New smartphone app could help diagnose tuberculosis and monitor treatment

Scientists recently investigated a smartphone-based app that records cough frequency and found it useful in identifying tuberculosis (TB) and tracking treatment success.

Cough is a primary symptom of many respiratory diseases, including tuberculosis (TB). Although treatable, monitoring the success of TB treatment is often a challenge for healthcare providers.

For example, the research points out that current TB treatment monitoring relies on sputum and culture conversion testing, which have low sensitivity and long turnaround times, are biohazards, and are prone to contamination.

Tracking the frequency of coughing can help doctors monitor the effectiveness of treatments, but this requires the person with TB to report exactly how many times they cough each day. The results can be challenging and inconclusive.

However, advances in smartphone app technology have opened new doors for diagnosing and monitoring certain medical conditions. For example, recent reports suggest that apps can detect strokes when they happen, and some apps can help diagnose autism.

The latest application development examined during clinical trials is the cough monitoring system called Hyfe AI. This smartphone-based app runs special software that calculates your cough frequency every hour. With this information, health care providers can have the data needed to diagnose TB and determine whether TB treatment is effective.

To test the utility of Hyfe AI, researchers recently conducted a study using the new app technology to monitor the coughs of people with suspected TB in Uganda, South Africa, the Philippines, Vietnam and India. The scientists then compared the cough patterns of people with microbiologically confirmed TB, clinical TB and other respiratory disease (ORD).

Their research is published in The International Journal of Tuberculosis and Lung Disease (IJTLD).

At the start of the study, participants aged 18 or older with a new or worsening cough lasting at least two weeks were given a smartphone with the Hyfe app installed. The application monitored the participant’s cough 24 hours a day for 14 days.

Of the 565 participants who completed cough monitoring, 144 had microbiologically confirmed TB, 48 were treated for clinical TB, and 373 had ORD.

Of the 144 participants with microbiologically confirmed TB, almost 94% (135) started TB treatment approximately one day after entering the study.

On the first day of observation, the average number of coughs was five coughs per hour. However, those with microbiologically confirmed TB had significantly higher hourly cough rates than those with ORD.

Despite this, the scientists found no significant difference in hourly cough frequency between those with clinical TB.

Additionally, by day 14 of cough monitoring, the overall hourly cough frequency in all groups had decreased to 3.5 – a significant decrease from the first day of the study.

Participants with TB showed unique cough frequency trajectories at the start of treatment that differed from the cough patterns of participants with ORD. In addition, cough frequency improved more rapidly among participants with microbiologically confirmed TB.

Although more research is needed, scientists suggest that cough-based TB screening could help reduce overtreatment of people without TB and improve detection of the condition.

“Continuous cough monitoring should be further investigated as a noninvasive biomarker for TB diagnosis and treatment monitoring,” the authors wrote.

A clinical trial investigating the performance of the Hyfe Cough Monitoring System is currently underway with an expected completion date of April 3, 2023.