October 2, 2025
Artificial Intelligence in Depression–Medication Enhancement (AID-ME): A Cluster Randomized Trial of a Deep-Learning–Enabled Clinical Decision Support System for Personalized Depression Treatment Selection and Management.
Dr. David Benrimoh and colleagues – including Douglas Researchers Drs. Manuela Ferrari and Anthony Gifuni – have recently examined the integration of artificial intelligence into depression care. Their study, published in The Journal of Clinical Psychiatry, investigates how an AI-enabled clinical decision support system can shape treatment selection and management for major depressive disorder.
Takeaway
An AI-powered clinical decision support system (CDSS) significantly improved remission rates and accelerated symptom improvement in patients with moderate to severe major depressive disorder (MDD), compared to standard guideline-informed care.
Background
MDD is a global health burden, and treatment often involves trial-and-error approaches that delay recovery. AI tools have shown promise in predicting treatment outcomes, but few have been tested in real-world clinical trials. The Aifred CDSS was developed to support clinicians by combining AI-generated remission probabilities for antidepressants with guideline-based treatment algorithms and measurement-based care.
Methods
This study was a multicenter, cluster-randomized controlled trial involving 47 clinicians across 9 sites in Canada and the U.S. Clinicians were randomized into two groups: an active group that used the Aifred CDSS and an active-control group that received guideline training and patient questionnaire data but did not use the CDSS.
A total of 74 adult outpatients with moderate to severe MDD were recruited. All patients had access to a patient portal to complete weekly questionnaires. Clinicians in the active group used the CDSS, which included:
- A clinical algorithm based on CANMAT guidelines.
- An AI module that predicted remission probabilities for 10 antidepressants.
- A dashboard for tracking patient progress.
Patients were followed for 12 weeks, with assessments at baseline, weeks 2, 4–6, 8, and 12. The primary outcome was remission, defined as a MADRS score <11 at study exit. Secondary outcomes included symptom response (≥50% reduction in MADRS), rate of symptom improvement, treatment adherence, and safety.
Results
Of the 74 patients, 61 completed at least two MADRS assessments and were included in the analysis. The remission rate was significantly higher in the active group (28.6%) compared to the control group (0%), with a P value of .01.
Patients in the active group also showed a faster rate of symptom improvement, with an average MADRS score reduction of 1.26 points per week versus 0.37 in the control group (P = .03). The total MADRS score reduction from baseline to study exit was 12.0 points in the active group and 4.9 points in the control group (P = .05), exceeding the threshold for clinical significance.
Treatment adherence was high in both groups (~95%), and no serious adverse events were attributed to the CDSS. Clinician and patient engagement with the platform was strong throughout the study.
Interpretation
The CDSS appears to enhance clinical decision-making by providing personalized treatment predictions and structured support for guideline-based care. Despite similar baseline characteristics and treatment options, patients in the active group experienced significantly better outcomes. These findings suggest that integrating AI into clinical workflows can improve the effectiveness and efficiency of depression treatment.
Conclusion
This study provides promising preliminary evidence that an AI-enabled CDSS can improve remission rates and accelerate symptom improvement in patients with moderate to severe MDD. While limited by sample size and early termination, the results support further research and potential clinical adoption of such tools.
About the study
“Artificial Intelligence in Depression–Medication Enhancement (AID-ME): A Cluster Randomized Trial of a Deep-Learning-Enabled Clinical Decision Support System for Personalized Depression Treatment Selection and Management” by David Benrimoh et al., published in The Journal of Clinical Psychiatry (2025).