Sound-based cough peak flow estimation in patients with neuromuscular disorders

Introduction/Aims: Cough impairment is common in individuals with neuromuscular disorders and is associated with respiratory infections and shorter survival. Cough strength is assessed by measuring cough peak flow (CPF) using a flow meter, but this method requires a complex device setup and trained staff. The aim of the study is to evaluate the reliability of a smartphone app to estimate CPF based on cough sounds in a cohort of individuals with neuromuscular disorders. Methods: Individuals with neuromuscular disorders underwent CPF measurement with a flow meter and a smartphone app. A CPF <270 L/min was considered abnormal. Results: Of the 50 patients studied, 26 had amyotrophic lateral sclerosis (52%), 15 had hereditary myopathies (30%), and 9 had myasthenia gravis (18%). The intraclass correlation coefficient (ICC) between the CPF measured with a flow meter and CPF estimated with cough sounds was 0.774 (p <.001) even if the patients had orofacial weakness (ICC = 0.806, p <.001). The smartphone app had 94.4% sensitivity and 100% specificity to detect patients with CPF of less than 270 L/min. Discussion: Our findings suggest that sounds measured with a smartphone app provide a reliable estimate of CPF in patients with neuromuscular disorders, even in the presence of with orofacial weakness. This may be a convenient way to monitor respiratory involvement in patients with neuromuscular disorders, but larger studies of more diverse patient cohorts are needed.