**4. Discussion**

Our data showed an impressive reduction in COVID-19-related deaths in older age groups in Ceará State, which is the population strata at highest risk for severe disease and death. Previous studies have shown that, by May 2021, more than 40,000 deaths had been prevented due to vaccination of the elderly population in Brazil with the Oxford-AstraZeneca and CoronaVac-Sinovac/Butantan vaccines [7]. Similar findings were found in the US after use of the first dose of Pfizer-BioNTech, particularly in older adults [9]. A study in Tennessee/USA showed a reduction of more than 95% in mortality in the vaccinated elderly population between December 2020 and March 2021 [10].

Considering the difficulties in the vaccine supply chain and their availability, it is important that the vaccines from both major producers showed a high effectiveness in reducing COVID-19-related deaths, even after a single dose. Furthermore, as predicted by Bolcato et al. in 2020, there may be problems that occur, such as insufficient production of vaccine doses for the entire population, or with different vaccination strategies and different times between doses, generating the need for difficult prioritization decisions [11]. In this context, the ability of a vaccine to protect against serious illness and death should be considered the most important outcome, since hospital admissions, especially in intensive care units, represent the greatest burden on health systems and has led several countries to face a collapse in their health systems. The global crisis generated by the coronavirus pandemic highlighted, once again, the importance of vaccination programs as effective public health measures, and brought about new mechanisms that may become models for future responses to regional epidemics and pandemics, with a greater variety of platforms and joint work to overcome challenges and accelerate vaccine development, manufacturing and delivery [12]. It is worth noting that the duration of protection after recovery from COVID-19 corresponds somewhat to the duration of protection provided by the vaccine [13].

For Hodgson et al. (2021), the beneficial effects of a vaccine can be assessed if the vaccine is effective in older adults and if there is a wide distribution of the vaccine [1]. The evaluation of asymptomatic SARS-CoV-2 infection is an important clinical outcome in the evaluation of vaccines, but is certainly of less public health importance than its effectiveness against death. In Italy, for example, the number of infections in nursing homes was particularly high, with a high mortality rate. Yet it must be recognized that the current situation of social disparity does not facilitate equal opportunities for all. As a result, the elderly will continue to experience moments of loneliness, despite efforts to reduce them [14].

Equal access to COVID-19 vaccines in all countries will continue to be a goal to be pursued. But the experience of previous pandemics suggests that access will be limited in low and middle income countries, despite the rapid development of some new candidate vaccines. Thus, the WHO proposal, with the COVAX Facility program, represents an attempt to facilitate multilateral cooperation to procure and distribute two billion doses of COVID-19 vaccines equitably in all countries of the world by the end of 2021 [15].

Our study is subject to some limitations, such as the use of secondary mortality data that may be subject to some errors. The smaller number of second doses by AstraZeneca in our study is basically due to the longer period between the two doses and, therefore, the population had not ye<sup>t</sup> received the second dose during the study period. We also observed that the population of people vaccinated in the age group over 90 years was higher than the estimated population for this age group, this fact is due to the last census being conducted in 2010. We used the population projection for the year 2021, but there was still a difference of 1900 more people vaccinated in the population over 90 years of age. The estimated population was adjusted to the vaccinated population and, thus, data should be interpreted with care. Data on the antibody response of vaccinated individuals were not available, which may limit interpretation of results. However, we obtained population-based data from a population with a high vaccination coverage, and the study results can, thus, be considered as robust and valid.
