Researchers have shown that the prevalence of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in Brazil varies significantly by city and region, with the highest prevalence observed in the North of the country and the lowest seen in South and Center-West regions.
Cesar Victora from the Universidade Federal de Pelotas and colleagues report that some of the highest prevalence was observed in cities located along a stretch of the Amazon river and among indigenous populations. Prevalence did not significantly vary by age group up to 79 years but did fall by about two-thirds among those aged 80 years or older.
The findings come from the first of a series of surveys planned for 133 large cities in Brazil.
A pre-print version of the paper is available in medRxiv*, while the article undergoes peer review.
Rio de Janeiro, Brazil April 29, 2020. An overview of an emergency field hospital is seen at Riocentro to receive the coronavirus covid-19 infections. Image Credit: Antonio Scorza / Shutterstock
Population-based data is needed
Despite the urgent need for population-based data on coronavirus disease 2019 (COVID-19), few studies based on national surveys are available.
Recently, population-based studies have been attempted in some countries, and a national serological survey carried out in Spain identified a range in prevalence, of less than 2% in some regions, and up to 11% in the city of Madrid.
In Brazil, the first case of COVID-19 was reported in São Paulo on the 27th of February, and as of 29th May, more than 440,000 cases and about 26,000 deaths had been reported.
Controversy has surrounded the government’s response, with Brazil’s president having opposed the introduction of social distancing measures and many hospitals being unable to cope with the number of patients affected, particularly in critical care.
Despite the extremely high number of official reports, many cases are going unreported due to testing being limited to severe illness only. Evidence also suggests that deaths due to COVID-19 are being undercounted.
“In light of the present crisis, there is an urgent need for population-based data on the pandemic,” writes Victora and colleagues.
What did the researchers do?
Victora and team conducted a nationwide seroprevalence survey between the 14th and 21st May among households in 133 cities across the 26 Brazilian states and the Federal District.
In each household, one person was randomly selected to undergo rapid serological testing for SARS-CoV-2 antibodies, before completing a short questionnaire to provide sociodemographic information.
The team had planned to sample 250 people in each of the cities, but in one city it was not possible to conduct any interviews; in 28 cities, it was possible to test between 1 and 99 people, and in 14 cities, it was possible to test between 100 and 199 were tested. In 44 cities, 200 to 249 people were tested and in 46 cities, 250 were tested. The overall number of people tested was 25,025.
“To our knowledge, this is the largest population-based study of the prevalence of antibodies to SARS-CoV-2 in geographical scope, and the second largest – after the national survey in Spain – in terms of sample size,” writes the team.
Seroprevalence varied significantly across the country
Seroprevalence of SARS-CoV-2 antibodies varied significantly across the country, from less than 1% in many cities in the South and Center-west regions to as high as 25% in a city called Breves located in the Amazon.
Of fifteen cities with the highest prevalence, eleven were in North Brazil, including six located along a 2,000 km portion of the Amazon river.
Overall, SARS-CoV-2 antibody prevalence across the 90 cities where at last 200 people were tested was 1.4%. The authors say that extrapolation of this 1.4% to the cities’ overall populations, resulted in an estimated 760,000 cases of infection, as opposed to the officially reported 104,782.
Prevalence was similar among men and women and did not significantly vary by age group up to the age of 79, although it did fall by around two thirds at age 80 or older.
The highest prevalence, of 3.7%, was observed among indigenous people, and the lowest prevalence, of 0.6%, was observed among white people, an effect that remained once analysis was confined to the North region, where indigenous people mostly reside.
“Most remarkable finding”
The team says that perhaps their most remarkable finding was the high prevalence observed in the six cities located along the Amazon River.
“In the city of Breves, the prevalence of 25% appears to be the highest ever reported anywhere so far,” write the researchers. “This finding of high prevalence in a tropical region contradicts common wisdom that continents such as Africa may be protected against COVID-19 due to high ambient temperature,” they add.
The team says one possible explanation is that the long river trips that take place from Belém to Breves and from Manaus to Tefé, for example, may represent periods of intense transmission due to overcrowded boats and many people sleeping or resting in close proximity.
“Except for road travel between Belém and Castanhal (67 km), all other transportation among the six cities is by riverboat or – for a minority who can afford it – by plane,” write Victora and colleagues.
The surveys will continue
The team says their study has shown how differently the COVID-19 pandemic has affected different regions of Brazil, with “rapid escalation in Brazil’s North and Northeast, and slow progression in the South and Center-West regions.”
The researchers say this is just the first in a series of national surveys that will now be repeated every three weeks to help track how the pandemic is progressing in the country.
medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.
- Victora C, et al. Remarkable variability in SARS-CoV-2 antibodies across Brazilian regions: nationwide serological household survey in 27 states. medRxiv 2020. Available at: doi: https://doi.org/10.1101/2020.05.30.20117531