Along with her team, Dr. Delphine Raucher-Chéné , a researcher at the Douglas Research Centre and Assistant Professor at McGill University’s Department of Psychiatry, has just published a scoping review investigating the use of digital tools in the context of bipolar disorder. The article, entitled “Digital tools for assessing bipolar disorder: A scoping review of the current landscape,” was published on April 21, 2026, in Neuroscience Applied.
This work was conducted in collaboration with Drs. Katie Lavigne & Geneviève Sauvé, also from the Douglas Research Centre.
KEY TAKEAWAYS
Digital tools for assessing bipolar disorder have expanded rapidly over the last decade, with a clear shift from self-report methods toward multimodal systems that combine smartphones and wearable sensors. Across the 214 studies included in this review, mood, sleep, and activity were the most frequently assessed domains, while cognition, physiological signals, and communication-based features remain comparatively underdeveloped. Most studies were conducted in high-income countries and involved younger, predominantly female samples, raising concerns about generalizability. Passive sensing approaches and hybrid designs are becoming increasingly common and have demonstrated strong feasibility and user adherence.
WHY THIS MATTERS
Bipolar disorder is a complex and chronic mental health condition characterized not only by mood fluctuations but also by cognitive impairments, sleep disturbances, and psychosocial difficulties. Traditional clinical assessments are typically brief and episodic, making it difficult to capture the dynamic and fluctuating nature of these symptoms in real-world settings. Digital tools enable continuous and ecologically valid monitoring, offering new opportunities to detect early warning signs of relapse, personalize treatment, and improve long-term outcomes for individuals living with the disorder.
WHAT THEY DID
The authors conducted a scoping review using established methodological frameworks from the Joanna Briggs Institute and PRISMA-ScR guidelines. They systematically searched multiple databases and grey literature sources up to March 2026. Of 4,418 identified records, 214 studies met the inclusion criteria. Eligible studies included participants at risk of or diagnosed with bipolar disorder and focused specifically on assessment using digital tools rather than interventions. The authors synthesized the findings narratively, mapping the types of technologies used, the populations studied, and the domains assessed.
WHAT THEY FOUND
The review found that smartphone applications, such as MONARCA, Monsenso, and Beiwe, were the most used tools, typically supporting active self-report and ecological momentary assessment. Wearable devices, including actigraphy watches, smart rings, and physiological sensors, were widely used for passive monitoring of sleep, activity, and physiological signals. Over time, data collection approaches have evolved, with 67 studies relying exclusively on active methods, 52 on passive sensing, and 94 using hybrid approaches that combine both. Mood was the primary domain assessed, followed by activity and sleep or circadian patterns. In contrast, relatively few studies examined cognitive functioning, physiological markers, or communication-related data such as speech and phone use. Participant samples tended to include young adults with a mean age of approximately 38 years, and about 72% of participants were female. Most studies did not adequately report illness stage or clinical state. Geographically, research was heavily concentrated in the United States and Europe.
WHAT IT MEANS
The findings suggest that digital monitoring of bipolar disorder is both feasible and increasingly sophisticated, with important advantages over traditional clinical approaches. Continuous, real-world data collection supports the development of digital biomarkers capable of identifying mood states and predicting relapse. However, the current evidence base is limited by demographic and geographic biases, inconsistent reporting practices, and a narrow focus on certain symptom domains. These challenges currently limit the clinical applicability and generalizability of digital monitoring tools in bipolar disorder.
WHAT’S NEXT
Future research should focus on improving standardization across tools, data collection methods, and reporting practices to enhance comparability and reproducibility. There is also a critical need to expand studies to more diverse and underserved populations and to develop culturally adapted and affordable technologies. Greater attention should be given to underexplored domains, including cognition, physiology, and communication patterns, to better capture the full complexity of bipolar disorder. Advances in multimodal digital phenotyping have the potential to support earlier detection, more personalized interventions, and improved longitudinal care if these challenges are addressed.