High Potential But Limited Evidence: Using Voice Data From Smartphones to Monitor and Diagnose Mood Disorders

Objective: This article evaluates the potential of smartphone audio data to monitor individuals recovering from mood disorders. Method: A comprehensive literature review was conducted based on searches in 9 bibliographic databases. Results: Seven articles were identified that used smartphone audio data to monitor participants with bipolar disorder from 4 to 14 weeks. The studies captured audio data in various contexts (e.g., in-person daily conversations, phone calls) and used common audio features (e.g., pitch and volume) to ascertain clinically relevant outcomes, including mood and social rhythm. Findings suggest that the utility of audio data in clinical and research contexts remains relatively unexplored and presents some challenges. For example, information on adherence and engagement among individuals recovering from bipolar disorder were often insufficient to gauge the generalizability of findings. Conclusions and Implications for Practice: Despite growing interest, additional research is required to confirm clinical utility of smartphone audio data for mood disorders.