A population-based study of the prevalence of COVID-19 infection in Espírito Santo, Brazil: methodology and results of the first stage

BACKGROUND COVID-19 is affecting almost the entire world, causing more than four hundred thousand deaths and undermining the health care systems, as much as the economy, of the afflicted countries. The strategies for prevention depend on largely lacking information, as infection prevalence and virus pathogenicity. This study aimed to determine the prevalence, the pathogenicity, and the speed of infection spreading in a large population in Brazil.

MATERIALS AND METHODS This is a serial cross-sectional study designed on a population basis and structured over houses as the sampling units. The sampling consisted of four visits at 15 days intervals in randomly selected census-designated sectors of the State major municipalities (reference municipalities) and two visits at 30 days intervals in smaller municipalities of the same regions of those of reference. At each visit, the investigators sampled houses and sampled one individual in each house for data collection. After the informed consent, the investigators performed a rapid antibody detection test (Celer Technology, Inc) and applied a questionnaire containing clinical and demographic questions.

RESULTS From May 13th to 15th, the investigators performed 6,393 rapid tests in 4,612 individuals of the reference municipalities, 1,163 individuals of the smaller municipalities, and 166 contacts of the positive individuals. Ninety-seven dwellers were positive in the reference municipalities, giving a prevalence of 2.1% (CI 95%: 1.67-2.52%). In the smaller municipalities, the figure was 0.26% (CI 95%: 0.05%-0.75%) (three positives). There was an association of the positive result with female sex (p = 0.013) and houses with five dwellers or more (p = 0.003). Seventy-eight positive individuals reported symptoms in the previous 15 days (80.4%), being anosmia (45.4%), cough (40.2%), and myalgia (38.1%) the more frequent. About one-third of them reported fever (28.9%).

CONCLUSIONS The results reveal a still small prevalence of infection in the study area, despite the significant number of sick people overloading the health system. The figures indicate an important underreporting in the area and a frequency that still can grow, making necessary public health actions for the containment of the transmission.