S140 Predictors of 24-h cough frequency in acute cough
- 1Division of Asthma, Allergy and Lung Biology, King's College London, London, UK
- 2Institute of Electronics and Telematics Engineering (IEETA), University of Aveiro, Aveiro, Portugal
- 3Department of Medical Physics, Leicester Royal Infirmary, Leicester, UK
- 4Department of Respiratory Medicine, Institute for Lung Health, Glenfield Hospital, Leicester, UK
Introduction 24-hour cough frequency monitoring is increasingly being used as an outcome measure to evaluate anti-tussive drugs. The optimal method of identifying patients with a significant cough frequency for inclusion into clinical trials is not known. We investigated a range of cough assessments screening tools that could be used for this purpose.
Methods 35 healthy subjects with acute cough due to upper respiratory infection (median (IQR) age 31 (23–35) years, 63% female, median (IQR) duration of cough 4 (2–6) days) were recruited as part of a larger study. All subjects underwent ambulatory 24-h cough frequency (CF24) monitoring with the Leicester Cough Monitor, health related quality of life with the Leicester Cough Questionnaire-acute (LCQ) and cough severity visual analogue scale (VAS). Receiver operating characteristic curve analyses were performed for the baseline screening tools LCQ, VAS and cough frequency in first hour (CF1) to identify patients with CF24=5, =7.5 and =10 coughs/hr.
Results The baseline geometric mean (logSD) CF24 was 14.7 (0.5) coughs/hr, CF1 25.9 (0.4) coughs/hr, mean (SEM) VAS 47 (3) mm and LCQ 14.3 (0.7). 4, 6 and 11 patients had CF24 <5, <7.5 and <10 coughs/hr respectively. The area under ROC curve (AUC) for VAS and LCQ for prediction of CF24 were poor, ranging from 0.51 to 0.58 and 0.50 to 0.68 respectively. The mean (SEM) AUC for CF1 was 0.92 (0.05), 0.79 (0.10) and 0.86 (0.06) for detecting 24-h cough frequency of =5, =7.5 and =10 coughs/hr respectively. The sensitivity and specificity for cough frequency recordings over 1 h to identify patients with significant 24-h cough frequency are presented in Abstract S140 table 1. A good specificity (to exclude patients with low 24-h cough frequency) was achievable at sensitivities ranging from 69 to 87%.
Conclusion Subjective cough severity and cough-specific quality of life are poor screening tools for predicting objective 24-h cough frequency. However, 1-h cough frequency can be used as a screening tool to identify patients with significant 24-h cough frequency for inclusion into future clinical trials.