Wash your hands with hydroalcoholic gel, smell it and access a QR code to answer a short questionnaire. These simple actions constitute the world’s first patented mass screening system for covid cases. A research group from the Rovira i Virgili University (URV) and the Pere Virgili Health Research Institute (IISPV) developed this method, which bases its operation on artificial intelligence techniques. The model determines on site which people are currently at low, medium, or high risk of having the disease, and the results offer a sensitivity of 97%. The first prototype of this device is being used in the emergency department of Hospital Universitari Sant Joan de Reus. The research results were published in the journal Scientific Reports.

The aroma of citrus fruits or apples are two of the first to be lost when the sense of smell is affected due to infection by the SARS-CoV-2 virus. Anosmia, that is, the loss of the ability to detect odors, has been one of the characteristic symptoms of covid since the beginning of the pandemic. But it’s not the only thing. Fever, headache, cough, malaise, sore throat… these are signs compatible with the coronavirus, but they can also correspond to a cold or the flu. How to know, without any diagnostic test, which virus is behind it?

The patented system is based on a hydroalcoholic gel to which a specific concentration of citrus essence has been added. “We knew from the results of previous research that this fragrance is one of the first to go unnoticed when the sense of smell is lost due to covid”, says Eduard Llobet, researcher at the Department of Electronic, Electrical and Automatic Engineering at URV who participated in the study. “We did tests with different concentrations until we defined which one we were looking for”, he adds.

From left to right, URV researcher Eduard Llobet i Youcef Aceli, from IISPV.

This test was carried out on about half a thousand patients who, during the second wave of the pandemic, went to the emergency room at the Hospital Universitario Sant Joan de Reus and primary care centers in the same city because they had symptoms. compatible with covid or because they were asymptomatic but in close contact with a positive case. They were asked to rub the gel on their hands and smell them after three seconds. The result was considered negative if the patient recognized a citrus fruit. And positive if you couldn’t smell the gel or couldn’t find a citrus scent.

Once this stage was completed, the people included in the study should fill out a short questionnaire with the result of the olfactory test and other data such as age, sex and presence or absence of various symptoms. “We gave each symptom a diagnostic value based on a calculation, and there were eight that we considered statistically significant in detecting the disease,” explains Youcef Azeli, an IISPV researcher who led the investigation. Once the answers were given, a PCR test was performed to verify the result.

“The system we developed is based on machine learning and based on the results of the questionnaire, it generated a model that allows detecting cases of covid on a massive scale when resources do not allow carrying out diagnostic tests”, says Albert Fernandez, researcher and developer at URV algorithm based on artificial intelligence. The data guarantee practically total sensitivity (97%), which makes it useful as a population screening method. “Commercialized antigen tests have an average of 80% sensitivity, which means that there are 20% false negatives. What we have developed is not a diagnostic test, but a screening system whose objective is to detect the greatest possible number of positives and avoid false negatives”, explains the team of researchers.

With this device, it is intended that, once the test has been carried out, the person can be clear about their level of risk of suffering from covid, which will help to cut the chains of transmission. “The objective is to protect the most vulnerable people who may still suffer the serious effects of covid, such as the elderly, pregnant women or immunosuppressed patients, and remind people with symptoms compatible with covid to take extreme precautionary measures”, defends the researcher . Youcef Azeli.

IISPV and URV developed a European patent for this prototype, which is part of the technological portfolio of both institutions. The aim is that devices with this gel can be installed in hospitals, homes, schools or public transport, so that people who use them are aware of the risk of spreading the virus they have. The URV Valorization Unit and the IISPV Innovation and Transfer Unit supported the research team in securing the mass screening system.

Bibliographic reference: Youcef Azeli, Alberto Fernández, Federico Capriles, Wojciech Rojewski, Vanesa Lopez-Madrid, David Sabaté-Lissner, Rosa Maria Serrano, Cristina Rey-Reñones, Marta Civit, Josefina Casellas, Abdelghani El Ouahabi-El Ouahabi, Maria Foglia-Fernández, Salvador Sarra & Eduard Llobet. A machine learning mass screening for COVID-19 based on symptoms and a simple olfactory test. Scientific Reports. DOI: https://www.nature.com/articles/s41598-022-19817-x.

Print, PDF and email version