Abstract

Introduction:  The COVID-19 epidemic has manifested geographically as clusters of high morbidity (hot zones) and as cold spots of low incidence, which have been explained based on social variables.

Objective:  To characterize morbidity patterns due to COVID-19 in Costa Rica from March 2020 to May 2022 and to explain them through social determinants of health in the geographical context.

Methods:  An ecological study at the district level was designed with data on vaccination against COVID-19, weekly reports on speed of advance of the epidemic, development level, and other demographic data. Thematic maps were constructed, and spatial morbidity patterns were identified and characterized, which were explained using linear and geographically weighted regression models.

Results:  In the Greater Metropolitan Area and surrounding area, clusters of hot spots were identified, and cold spots flanked these high-incidence areas. The linear regression model, built from the variables: average number of vaccines per person, speed in weekly case reporting, social development in its economic, educational, and health dimensions, as well as the proportion of overcrowded homes and people born in the outside, explained more than 70 % of the spatial variations of the incidence of cases (crude and standardized by age and sex). The geographically weighted model corrected autocorrelation problems, improving the explanatory capacity to 82 %.

Conclusions:  morbidity during the COVID-19 epidemic until May 2020 was configured spatially through well-established clusters of hot and cold spots. This structure could be explained from the social determinants of health, proving that effects on morbidity are generated, differentiated territorially.

Key words: epidemic; risk; territory; clusters; geographic regression.