Improving Treatment Adherence in Major Depressive Disorder via Precision Psychiatry

Article Impact Level: HIGH
Data Quality: STRONG
Summary of  JAMA https://doi.org/10.1001/jama.2026.1327 
Dr. Andrea Cipriani  et al.

Points

Researchers evaluated the PETRUSHKA AI tool in a trial of over five hundred adults to determine if personalized antidepressant selection improves adherence compared to standard psychiatric care.
The study found that patients using the digital tool were forty percent less likely to stop their medication during the first eight weeks of treatment due to side effects.
Longitudinal results at twenty-four weeks showed that participants in the intervention group experienced significantly greater improvements in both depressive and anxiety symptoms than those in the control group.
The tool integrates clinical evidence with specific patient preferences in a three-minute consultation to support shared decision-making in both primary care and specialized mental health settings.
While the trial demonstrated clear benefits for treatment continuation investigators acknowledged that missing data and the open-label study design are factors that require further long-term validation.

Summary

This research evaluated the clinical utility of PETRUSHKA, an AI-driven decision-support tool designed to personalize antidepressant selection for patients with major depressive disorder (MDD). The study addressed the high attrition rates common in psychiatric care, which are often driven by sub-optimal drug matching and intolerable side effects. By integrating clinical and demographic variables with individual patient preferences, the tool aims to transition antidepressant prescribing from a trial-and-error approach to a precision psychiatry model.

In a randomized clinical trial involving over 500 adults across the UK, Brazil, and Canada, the use of PETRUSHKA significantly improved treatment adherence. Participants in the intervention group were approximately 40% less likely to discontinue their medication within the first eight weeks of treatment compared to those receiving usual care. Longitudinal data at 24 weeks indicated superior mental health outcomes, with patients reporting greater reductions in both depressive and anxiety symptoms. The tool’s three-minute administration time suggests high feasibility for integration into time-constrained primary care environments.

Despite these positive outcomes, investigators noted that the lack of a double-blind design and a substantial volume of missing data may limit the generalizability of the results. However, the reduction in discontinuation due to adverse effects suggests that the tool successfully prioritizes patient-specific tolerability profiles. As the first demonstrated effective clinical prediction tool in mental health, PETRUSHKA represents a shift toward shared decision-making. Future research is required to evaluate long-term cost-effectiveness and the tool’s impact on specialized psychiatric resource allocation.

Link to the article: https://jamanetwork.com/journals/jama/article-abstract/2845910

References

Cipriani, A., Fernandes, K. B. P., Mulsant, B. H., Efthimiou, O., Williams, N., Mort, S., Elgarf, R., Liu, Q., Haque, N., Potts, J., Ede, R., Fox, R., Liboni, M., Nesi Cavicchioli, D. A., Simon, J., Smith, K. A., Zangani, C., Li, Z., Taylor, U., … Tomazini, V. F. (2026). A decision-support system to personalize antidepressant treatment in major depressive disorder: A randomized clinical trial. JAMA. https://doi.org/10.1001/jama.2026.1327