Sweat samples

From October 2020 until April 2021, sampling was organized in Belgium in different hospitals. Positive samples came from CHU-Liege, CHU-ND-Bruyeres, CHU St-Pierre Brussels, St-Pierre Ottignies, UZ Gent, AZ Glorieux Ronse, AZ Oudenaarde, OLVZ Aalst, Jan Yperman Ieper, UZA Antwerp, ZNA Stuivenberg, GZA Anvers, AZ Klina Brasschaat, Jan Palfijn Gent, AZ St-Vincentius Deinze, St-Trudo St-Truiden, AZ Jan Portaels Vilvoorde, AZ Alma Eeklo. Negative samples came from the Kiwanis organization, who organized sampling in different cities in Belgium. Additionally, different care centers (hospitals, senior homes) organized sampling: WZC Armonea Wilrijk, WZC Armonea Spanjeberg, Zorg-Saam WZC Oostakker, WZC Curando Ruiselede, CHU-Liege, CHU St-Pierre Brussels. A list of metadata was collected from each patient/participant, including date, age, biological gender, weight, height, Body Mass Index, ethnicity, postal code, deodorant use, deodorant use frequency, hygiene habits, frequency of underarm washing, medication use, hormonal contraception use, antibiotics use, smoking, comorbidities, (hospital) location of sampling, SARS-CoV-2 symptoms, Ct-value of qPCR result. In the essence of time, dogs used in the present study were trained on a large and diverse set of samples including different hospitals/elderly homes, young and old people, male and female persons, smoker and non-smoker; deodorant user and no underarm cosmetic users.

All sweat donors had their SARS-CoV-2 status (negative or positive) confirmed by qPCR. For positive samples, only patients with clear symptoms (hospitalized) and qPCR results of <30 cycles were preferred. Patients were tested multiple times in the hospital. Patients with no PCR-confirmed test and/or vaccinated against SARS-CoV-2 were excluded. Patients in hospitals with clinical signs related to SARS-CoV-2 (respiratory symptoms, fever) but negative (qPCR) to SARS-CoV-2 were also included. With each donor, a complete but anonymized clinical metadata file was completed. The sweat sampling was performed by trained doctors and/or nurses for safety reasons. It consisted of 5 cotton balls or sterile compresses placed under the 2 armpits of the patient/donor during 15–30 min. For a subset of patients, the sampling was repeated on different days, as long as patients were still SARS-CoV-2 positive. The sampler wore nitrile gloves and a coverall (biological hazard) and handled the samples with a clamp, before putting them in a glass jar or in a closed plastic bag (ziplock). Within 1 h, the plastic bag or the glass jar were frozen (−20°C or colder) and stored until the training of the dogs. Temperature inside the freezer was constantly recorded and was found to be stable.

The tested samples during validation and post-validation were obtained from the original SARS-CoV-2 virus (WIV04 / 2019). At a later stage, samples were obtained from vaccinated people at CHU Saint-Pierre Brussels at least 3 weeks after the second dose of their vaccine (Comirnaty, BioNTech-Pfizer). These samples were also presented to the dogs, together with a positive control sample.