PREACT-digital: study protocol for a longitudinal, observational multicentre study on digital phenotypes of non-response to cognitive behavioural therapy for internalising disorders

Jul 10, 2025BMJ open

Tracking digital signs of poor response to cognitive behavioral therapy for internalizing disorders over time in multiple centers

AI simplified

Abstract

An estimated 468 patients will participate in the PREACT-digital study, which aims to identify predictors of non-response to cognitive behavioural therapy (CBT) for internalising disorders.

  • The study combines with to monitor daily variations in factors like mood and physical activity.
  • Dynamic markers from these assessments may help predict which patients do not respond to CBT.
  • Data collected includes heart rate, sleep patterns, and emotional regulation strategies.
  • Predictive analyses will use both basic algorithms and advanced machine learning techniques to capture complex patterns.
  • Exploratory analyses will investigate the relationships between variables over time and assess changes in symptoms.

AI simplified

Key numbers

468
Expected Sample Size
Total number of patients anticipated to participate in PREACT-digital.
350
Long Version Participants
Number of patients expected to choose the long version of the study.
>75%
Predictive Accuracy Goal
Target predictive accuracy for identifying non-response after 20 therapy sessions.

Full Text

What this is

  • PREACT-digital is a longitudinal, observational study focused on predicting non-response to cognitive behavioural therapy (CBT) for internalising disorders (ID).
  • It combines () and to gather real-time data on patients' emotional and physical states.
  • The study aims to identify dynamic markers that can inform personalized treatment approaches and improve therapy outcomes.

Essence

  • PREACT-digital seeks to enhance the prediction of non-response to CBT in patients with internalising disorders by utilizing real-time data collection methods. The study integrates and to capture daily fluctuations in emotional and behavioral states, aiming to inform personalized therapy.

Key takeaways

  • The study will involve approximately 468 patients, with 350 expected to complete a long-term version of the study. Participants will provide data through a smartwatch and a customized app, enabling continuous monitoring of their emotional and physical conditions.
  • The central goal is to achieve a predictive accuracy greater than 75% for identifying non-response to therapy after 20 sessions. This will be assessed using a combination of data and features collected during a 14-day assessment phase.
  • Findings from this study aim to contribute to the development of more personalized and effective CBT approaches, potentially allowing for real-time interventions based on patients' data collected through wearable technology.

Caveats

  • The absence of a healthy control group limits the ability to distinguish digital patterns specific to clinical populations from those present in non-clinical individuals. Future studies may need to incorporate control groups for better specificity.
  • The extensive assessment schedule may burden participants, potentially affecting adherence and retention in the study. Managing participant engagement will be crucial to ensure the robustness of the data collected.

Definitions

  • Ecological Momentary Assessment (EMA): A method involving real-time data collection through repeated self-reports, capturing individuals' experiences in their natural environments.
  • Passive Sensing: Continuous data collection on behavioral and physiological metrics using wearable devices or smartphones, without active input from the user.

AI simplified

what lands in your inbox each week:

  • 📚7 fresh studies
  • 📝plain-language summaries
  • direct links to original studies
  • 🏅top journal indicators
  • 📅weekly delivery
  • 🧘‍♂️always free