Passive Sensing for the Prediction of Suicidal Thoughts and Behaviors: A Systematic Review and Recommendations for Future Research

Abstract

Predicting suicidal thoughts and behaviors (STB) remains challenging. The use of passive sensing data gathered through smartphones and wearables may contribute to overcoming current limitations in STB prediction. In this systematic review, we explored the feasibility and predictive validity of passive sensing for STB. On October 18, 2022, we systematically searched Medline, Embase, Web of Science, PubMed, and PsycINFO. Studies were eligible if they reported on the association between STB and passive sensing through smartphones or wearables, or on the feasibility of passive sensing in this context. The risk of bias was assessed by two independent researchers using the PROBAST tool. Out of 1765 unique records, we identified eight prediction studies, six feasibility studies, and five protocols. Studies found that electrodermal activity, sleep characteristics, heart rate variability, and app usage were associated with STB. However, results on the incremental value of passive data beyond self-report are inconsistent. Risk of bias ratings revealed major shortcomings in methodology and reporting. Studies indicated that passive sensing is feasible in terms of user satisfaction and adherence. In conclusion, there is only limited evidence on the predictive value of passive sensing for the prediction of STB. We highlight important quality characteristics for future research.