Skip to main content

Teach, and teach and teach: does the average citizen use masks correctly during daily activities? Results from an observational study with more than 12,000 participants

Our take —

This study, available as a preprint and thus not yet peer reviewed, conducted direct observations of 12,588 individuals in Baixada Santista, Brazil. The study found suboptimal (<50%) and inappropriate (42.7%) face mask use in public settings.

Study design


Study population and setting

Between June 17 and 19, 2020, investigators conducted direct observations in commercial areas of five cities in Baixada Santista, Brazil. While making daily hour-long observations, investigators counted the number of individuals they observed and the proportion that were wearing face masks. Observations of the following criteria were noted for each recorded individual: 1) mask covering nose and mouth; 2) masks with exposed nose and/or mouth; 3) no mask use; 4) poorly fitted masks; and 5) touching masks/face while in use.

Summary of Main Findings

Across 12,588 observations, 41.5% wore masks covering nose and mouth, with some variability across cities (range: 39.1 – 63.5%). Nearly one-fifth (15.5%) of observants did not wear any face covering. Among those observed wearing masks, 24.9% had visibly exposed noses and/or mouths, and 17.8% touched their mask while in use.

Study Strengths

The authors collated a large sample of observation data, rather than self-reported survey data, to ascertain prevalence and appropriate use of face masks in public settings.


Facemask use was assessed at a single time point raising concerns about the quality of data captured and inferences that can be drawn about an individual’s routine mask use from a brief behavioral observation. Additionally, the authors collected data only on observed mask use, limiting investigation of patterns and associations between mask use and observant characteristics (i.e., age, race, type of mask). Recorded mask use behaviors were assigned to discrete categories when, in reality, individuals could exhibit behaviors falling into multiple categories, which is a source of potential measurement error.Furthermore, some characteristics like, whether a mask is firmly fitted, would be difficult to ascertain from afar. Observations were limited to large commercial streets so behavior may differ in different settings.

Value added

The study is among the first to report on prevalence and appropriate use of face coverings in public using direct observation in a setting with high COVID-19 transmission.

This review was posted on: 18 July 2020