People worried they might have COVID-19 could soon use a simple phone app to see if they are positive for the virus.
Early results of ResApp's trial of more than 700 people have been released but are still to be peer-reviewed.
The trial, which was funded by the developer, found if people coughed into the app it could detect COVID-19 in 92 per cent of cases.
The trial included people with both the Delta and the Omicron strain of the virus and the screening was conducted across homes, testing clinics and hospitals.
A spokesman for the company said the results exceeded "real-world measured sensitivity of rapid antigen testing", which refers to the RAT results achieved when people do the tests at home.
The Brisbane company has already received regulatory approval for smartphone apps to screen cough sounds to diagnose respiratory illnesses including pneumonia and asthma, and the breathing of someone with sleep apnoea.
Smartphone screening for COVID-19 could increase access to testing, particularly in areas where it is difficult to access RAT and PCR tests.
It could also reduce parents' concerns about the challenges of performing nasal or throat swabs on children, particularly those with disabilities such as autism.
Value in ruling out COVID-19
The US-based COVID-cough pilot clinical study aimed to collect data to train an algorithm to identify COVID-19 through cough sounds recorded on a smartphone.
It used a gold-standard PCR pathology test as a reference standard.
The trial was expanded to include patients in India last year and included 741 patients, among which 446 were COVID-19 positive.
A company spokesperson said the value in the app was its potential to accurately rule out COVID-19, reduce the need for PCR and RAT tests and improve access to surveillance testing.
Technology will improve telehealth
Lung Foundation chairperson Lucy Morgan said using cough signatures to diagnose disease could prove a valuable tool for telehealth.
"This sort of technology is lifting the virtual consultation out of a phone call and improving the quality of a medical assessment," she said.
"None of these tests are perfect, so this is a very simple, potentially very inexpensive way of helping a clinician and patient to rule out COVID as a cause of cough that will really help us worldwide to improve our diagnostic rates and the way we care for people with COVID," Dr Morgan said.
She said more information was needed about how the app would work if a patient had multiple diseases.
"What happens if you have a chronic lung disease … where you cough all the time and then you get COVID on top of that, can the app determine the cause of a worsening cough?"
Global benefits
Advisor to ResApp Health and chair of epidemiology at Deakin University Catherine Bennett said the technology promised to make screening more accessible by providing a telehealth diagnostic tool that was not reliant on supply or laboratory capacity.
"That's really important globally as we look at diagnostics and screening of people for COVID and knowing that's something you can scale up to population level quite quickly," she said.
Professor Bennett said she expected results from people using the app at home would be similar to those achieved in a clinical setting.
"It's a level of reliability that previously we've only really expected from a test in the hands of a health professional or specifically from a PCR test.
"We know with rapid antigen tests used at home, you don't get the same sensitivity and specificity that you get in the trials, it drops away."
Trial diversity important
Respiratory medicine expert Guy Marks said more data was needed but he welcomed new ways to test for COVID-19.
"To reduce the amount of more expensive testing that we need to do with a preliminary test would be useful and there is history of listening to breath sounds and cough sounds to attempt to make a diagnosis," Professor Marks, from the University of New South Wales, said.
However, he said more information was needed about the people in the clinical trial.
Professor Marks said it was important the trial included people who were asymptomatic, had minor and severe symptoms as well as people with other health conditions.
"We need more information about the performance of this test in a diverse range of patients," he said.
