicenter trial including 161 older adults (≥65 years) with moderate to severe depressive symptoms (PHQ-9 ≥ 10). Outcome measures were depression (response on the QIDS-SR), quality-adjusted life-years (QALYs) and societal costs. Missing data were imputed using multiple imputation. Cost and effect differences were estimated using bivariate linear regression models, and statistical uncertainty was estimated with bootstrapping. Cost-effectiveness acceptability curves showed the probability of cost-effectiveness at different ceiling ratios. Societal costs were statistically non-significantly lower in BA compared to TAU (mean difference (MD) -€485, 95 % CI -3861 to 2792). There were no significant differences in response on the QIDS-SR (MD 0.085, 95 % CI -0.015 to 0.19), and QALYs (MD 0.026, 95 % CI -0.0037 to 0.055). On average, BA was dominant over TAU (i.e., more effective and less expensive), although the probability of dominance was only 0.60 from the societal perspective and 0.85 from the health care perspective for both QIDS-SR response and QALYs. Although the results suggest that BA is dominant over TAU, there was considerable uncertainty surrounding the cost-effectiveness estimates which precludes firm conclusions. [Display omitted] •Behavioral activation was dominant (less costly, more effective) compared to treatment as usual from a societal perspective.•Behavioral activation was also dominant from a health care perspective.•Behavioral activation is probably cost-effective compared to treatment as usual.•Differences in societal costs and effects were not statistically significant.