Scientists in the United Arab Emirates (UAE) are working on a COVID-19 breathalyser test that can return a result in less than a minute. Adel Al Redha, Emirates’ Chief Operating Officer of Emirates Airline says such as device would “hugely benefit the re-opening of international travel and other economic activity,” by taking pressure off specialist labs and giving governments the confidence to ease border restrictions and social distancing rules.
Similar tests have been developed by researchers from Israel and elsewhere but so far these non-invasive and incredibly fast tests have failed to achieve the kind of accuracy as that offered by gold-standard polymerase chain reaction (PCR) tests or even rapid antigen and LAMP tests.
A PCR test can take several hours to process, a LAMP test takes around an hour and a rapid antigen test up to 30 minutes. A breathalyser test would, however, take less than 60 seconds and be far less invasive that most other tests that require swabs of the throat and nasal passage.
“A person only needs to blow into a disposable mouthpiece connected to a high-precision breath sampler,” the official WAM news agency explained. “The exhaled breath is collected into a breath sampler and fed into a cutting-edge mass spectrometer for measurement. A machine learning software analyses the volatile organic compound (VOC) profile and generates the result in less than a minute.”
The main problem, however, is that a pilot study of the technology achieved a sensitivity rate of 93 per cent and specificity of 95 per cent with its machine learning algorithm. Sensitivity refers to the number of true positives the test picks up, while specificity refers to the number of false positives the test might return.
According to the respected medical journal the BMJ, if 6 in 100 people actually had COVID-19, this test would find all the positive cases but would also incorrectly flag 5 people as testing positive.
Developed by Singapore-based Breathonix, the device is now being put through its paces in a larger clinical study of 2,500 people in Dubai to see whether its machine learning algorithm can further improve the accuracy of the test.
If the trials prove successful, the device could win the backing of Dubai’s health authority and then be pushed into action to quickly screen airline passengers for COVID-19.
