Differential symptoms among COVID-19 outpatients before and during periods of SARS-CoV-2 Omicron variant dominance in Blantyre, Malawi: a prospective observational study

BackgroundIt is widely reported that the SARS-CoV-2 Omicron variant has resulted in high number of cases, but relatively low incidence of severe disease and deaths, compared to the pre-Omicron variants of concern. We aim to assess the differences in symptom prevalence between Omicron and pre-Omicron infections in a sub-Saharan African population. MethodsIn this cross-sectional observational study, we collected data from children and adult outpatients presenting at two primary healthcare facilities in Blantyre, Malawi. Eligible participants were aged >1month old, with signs suggestive of COVID-19, and those not suspected of COVID-19. Nasopharyngeal swabs were collected for SARS-CoV-2 PCR testing and positive samples whole genome sequenced to identify the infecting variant. The primary outcome was the likelihood of presenting with a given symptom in individuals testing positive during the period in which Omicron-dominated (December 2021 to March 2022) with those infected during the pre-Omicron period (August 2021 to November 2021). FindingsAmong 5176 study participants, the median age was 28 years (IQR 21-38), of which 6.4% were under 5, 9.2% were 6 to 17 years, 77% were 18 to 50 years, and 7.1% were above 50 years old. Prevalence of SARS-CoV-2 infection was 23% (1187/5188), varying over time, with peaks in January 2021, July 2021 and December 2021, driven by the Beta (B.1.351), Delta (B.1.617.2) and Omicron (BA.1/2) variants, respectively. Headache (OR 0.47[CI 0.29 - 0.79]), cough (OR 0.37[CI 0.22 - 0.61]), fatigue (OR 0.20[CI 0.08 - 0.48]) and abdominal pain (OR 0.38[CI 0.18 - 0.78]) were less common in participants infected during the Omicron-dominant period than during pre-Omicron period. Fever was more common in participants infected during the Omicron-dominated period than during pre-Omicron period (OR 2.46[CI 1.29 - 4.97]). COVID-19 vaccination, accounting for number of doses and days since last dose, was not associated with a reduced risk of PCR-confirmed SARS-CoV-2 infection (1 dose, OR 1.10[CI 0.39 - 2.66]; 2 doses, OR 1.11[CI 0.40 - 2.57]; all p=0.8). InterpretationIn this Malawian population, the prevalence of clinical symptoms associated with Omicron infection differ from those of pre-Omicron infections and may be harder to identify clinically with current symptom guidelines. To maintain robust surveillance for COVID-19 and emerging variants, case definitions and testing policies will need to be regularly reviewed to ensure case ascertainment.