ABSTRACT
INTRODUCTION: Symptomatic Coronavirus Disease 2019 (COVID-19) screening has been a cornerstone of case identification during the pandemic. Despite the myriad of COVID-19 symptoms, symptom screens have primarily focused on symptoms of influenza-like illnesses such as fever, cough, and dyspnea. It is unknown how well these symptoms identify cases in a young, healthy military population. This study aims to evaluate the utility of symptom-based screening in identifying COVID-19 through three different COVID-19 waves. MATERIALS AND METHODS: A convenience sample of 600 military trainees who arrived at Joint Base San Antonio-Lackland in 2021 and 2022 were included. Two hundred trainees with symptomatic COVID-19 before the emergence of the Delta variant (February-April 2021), when Delta variant was predominant (June-August 2021), and when Omicron was the predominant variant (January 2022) had their presenting symptoms compared. At each time point, the sensitivity of a screen for influenza-like illness symptoms was calculated. RESULTS: Of the 600 symptomatic active duty service members who tested positive for COVID-19, the most common symptoms were sore throat (n = 385, 64%), headache (n = 334, 56%), and cough (n = 314, 52%). Although sore throat was the most prominent symptom during Delta (n = 140, 70%) and Omicron (n = 153, 77%), headache was the most common before Delta (n = 93, 47%). There were significant differences in symptoms by vaccination status; for example, ageusia was more common in patients who were not completely vaccinated (3% vs. 0%, P = .01). Overall, screening for fever, cough, or dyspnea had a 65% sensitivity with its lowest sensitivity in the pre-Delta cases (54%) and highest sensitivity in Omicron cases (78%). CONCLUSIONS: In this descriptive cross-sectional study evaluating symptomatic military members with COVID-19, symptom prevalence varied based on predominant circulating COVID-19 variant as well as patients' vaccination status. As screening strategies evolve with the pandemic, changing symptom prevalence should be considered.