COVID-19 Reinfections in Mexico City: Implications for public health response

Abstract

Background: SARS-CoV-2 pandemia continues to be important even when more than 60% of the global population has been vaccinated. As the pandemia evolves the number of reinfection cases will continue to increase as new variants are generated that evade the immune response. Understanding reinfections is important to guide the public health system and to inform decision-makers. Methods: We downloaded clinical outcome and severity of infection data from the SISVER (respiratory disease epidemiological surveillance system) database. We sequenced SARS-CoV-2 samples, identified SARS-CoV-2 lineage and upload this genomic data to GISAID. We analyzed time and lineage between index infection and reinfection. We also analyzed the clinical outcome, severity of infection and vaccination status during reinfections. Findings: In this study we confirmed that each wave of SARS-CoV-2 infections was characterized by a different viral variant showing a prevalence higher that 95%. We found that the fraction of reinfection is not linearly related to the average time of separation between waves with 40% of all the reinfections occurring at wave 5, the only wave with more than one SARS-CoV-2 variant with a prevalence higher than 80%. Regarding type of care 2.63% patients were considered ambulatory during the reinfection even when they were hospitalized during the index infection and only 0.78% presented the opposite behavior. Moreover, 6.74% reinfections transitioned from asymptomatic to mild or severe or from mild to severe; and 8.95% transitioned from severe to mild or asymptomatic or from mild to asymptomatic. The highest number of reinfections have occurred in unvaccinated patients (41.6%), followed shortly by vaccinated patients (31.9%). However, most reinfections occurred after wave 4 when the national vaccination efforts have reached 65% of the general population. Interpretation: The analyzed data suggests a diminished severity of infection during reinfection either if transitions in disease severity or transitions in type of patient care are considered. Finally, we also observed an overrepresentation of unvaccinated patients in reinfections.