Excess mortality from COVID 19 in Costa Rica: a registry based study using Poisson regression

Abstract

Excess mortality from COVID 19 in Costa Rica: a registry based study using Poisson regression Romain Fantin,a ,b ,c ,∗ Cristina Barboza-Solís,c Allan Hildesheim,b and Rolando Herrerob a Centro Centroamericano de Población, Universidad de Costa Rica, San Pedro, Costa Rica b Agencia Costarricense de Investigaciones Biomédicas – Fundación Inciensa, San José, Costa Rica c Facultad de Odontología, Universidad de Costa Rica, San Pedro, Costa Rica Summary Background Official death toll related to COVID-19 has been considerably underestimated in reports from some Latin American countries. This study aimed to analyze the mortality associated with the COVID-19 pandemic in Costa Rica between March 2020 and December 2021. Methods A registry based study based on 2017–2021 data from the National Institute of Statistics and Census was designed (N = 128,106). Excess deaths were defined by the WHO as “the difference in the total number of deaths in a crisis compared to those expected under normal conditions”; and were estimated using a Poisson regression, and mortality and years of potential life lost (YPLL) rates were calculated. Findings The COVID-19 pandemic represented 15% of the deaths in Costa Rica between March 2020 and December 2021. The mortality rate related to COVID-19 was 83 per 100,000 person-years. Between March and July 2020 (low- incidence period), observed number of deaths was 9%-lower than expected, whereas it was 15% and 24% higher than expected between July 2020 and March 2021 (high incidence period - no vaccination), and between March 2021 and December 2021 (high incidence period – progressive vaccination) respectively. Between July 2020 and December 2021, excess deaths observed and COVID-19 deaths reported were comparable (7461 and 7620 respectively). Nevertheless, there were more deaths than expected for conditions that predispose to COVID-19 deaths. YPLL and mortality rates increased with age, but significant excess deaths were observed in all age-groups older than 30–39 years. No large differences were noted by districts’ socioeconomic characteristics although excess death rate was lower in rural compared to urban areas.