Magnitude and Moderators of Digital Placebo Effects on Depressive Symptoms: A Systematic Review and Meta-analysis

Digital health tools, or “digital therapeutics” ((DTx) such as phone apps or web-based programs, are increasingly being tested in clinical trials to help people with mental health problems. Depression is one of the most common mental health conditions, affecting many millions of adults worldwide and often causing low mood, loss of interest, sleep problems, and difficulty with daily life. As more DTx are tested in clinical trials, researchers need good comparison groups to understand whether an app truly helps.

In drug trials, a “placebo” is an inactive treatment that looks like the real one. In digital trials, a “digital placebo” can be an app or digital program that looks and feels similar to the active program but does not include the key features intended to treat depression. Digital placebos are used so that participants (and ideally the research team assessing outcomes) do not know which intervention was given, helping reduce bias. However, digital placebos may still improve symptoms—for example through hope of benefit, attention, reminders, or feeling supported—and the size of this improvement is not yet well understood.

In this project, we will carry out a systematic review and meta-analysis. A systematic review means we will search for, screen, and review the results from all relevant clinical trials. A meta-analysis means we will combine data from different studies to estimate the average effect more precisely.

We will do two main things. First, we will describe the key characteristics of the clinical trials that used a digital placebo as the comparison group (such as the type of digital placebo used, how blinding was attempted, who was included, and how outcomes were measured). Second, we will estimate whether digital placebos lead to a measurable improvement in depressive symptoms and, if they do, what factors (“moderators”) make the placebo effect larger or smaller.

To make results comparable across trials, we will focus our main meta-analysis on studies that measured depression using the Patient Health Questionnaire-9 (PHQ-9), a widely used 9-item questionnaire completed by the patients. We will include studies of adults aged 18 years and older, but we will not restrict inclusion to people with a formal depression diagnosis so that we can study a broader range of participants. We will extract PHQ-9 scores before and after the treatment intervention. We will then calculate the average change in PHQ-9 in digital placebo groups.

This research is needed because digital placebo effects can make an active digital treatment look more effective than it truly is. By quantifying the digital placebo effect and identifying what influences it, we will help improve the design and interpretation of future digital mental health trials, supporting better evidence for patient care.