COVID-19 Diagnosis based on Taste Disorders: A Case-Control Study

Author(s): Kamil Adamczyk, Michal Herman, Janusz Fraczek, Robert Piec, Barbara Szykula-Piec, Artur Zaczynski, Rafal Wójtowicz, Krzysztof Bojanowski, Ewa Rusyan, Zbigniew Kró, Waldemar Wierzba, E

Background: We aimed to assess taste disturbances in COVID-19 patients. Additionally, we hypothesized that it is possible to establish a reliable and inexpensive screening method for SARS-COV-2 infection using a flavor test.

Methods: The ability to taste the sweet, salty, sour and bitter flavors was assessed in 52 COVID-19 patients and 36 controls using flavor tablets (sucrose, NaCl, ascorbic acid and grapefruit extract - 99% naringin) with flavor concentrations established previously in healthy subjects.

Results: Subjective smell or taste disturbances were reported by 65% of subjects in the test group compared with 8% in the control group. A numerical difference was demonstrated for all flavor tests recognitions between COVID-19 patients and the control group; however, the difference was only significant with the sweet and salty flavor tablets. Complete ageusia was very uncommon in the study group and occurred in only 1 out of 52 subjects. Only one predictive model – the medical questionnaire, consisting of questions regarding self-reported loss of taste, smell, or fever - displayed good performance in COVID-19 diagnosis (ROC AUC = 0.82). The performance of other predictive models, including the taste testers, was lower.

Conclusion: Among all tested flavors, the sweet and salty flavors perception was impaired in COVID-19 patients. Despite the objective taste tests being carried out, the medical questionnaire was the predictive model with the best performance, higher than any single taste test or even multiple tastes test.