AI to help personalise treatment for depression as part of major trial

The revolutionary digital tool, called PETRUSHKA, has been developed and is being trialled by a team at the University of Oxford’s Department of Psychiatry, and is funded by the National Institute for Health and Care Research (NIHR).

PETRUSHKA is an AI algorithm – informed by data from more than a million people – which recommends in real-time the best antidepressants for each individual, during the standard care consultation. The tool personalises antidepressant treatment using information from each patient, such as their age, gender and symptom severity, and incorporates their preferences about side effects. Researchers say this is an innovative way to empower patients and share decision-making during the treatment process.

In the UK 8.6 million people are prescribed antidepressants every year, but around 80 per cent of them stop the treatment after just a few weeks. This often happens because antidepressants are prescribed without a clear understanding of which drug is the most appropriate medication for each individual. So, people stop the antidepressant early because they are prescribed a drug which might work for an “average person”, but has not been tailored to them individually.

Professor Andrea Cipriani, Chief Investigator for PETRUSHKA and Lead of the NIHR Oxford Health Biomedical Research Centre Data Science Theme, says:

In real world practice, antidepressants are usually prescribed based on the clinician’s knowledge. There are more than 30 antidepressants on the market, but nine times out of 10, GPs prescribe one of just 4 antidepressants. This means that patients may not get the treatment that would be most effective or sustainable for them.

We want to treat individuals, not averages. We want to be as precise as possible and identify the best treatment for each patient sooner based on what we know works for patients with similar characteristics, rather than wasting time trying treatments that might be less effective.”

Patients can now sign themselves up to be part of the trial, rather than waiting for the GP to put them forward, and will go through a quick screening process via telephone. All visits will be with a consultant psychiatrist and can be done either in person or remotely. The trial is also being run internationally at facilities in Canada and Brazil.

The PETRUSHKA team hope to recruit around 200 more participants by the summer. The trial compares the PETRUSHKA tool with usual care and will involve a total of 504 participants.

Nyla Haque, PETRUSHKA Trial manager, added:

The study lasts 24 weeks in total, but after eight weeks we will see how many participants are still taking the allocated treatment. This measure will tell us how acceptable and well tolerated a treatment is.  We will also collect information from the participants about their mood, anxiety, quality of life and any side effects during the entire duration of the trial.”

Prospective participants must be aged between 18 and 74 years of age, with depression, and who are willing to try an antidepressant.