TheRapeutic Interactions in Psychological therapy Of anxiety and Depressive disorders (TRIPOD)

Background Despite the well-established efficacy of psychological therapies for anxiety and depressive disorders, effectiveness varies greatly between individual therapists, especially in routine practice. The most effective therapists have recovery rates 10 times the average. Conversely, the least effective therapists can actually make their patients' problems worse.

Differences between therapists grow where patients' problems are more severe or if therapy is short-term. Therapists also do not get more effective with time or experience.

Overall, the field lacks a systematic means of measuring and monitoring the content of psychological therapy for effective delivery. All tools used in practice have lengthy, resource-intensive coding processes, which makes routine use prohibitive.

Aims and objectives This project aims to develop an automated feedback tool to monitor and enhance the effectiveness of psychological therapies for anxiety and depressive disorders.

The project builds on a reliable, published tool used to assess patient-therapist interactions at the first sessions of psychological therapy, which could predict a range of long-term health outcomes. In three stages this project will: Automate the process of rating therapeutic interactions. Validate the prognostic power of the interaction rating tool when automated.

Translate these results into a usable feedback intervention for practising psychological therapists.

Methods Automate: A type of artificial intelligence called Natural Language Processing will be used to automate the interaction rating process and improve the predictive validity.

This has already started; current NIHR funding supports a multi-disciplinary team to use machine learning to begin the automated rating process and this will be refined during the clinical lectureship.

Validate: Prognostic modelling will then be used to assess the predictive validity of automated ratings of 400 patients' first session interactions in Cognitive Behaviour Therapy (CBT) and counselling for depression.

Translate: Using data from these prognostic models, a feasibility study will offer 20 practising NHS psychological therapists a training intervention where they are given feedback on the effectiveness of interactions in their psychological therapy sessions.

The clinical outcomes they achieve prior to the intervention will be compared to those achieved after the intervention to gain initial data on clinical effectiveness. Retention and compliance will also be assessed alongside therapist experience interviews to inform implementation.

Timelines for delivery The first 18 months will be split into an initial testing and refinement phase, followed by a validation modelling study.

The final 18 months will be used for feasibility testing and dissemination with the aim of being ready to start a full trial of the intervention after the funding ends.

Anticipated impact and dissemination The project includes a stakeholder panel of patients, psychological therapy practitioners, managers, trainers and trainees to ensure that the study is designed in a way that is applicable to all these parties with strong potential for implementation.

There will be annual dissemination events where results will be presented, which will begin implementation as study results are produced. Presentation at international conferences and peer reviewed publications will further support dissemination.

Local psychological therapy services are collaborators in the study to inform implementation, where appropriate, after the study.