JMIR Publications

ABSTRACT

Background:

There is certain evidence on the efficacy of smartphone-based mental health interventions. However, mechanisms of action remain unclear. Placebo effects contribute to the efficacy of face-to-face mental health interventions, and may also be a potential mechanism of action in smartphone-based interventions.

Objective:

We investigated whether different types of efficacy expectancies as potential factors underlying placebo effects could be successfully induced in a smartphone-based digital placebo mental health intervention, ostensibly targeting mood and stress.

Methods:

We conducted a randomized, controlled, single-blinded superiority trial with a multi-arm parallel design. Participants underwent an Android-based smartphone-based digital placebo mental health intervention for 20 days. We induced prospective efficacy expectancies via initial instructions on the purpose of the intervention and retrospective efficacy expectancies via feedback on the success of the intervention at days 1, 4, 7, 10, and 13. 132 healthy participants were randomized to a prospective expectancy only (n=33), a retrospective expectancy only (n=33), a combined expectancy (n=34), or a control condition (n=32). As endpoint, we assessed changes in efficacy expectancies with the Credibility Expectancy Questionnaire, before the intervention and on days 1, 7, 14, and 20. For statistical analyses, we used a random effects model for the intention-to-treat sample, with intervention day as time variable and condition as two factors: prospective expectancy (yes vs. no), and retrospective expectancy (yes vs.no), allowed to vary over participant and intervention day.

Results:

Credibility (b = -1.63, 95%confidence interval (CI) [-2.37, -0.89], P < .001) and expectancy (b = -0.77, 95%CI [-1.49, -0.05], P = .04) decreased across intervention days. For credibility and expectancy, we found significant three-way interactions intervention day*prospective expectancy*retrospective expectancy (b = 2.05, 95%CI [0.60, 3.50], P < .006 resp. b = 1.55, 95%CI [0.14, 2.95] P = .03), suggesting that efficacy expectancies changed throughout intervention days and depending on condition.

Conclusions:

To our knowledge, this is the first empirical study investigating whether efficacy expectancies could be successfully induced in a specifically designed placebo smartphone-based mental health intervention. Our findings may pave the way to diminish or exploit digital placebo effects and help to improve efficacy of digital mental health interventions. Clinical Trial: ClinicalTrials.gov Identifier: NCT02365220. Registered February 18, 2015.