A novel alert system for impending acute exacerbation of COPD based on cough monitoring

Abstract

Introduction: Acute exacerbations of COPD (AECOPD) often lead to health service utilisation and have associated morbidity and mortality. Telemonitoring is proposed to facilitate early treatment by identifying and alerting clinicians to deterioration; but existing strategies have proven disappointing. Cough commonly increases at the time of AECOPD. We have previously described a system to monitor cough trends over time. Aim: To assess cough count around AE-COPD and develop a system to alert clinicians to a meaningful change in cough frequency. Methods: This prospective longitudinal study of cough monitoring included COPD patients that had ≥2 AECOPD within the last year. Participants underwent continual cough monitoring and completed a daily questionnaire for 90 days. AECOPD were identified by self-reporting, prescribing and hospital attendance data. A novel alert system based on cough monitoring was compared with a symptom based strategy. Results: 28 patients completed the study (16 males; median age 66 [range 46-86]; mean [SD] FEV-1 % predicted 36 [18]%; CAT score 28 [8]) and had a total of 43 moderate and 15 severe AECOPD. Mean cough count increased during AE-COPD. Alerts based on cough monitoring occurred in the prodrome of 45% of AE-COPD with 1 false alert every 100 days. Questionnaire based alerts occurred in the prodrome of 88% of AE-COPD with 1 false alert every 10 days. Conclusion: Cough monitoring can detect change in cough frequency around AECOPD. The low false alert rate with cough monitoring means it has the potential to be a clinically useful tool to identify an impending AE-COPD and facilitate an early response. The utility of symptom monitoring is limited by frequent false alerts.