Decision Tool Reduces Antidepressant Dropout in Major Depressive Disorder
Key Clinical Summary
- A web-based decision-support system (PETRUSHKA) significantly reduced early antidepressant discontinuation in adults with major depressive disorder (MDD).
- In a multicenter randomized clinical trial, 17% of patients in the PETRUSHKA arm discontinued treatment by 8 weeks vs 27% with usual care (adjusted relative risk [RR], 0.62; P = .007).
- The tool also was associated with greater improvements in depressive and anxiety symptoms at 24 weeks.
A clinical decision-support system called PETRUSHKA (Personalizing Antidepressant Treatment for Unipolar Depression Combining Individual Choices, Risks, and Big Data) was found to improve antidepressant continuation in adults with major depressive disorder (MDD). The tool also improved depression and anxiety symptoms. Findings from the multicenter randomized clinical trial were published in JAMA.
Study Findings
The international trial included 540 adults aged 18 to 74 years diagnosed with major depressive disorder across 47 sites in Brazil, Canada, and the United Kingdom. Patients were randomized 1:1 to treatment guided by the PETRUSHKA tool or to usual clinical care. The primary outcome was treatment discontinuation for any reason at 8 weeks.
Of the 493 participants included in the primary analysis (median age 35 years; 58% female), discontinuation at 8 weeks occurred in 41 of 241 (17%) in the PETRUSHKA group compared with 69 of 252 (27%) in the usual care group (adjusted relative risk [RR], 0.62; 95% CI 0.44–0.88; P = .007). Discontinuation due to adverse events at 8 weeks was also lower with PETRUSHKA (9.1% vs 15.5%; adjusted RR 0.59; 95% CI 0.36–0.97; P = .04).
Secondary outcomes at 24 weeks demonstrated mean depressive symptom scores (PHQ-9) of 7.1 (SD, 5.4) with PETRUSHKA versus 9.2 (SD, 6.5) in usual care (adjusted between-group mean difference, −1.92 [95% CI, −3.06 to −0.78]; P < .001). Anxiety symptoms (GAD-7) were similarly improved (mean, 4.6 vs 5.8; adjusted between-group mean difference, −1.39 [95% CI, −2.26 to −0.52]; P = .002).
Authors note several study limitations, including lack of double-blind design, the use of individual rather than cluster randomization, PETRUSHKA lacking complete information on all available antidepressants, and missing information for secondary outcomes.
Clinical Implications
Major depressive disorder remains a leading cause of disability worldwide, and early discontinuation of antidepressant therapy is a persistent challenge that undermines treatment effectiveness. Personalized selection of antidepressants incorporating patient characteristics and preferences is recommended but often difficult to implement in routine practice.
The PETRUSHKA tool integrates clinical and demographic predictors with patient preferences to generate a ranked list of antidepressant options, potentially enhancing shared decision-making. In this trial, use of the tool was associated with a clinically meaningful reduction in discontinuation rates at 8 weeks—a critical early treatment period—and greater long-term symptom improvement at 24 weeks compared with usual clinician-guided care.
For clinicians, adopting validated decision-support systems could streamline personalized treatment planning and improve adherence, particularly in settings with high variability in antidepressant prescribing. Nonetheless, further studies with broader populations and rigorous blinded designs are needed to verify the generalizability and sustained benefits of such tools before widespread implementation.
Expert Commentary
According to the authors, “The PETRUSHKA tool was effective in personalizing antidepressant treatment and increased the likelihood of continuing treatment through 8 weeks.” They acknowledge that the absence of a double-blind design and missing data may influence interpretation but emphasize the potential for decision-support systems to enhance clinical practice.
Reference
