Objective: To evaluate the efficacy of alcohol sniff test as a predictor of SARSCoV2 infection in patients with flu syndrome.

Methods: A cross-sectional observational study was conducted between September and December 2020 in the employees of a tertiary hospital who presented mild influenza syndrome. A total of 103 individuals participated in the study, divided into three groups: flu-like syndrome and RT-PCR test positive for COVID-19; flu syndrome and negative RT-PCR test for COVID-19 and an asymptomatic control group. All patients were submitted to olfactory evaluation through the alcohol sniff test.

Results: Of the 103 individuals studied, 35 (33.98%) had flu-like symptoms and positive RT-PCR, 38 (36.89%) had flu-like symptoms and negative RT-PCR and 30 (29.12%) were asymptomatic. The overall mean distance of the AST test was 10±8.2cm. There was a statistically significant difference between the mean distance of the COVID+ groups (4.35±4.1cm) and the control group (20±4.3cm) (p<0.05). This relationship was also maintained between the groups COVID+ (4.35cm) and COVID− (9±7.5cm) (p<0.05). For a cut-off of 10cm, the AST presented sensitivity of 88% and specificity of 41%, leading to an odds-ratio of 9.7 (95% CI 3.3–28.1) (p<0.001).

Conclusion: The alcohol sniff test presented high sensitivity and odds ratio for COVID-19 screening in patients with mild influenza syndrome in the context of pandemic.

Keywords: Olfaction; COVID; Otorhinolaryngology.