A paper from the University of Southampton examining how best to treat psychosis has concluded that a greater range of individually targeted therapies could improve outcomes for patients. The research questions if cognitive behaviour therapy for psychosis (CBTp) should remain the dominant treatment and suggests that, in the future, big data and artificial intelligence may help to develop a range of more bespoke therapies.
CBTp was introduced in the 1990s and after evaluation in a large number of clinical trials, it became an established treatment for psychosis. Now, psychologists at the universities of Southampton and Sheffield have asked if less complex, less costly approaches may be as, or more, effective.
Lead author on the paper, professor Katherine Newman-Taylor of the School of Psychology at the University of Southampton, explained: “Our article asks whether CBTp benefits people with early psychosis and those with schizophrenia-related diagnoses in terms of clinical, functioning, and recovery outcomes. Also, for young people with mental health conditions who are at high risk of developing psychosis.
“While acknowledging the benefits CBTp can have for some, we wanted to consider if we should now look elsewhere to improve outcomes and if refining existing therapies could better meet the needs of people with psychosis.”
Psychosis is when a person perceives or interprets reality in a very different way from others. It may involve hallucinations, delusions, and disorganised thinking and speech. The term psychosis describes symptoms across a range of conditions but is typically associated with the diagnosis of schizophrenia.
Psychosis can lead to feeling scared, anxious, threatened, confused, and overwhelmed. CBTp works by helping people to make sense of their early life experiences, and current thoughts, feelings, and behaviours, for example when hearing voices or in the grip of paranoia. Therapy involves working collaboratively to build the person’s ability and confidence that they can do what’s important to them, even if the voices, paranoia, and other symptoms of psychosis persist.
The Southampton and Sheffield researchers examined two umbrella reviews conducted by other researchers in 2019 and 2023. An umbrella review provides a very high level analysis of a wide range and large number of past research papers to help reach conclusions about a topic or issue.
The team used these recent umbrella reviews to give a “bird’s eye view” of the effectiveness of CBTp to treat psychosis in different groups of people. Their findings are published in Psychology and Psychotherapy: Theory, Research and Practice. The paper concludes that large-scale analysis of treatment outcomes from pooled data is masking important nuances. While many are benefitting from CBTp, some patients only experience modest outcomes and others may be harmed by it.
The team says that by focusing on the therapeutic relationship and particular processes, such as worry and past trauma, clinicians would be able to help people more effectively. They also propose the development of large datasets, interpreted by sophisticated AI machine-learning tools, to help aid decisions about treatments. These may include CBTp alongside other approaches, such as working with the whole family, and setting up informal peer support networks early in the treatment process.
Newman-Taylor concluded: “We predict that over the next 10 years, large, continually evolving datasets, built from patient experience, will be used to shape precision psychological therapies. Using data to determine treatment outcomes will help us to choose the right evidence-based therapy for an individual. However, it is vital that we use these methods to make decisions jointly with patients, and only work with organisations who we trust to manage our health data securely and ethically.”