Henry M.G. Glyde1Alison M. Blythin2 Tom M.A. Wilkinson3Ian T. Nabney4
James W. Dodd5
- 1. EPSRC Centre for Doctoral Training in Digital Health and Care, University of Bristol, Bristol, UK
- my mHealth Limited, Bournemouth , UK
- my mHealth and Clinical and Experimental Science, University of Southampton, Southampton, UK
- 4. School of Engineering Mathematics and Technology, University of Bristol, Bristol, UK
- 5. Academic Respiratory Unit, Translational Health Sciences, Bristol Medical School, University of Bristol, Bristol, UK
Abstract
Background
Acute exacerbations of COPD (AECOPD) are episodes of breathlessness, cough and sputum which are associated with the risk of hospitalisation, progressive lung function decline and death. They are often missed or diagnosed late . Accurate timely intervention can improve these poor outcomes. Digital tools can be used to capture symptoms and other clinical data in COPD. This study aims to apply machine learning to the largest available real-world digital dataset to identify AECOPD Prediction tool which could be used to support early intervention improve clinical outcomes.
Objective
To create and validate a machine learning predictive model that forecasts exacerbations of COPD 1-8 days in advance. The model is based on routine patient-entered data from myCOPD self-management app.
Method
Adaptations of the AdaBoost algorithm were employed as machine learning approaches. The dataset included 506 patients users between 2017-2021. 55,066 app records were available for stable COPD event labels and 1,263 records of AECOPD event labels. The data used for training the model included COPD assessment test (CAT) scores, symptom scores, smoking history, and previous exacerbation frequency. All exacerbation records used in the model were confined to the 1-8 days preceding a self-reported exacerbation event.
Results
TheEasyEnsemble Classifier resulted in a Sensitivity of 67.0% and a Specificity of 65% with a positive predictive value (PPV) of 5.0% and a negative predictive value (NPV) of 98.9%. An AdaBoost model with a cost-sensitive decision tree resulted in a a Sensitivity of 35.0% and a Specificity of 89.0% with a PPV of 7.08% and NPV of 98.3%.
Conclusion
This preliminary analysis demonstrates that machine learning approaches to real-world data from a widely deployed digital therapeutic has the potential to predict AECOPD and can be used to confidently exclude the risk of exacerbations of COPD within the next 8 days.
Permission to use received from Henry Glyde
