21 August 2026
Hyfe's CoughMonitor Suite captured continuous, objective cough data on standalone wrist-worn watches, with no study phone or mobile network required during each monitoring window, at a fraction of the setup and staff burden of the team's other digital health tools.
Partner: University of California, San Francisco (UCSF) — Sophie Huddart, PhD, MSc, with UC Irvine and WALIMU (Uganda)
Study type: Prospective pilot cohort, decentralized continuous cough monitoring, nested in the R2D2 TB Network
Therapeutic area: Tuberculosis / respiratory infectious disease
Setting: Primary care centers, Kampala, Uganda
Platform: Hyfe CoughMonitor Suite (CMS)
Key outcome:
In a decentralized TB pilot in Kampala, Hyfe's CoughMonitor Suite captured continuous, objective cough data on standalone wrist-worn watches, with no study phone or mobile network required during each monitoring window, at a fraction of the setup and staff burden of the team's other digital health tools.
The team needed objective cough data from people beginning TB treatment, captured at two short windows: at treatment initiation and again around day 30. The work had to run in real-world primary care in Kampala, not a controlled clinic, and it had to fit alongside the demands of a busy research program.
The central question for this pilot was practical rather than scientific. The team had used a range of digital health tools for research data collection, and wanted to see how the new wrist-worn cough watches would fare in their setting. Specifically: could the watches capture high-quality data on their own, with no connection to a mobile network or a paired phone, across each three-day observation window?
Hyfe deployed the CoughMonitor Suite. Participants wore the watch continuously for at least three days at each of two timepoints, treatment initiation and approximately day 30. The watch passively detects sounds consistent with cough and records only the timestamp of each event, so no audio leaves the device.
The distinctive part was how little the deployment needed to work. The watches were sent out on participants and ran standalone for the full observation window, with no paired study phone and no mobile-network connection required during monitoring, then returned for the team to extract data. That removed the phones, chargers, and hands-on staff time that the team's other tools depended on.
Hyfe supported the deployment with training sessions and SOPs, and stayed responsive through the study: when participants said they wanted the watch to display the time so they would not have to wear a second watch, Hyfe pushed that software update within a week. The team used the CoughMonitor Suite dashboard to check data quality and wear time, confirm successful data transfer, and monitor collection across all of their sites.
| Metric | Result |
|---|---|
| Participants with cough data | 40 |
| Monitoring windows per participant | 2 (treatment initiation and ~day 30) |
| Minimum continuous wear per window | 3 days |
| Total monitoring days captured | 462 |
| Total coughs captured | 112498 |
| Mean / median daily wear time | Mean: 11 hours |
| Successful data transfer / data quality | All watches synced and no recording sessions were lost |
We found the wearable cough monitor easy to deploy in our study setting, requiring limited infrastructure and staff time to get up and running. Our participants had no trouble wearing the devices while going about their daily activities and we received high-quality data from the get-go."
Study Investigator
UCSF
For sponsors and CROs running decentralized or multi-site respiratory studies, this pilot shows that objective cough monitoring can be stood up in real-world, low-resource settings with minimal infrastructure and staff time. Because the watches run standalone, with no paired phone or network dependency during monitoring, the deployment carried low operational overhead and, in the team's experience, a very low participant refusal rate, making it feasible to monitor larger numbers of participants at once. Objective, quantitative cough data sits alongside patient-reported outcomes to give a fuller picture of participant health, and the possibility of real-time data syncing opens the door to future study designs that use cough data for real-time, adaptive decision-making.
