Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.

Digital interventions can be an important instrument in treating substance use disorder. However, most digital mental health interventions suffer from early, frequent user dropout. Early prediction of engagement would allow identification of individuals whose engagement with digital interventions may be too limited to support behaviour change, and subsequently offering them support. To investigate this, we used machine learning models to predict different metrics of real-world engagement with a digital cognitive behavioural therapy intervention widely available in UK addiction services. Our predictor set consisted of baseline data from routinely-collected standardised psychometric measures. Areas under the ROC curve, and correlations between predicted and observed values indicated that baseline data do not contain sufficient information about individual patterns of engagement.

Original publication




Journal article


Studies in health technology and informatics

Publication Date





967 - 971


Centre for Health Informatics, School of Health Sciences, University of Manchester, UK.


Humans, Substance-Related Disorders, ROC Curve, Mental Health, Machine Learning