Self-reported symptoms as predictors of SARS-CoV-2 infection in the general population living in the Amsterdam region, the Netherlands

Introduction Most COVID-19 symptoms are non-specific and also common in other respiratory infections. We aimed to assess which symptoms are most predictive of a positive test for SARS-CoV-2 in symptomatic people of the general population who were tested. Methods We used anonymised data of all SARS-CoV-2 test results from the Public Health Service of Amsterdam from June 1,2020 through August 31, 2021. Symptoms were self-reported at time of requesting a test. Multivariable logistic regression models with generalized estimating equations were used to identify predictors of a positive test. Included symptoms were: cough, fever, loss of smell or taste, muscle ache, runny nose, shortness of breath, and throat ache; adjustments were made for age and gender, and stratification by month. Results Overall, 12.0% of 773,680 tests in 432,213 unique individuals were positive. All symptoms were significantly associated with a positive test result, the strongest positive associations were: cough (aOR = 1.78, 95%CI = 1.75-1.80), fever (aOR = 2.11, 95%CI = 2.07-2.14), loss of smell or taste (aOR = 2.55, 95%CI = 2.50-2.61), and muscle ache (aOR = 2.38, 95%CI = 2.34-2.43). The adjusted odds ratios for loss of smell or taste slightly declined over time, while that for cough increased. Conclusion Cough, fever, loss of smell or taste, and muscle ache appear to be most strongly associated with a positive SARS-CoV-2 test in symptomatic people of the general population who were tested.