Abstract
BACKGROUND AND OBJECTIVE: International guidelines emphasise the importance of objective cough assessment for evaluating therapeutic efficacy. Cough negatively impacts quality of life and generates high costs for healthcare systems. However, the accuracy of cough frequency monitors remains unclear. This study aimed to identify tools developed to monitor cough frequency and evaluate their accuracy. METHODS: Observational studies involving adults, children, or infants who used cough frequency monitors were included from the Pubmed/MEDLINE, EMBASE, and Web of Science databases published up to 24 November 2024. Two independent reviewers performed the screening, data extraction, and quality assessment of the studies using QUADAS-2. Summary estimates of diagnostic accuracy were calculated with meta-analysis using bivariate mixed effects regression. RESULTS: Nineteen studies were included in the review. In the 16 studies included in the meta-analysis with 52,612 events, sensitivity was 89% (95% CI 0.84-0.93, I (2) 89.3%) and specificity was 99% (95% CI 0.98-1.00, I (2) 88.2%). The area under the curve (AUC) of 0.96 indicated excellent discriminative capacity. In the 11 devices and 4 applications identified, individual accuracy was > 0.9. CONCLUSION: The tools are accurate in distinguishing between cough and non-cough sounds, but their clinical application still presents technical challenges and requires investment. Future research should improve the monitors, maintaining high accuracy but allowing for agile, cost-effective assessment.