A new digital tool designed to personalise antidepressant treatment for individuals with depression could transform the way medication is prescribed. The PETRUSHKA tool, developed as part of a large-scale international study, uses a data-driven approach to help clinicians and patients choose the most suitable antidepressant based on personal preferences and clinical factors. The findings were published in the Canadian Journal of Psychiatry.
Depression affects millions worldwide, yet the process of prescribing antidepressants often relies on a trial-and-error approach. Many patients are prescribed medications that may not be effective for them or cause distressing side effects, leading to high discontinuation rates. The PETRUSHKA tool seeks to address this by incorporating patient-specific data and predictive modelling to provide tailored recommendations.
Developed by researchers at the University of Oxford and other international institutions, the tool gathers information from clinical trials and real-world health records to generate a ranked list of suitable antidepressants. It factors in the patient’s medical history, previous treatment experiences, and preferences regarding side effects. The goal is to support shared decision-making between clinicians and patients, ensuring that treatment choices align with individual needs.
The PETRUSHKA trial was conducted across multiple sites in the UK, Canada, and Brazil. Patients with a diagnosis of unipolar depression participated in the study, with some receiving treatment recommendations through the tool while others followed usual care procedures. The primary aim was to determine whether those using the tool were more likely to adhere to their prescribed antidepressant compared to those receiving standard care.
Preliminary findings indicate that patients who used the PETRUSHKA tool were more likely to continue with their prescribed medication for at least eight weeks. This suggests that personalised recommendations may improve adherence and treatment outcomes. The tool also provides visual representations of the likelihood of different side effects, helping patients make informed decisions about their treatment.
Despite its promise, the tool has some limitations. It only provides recommendations for antidepressant monotherapy, meaning it does not cater to patients requiring combination treatments or those with treatment-resistant depression. Additionally, some antidepressants are not included in its algorithm due to limited access to clinical trial data. Researchers acknowledge these constraints and suggest that future updates could incorporate genetic data and other biomarkers to refine the tool’s predictions further.
The move towards precision psychiatry has gained momentum in recent years, with experts calling for more personalised approaches to mental health treatment. Traditional prescribing methods often overlook individual differences in drug metabolism and response, which can lead to prolonged periods of ineffective treatment. By leveraging machine learning and large-scale clinical data, the PETRUSHKA tool represents a step towards more targeted interventions.
