Cost-effectiveness of a novel e-health depression service.

OBJECTIVES: A recent trial assessed feasibility of an e-health service (" Improvehealth.eu ") to support depression care and reported positive outcomes. Our objective was to examine cost-effectiveness of the Improvehealth.eu service. A baseline model was used to evaluate cost and effects of the intervention. Given the high uncertainty in the input space, a series of alternative scenarios were evaluated to challenge the result. The aim was to find if conservative or even pessimistic estimates and assumptions could result in a change of the cost-effectiveness from the baseline model.
MATERIALS AND METHODS: A probabilistic depression model combined with bootstrapping was built and populated with data from the literature and from the pilot efficacy trial of the e-health service. The core of the model was a stochastic mapping function that translated depression-specific outcomes to quality-adjusted life years. Correlated sampling was used to obtain unbiased and consistent piecewise linear transformation of Beck Depression Inventory scores to utilities. The results are shown as cost-effectiveness acceptability curves with value of information data. An extreme scenario analysis was then performed to deal with parameter, structural, and modeling uncertainty.
RESULTS: Cost-effectiveness of the e-health service was favorable because of low cost and high efficacy of the intervention. Apart from the most pessimistic one, none of the 13 alternative scenarios changed the preferred alternative.
CONCLUSIONS: Improvehealth.eu is cost-effective relative to usual care, given the available efficacy data. Results of the health economic evaluation were robust to alternative assumptions, despite considerable uncertainty in input data.