The use of cough frequency monitors effectively discriminates between cough and non-cough sounds, results of a systematic review and meta-analysis have shown. However, their clinical application remains technically challenging and will require investments.
“At present, their restricted use in research environments and lack of investment make expansion into clinical practice technically challenging,” the investigators said.
PubMed, Embase, and Web of Science databases were accessed for observational studies involving adults, children, or infants who used cough frequency monitors published up to 24 November 2024. Two independent reviewers screened the studies, extracted data, and assessed study quality.
The investigators calculated summary estimates of diagnostic accuracy with meta-analysis using bivariate mixed effects regression.
Nineteen studies met the eligibility criteria, and 16 were included in the meta-analysis. A total of 52,612 events were examined. The review identified 11 devices, three mobile applications (apps), and one app that can be installed on a mobile phone or computer. [Respirology 2026;31:459-470]
The cough frequency monitors had a sensitivity of 89 percent (95 percent confidence interval [CI], 0.84‒0.93; I2, 89.3 percent) and specificity of 99 percent (95 percent CI, 0.98‒1.00; I2, 88.2 percent). The area under the curve was 0.96, suggesting excellent discriminative capacity. Individual accuracy in the 11 devices and four apps examined was >0.9.
The meta-analysis revealed “excellent accuracy of the cough frequency monitors,” the investigators said. “The individual accuracy of the studies did not demonstrate inferiority in any of the devices evaluated.”
Customization
Peak cough flow varies from one patient to another, depending on age, gender, and height. Volitional cough in healthy individuals may yield higher peak values, especially when compared with those of patients with neuromuscular diseases. Such difference may justify the slightly higher AUC in the subgroup of healthy individuals, according to the investigators. [Chest 1993;104:1553-1562]
An earlier study identified customization of cough frequency monitors based on the health status of an individual. This approach can aid in the development of devices or apps that monitor cough and diagnose diseases causing cough or act as a marker of severity. [J Thorac Dis 2020;12:5207-5223]
“The potential adaptation of mobile devices through validated applications for cough monitoring is therefore very attractive,” the investigators said. “However, cough processing stages must be considered.”
Different recording methods can affect the sensitivity of the cough monitor when the device was evaluated separately, according to the investigators. [J Clin Monit Comput 2025;39:157-167; Comput Biol Med 2018;100:176-185; JMIR Form Res 2023;7:e38439]
Clinical use
“To our knowledge, no cough frequency monitor has been approved by a regulatory agency for clinical use,” the investigators said.
One cough monitoring device, VitaloJAK, had been validated, with a sensitivity ranging from 99.2 percent to 100 percent. However, this tool was not included in the current review because the validation study was a conference abstract with no complete data for analysis. [Thorax 2012;67:A131]
“Some monitors, such as VitaloJAK, are approved for sound recording, but not as validated cough frequency monitors,” the investigators said. “This approval does not guarantee the accuracy of the software for drug evaluation or for decision-making in clinical practice.”
Further studies are warranted to improve the accuracy of cough frequency monitors, “with the goal of enabling clinical availability for patients with cough through the development of devices or mobile apps that are easy to use, cost-effective, and monitor coughing reliably and in less time,” according to the investigators.