A new GC-IMS breath test has been developed that can detect COVID-19 with 80% accuracy. The results are available almost immediately, making it a potential on-site test. Previously, diseases such as diabetic ketoacidosis could be diagnosed by the smell of exhaled air. Diabetic alert dogs can even detect the metabolic disorder in its early stages and warn patients. However, this type of diagnosis has not been used for COVID-19. Some dogs can detect COVID-19 in the urine of infected people, but the long training time makes it difficult to use them clinically.

Researchers from the University of Edinburgh and the Dortmund Clinic have therefore investigated whether volatile organic compounds (VOCs) produced during lung infection with SARS-CoV-2 can be detected in exhaled air using gas chromatography. The independent research groups used an identical measuring instrument that combines ion mobility spectrometry and gas chromatography. In addition to COVID-19 patients, breath analyses were also performed on people with other respiratory diseases such as COPD and asthma, as well as bacterial or viral respiratory infections.

Compared to previous methods, the test procedure is simple and comparable to an alcohol test. The air blown through a small disposable collection tube is taken with a syringe and introduced into the GC-IMS. The analysis of the sample is carried out without additional reagents. Results are available within a few minutes. According to the study authors, this method is also suitable as an on-site rapid test. Aerosols are not released, so the risk of infection for the testing personnel is relatively low.

The accuracy of the test was 80% in Edinburgh and 81.5% in Dortmund. Further studies with larger groups of subjects will now verify the accuracy of the test procedure and investigate whether the results are influenced by other factors such as people’s diet. This new test could be a game-changer in the fight against COVID-19, providing a quick and easy way to detect the virus on-site.

Leave a Reply

Your email address will not be published. Required fields are marked *