New study from Brazil doesn’t show that COVID-19 vaccines increase risk of death in the long term

Since their rollout in December 2020, the safety of COVID-19 vaccines has been a frequent target of misinformation. Science Feedback repeatedly addressed claims about the alleged dangers of these vaccines, showing that many such assertions were misleading, unsupported by evidence, or outright false (see examples here, here, and here).

A new claim from December 2024 continued this trend, alleging that data from Brazil demonstrated that COVID-19 vaccines are “catastrophic” and increase long-term mortality risk among those who received them.

Chiropractor Patrick Flynn posted on Facebook that “for a negligible benefit against COVID, after one year the benefit was reversed, and the vaccinated showed a substantially greater risk of death.” Similarly, epidemiologist Nicolas Hulscher referred to the vaccines as “life-reducing injections.” Both Flynn and Hulscher have a history of spreading inaccurate claims about vaccines.

This claim cites a December 2024 study published in Frontiers in Medicine by Brazilian researchers Nádia Cristina Pinheiro Rodrigues and Mônica Kramer de Noronha Andrade as supporting evidence^[1]. The study examined all-cause mortality in individuals who had severe COVID-19, assessing their risk in the medium term (which they defined as three months to one year after symptoms of severe disease) and the longer term (defined as beyond one year).

However, the claim incorrectly represented the studies’ findings, generalizing observations from a highly specific population to the general population. Furthermore, the study itself isn’t equipped to draw a causal relationship between vaccination and an increased mortality risk due to potential bias. We’ll explain why below.

What the study did and what it found

The study by Pinheiro Rodrigues and Kramer de Noronha Andrade investigated the survival of patients who experienced severe COVID-19 between 2020 and 2023. Specifically, the authors analyzed data from a Brazilian nationwide database of individuals diagnosed with COVID-associated severe acute respiratory syndrome (SARS). They defined a case of SARS as an “Individual with Flu-like Syndrome who presents: dyspnea/respiratory distress or persistent pressure or pain in the chest or oxygen saturation less than 95% in room air [i.e. without ventilation] or bluish color (cyanosis) of the lips or face”.

Thus, the study’s participants are COVID-19 patients who developed a severe form of the disease. The authors described them as “very severe patients with SARS.”

To focus on medium- and long-term survival, the study restricted its analysis to patients who were still alive three months after their SARS episode. As a result, patients who died during or shortly after their severe COVID-19 illness were excluded from the study.

The authors then conducted a statistical analysis comparing the all-cause mortality risk of vaccinated and unvaccinated patients, adjusting for potential confounding factors such as age, comorbidities, and socioeconomic status.

Their findings revealed that in the medium term, which they defined as three months to one year after SARS, vaccinated individuals had a 4 to 8% lower mortality risk compared to unvaccinated individuals.

However, in the long term (defined as one year or more after SARS), the mortality risk was significantly higher among vaccinated individuals, the risk being 69 to 94% higher compared to unvaccinated individuals.

The study didn’t assess the risks and benefits of COVID-19 vaccines for the general population

The claim that COVID-19 vaccines brought negligible benefits to the public while increasing their risk of death relied solely on the study by Pinheiro Rodrigues and Kramer de Noronha Andrade. However, this study didn’t support such a sweeping conclusion.

First, as previously explained, the study focused exclusively on individuals who experienced severe COVID-19, specifically those with severe acute respiratory syndrome (SARS). This was a small and highly specific subset of the population, and findings from this group can’t be generalized to the broader population.

Second, the study didn’t compare the overall risks and benefits of COVID-19 vaccines. A proper risk-benefit analysis would have required quantifying both the number of deaths potentially averted by vaccination and any deaths that vaccines might have caused. The study didn’t include such an analysis and, therefore, couldn’t inform broader conclusions about the vaccines’ overall impact on public health.

In contrast, other studies and analyses have directly assessed the risks and benefits of COVID-19 vaccination^[2,3]. For instance, analyses in the U.S. estimated that hundreds of thousands of deaths among unvaccinated individuals could have been prevented if those individuals had been vaccinated.

The study didn’t include all-cause deaths that occurred in the first three months after SARS

The authors only included individuals who were still alive three months after their SARS episodes. As stated in the study’s Methods section, this was done on purpose to “capture only deaths that are not directly attributable to COVID-19 from post-immediate hospitalization and discharge.”

The authors also noted that they had performed a separate analysis of deaths occurring within the first three months after SARS. In that earlier analysis, they found that vaccination reduced the risk of death^[4].

In fact, research shows that COVID-19 vaccines reduce the risk of death^[1,4,5,6]. It’s therefore highly plausible that proportionally more unvaccinated people died in the first three months after severe COVID-19 compared to vaccinated individuals. In other words, the protective benefit of vaccination is likely strongest in the early phase following severe disease^[7-9].

Because these early deaths were excluded from the analysis, it’s not possible to conclude whether there were more deaths among people who received a COVID-19 vaccine than among unvaccinated people, as the claim asserted.

Moreover, excluding the deaths occurring during the first three months may introduce a survivorship bias, potentially influencing the study’s long-term findings. We reached out to the study’s authors for comment but didn’t have a reply to our questions at the time of publication.

If vaccinated and unvaccinated individuals died at different rates during this early phase, the groups that survived to be included in the analysis might differ significantly in terms of baseline health and resilience. As a result, the long-term mortality comparison may have been conducted on two distinct populations, which could skew the observed difference in long-term mortality risk.

Restricting the study to SARS survivors biases the relationship between vaccination and mortality risk

Another important consideration is that the inclusion criterion used in the study—experiencing COVID-related SARS—may introduce a bias that influences the observed association between vaccination status and mortality risk.

The likelihood of developing severe COVID-19 (SARS) depends on an individual’s baseline health and resilience. Vulnerability to disease varies between people. Frail individuals or those with specific genetic predispositions are more vulnerable to disease and more likely to develop severe COVID-19 when infected, compared to healthier individuals.

At the same time, extensive research has shown that vaccination reduces the risk of experiencing severe COVID-19. This means that both vaccination status and baseline health influence the likelihood of developing SARS.

When the inclusion in a study is conditioned on a criterion—developing SARS—that is influenced by two factors—vaccination status and baseline health—it can introduce what is known as index event bias or collider bias^[10,11]. Collider bias can create a spurious relationship between vaccination status and the outcome (death more than one year after SARS).

One consequence of this bias is that the vaccinated and unvaccinated groups in the study differ by more than just their vaccination status. Because vaccination reduces the risk of developing SARS, some vaccinated people who contracted COVID-19 may have avoided SARS and were excluded from the study. Those same people, had they been unvaccinated, might have developed SARS and been included. This selection process distorts the comparison between vaccinated and unvaccinated groups, potentially leading to misleading conclusions about long-term mortality risk.

In other words, the observed increase in mortality risk among vaccinated people one year after SARS may be influenced by collider bias due to the study’s inclusion criterion rather than a causal relationship.

Conclusion

In summary, the claim that COVID-19 vaccines are “catastrophic” and put people at higher risk of dying misrepresents the findings of the cited study. While the study observed a higher long-term mortality risk among vaccinated individuals who had survived severe COVID-19, this observation is limited to a specific population of fragile patients and cannot be generalized to the general population.

Furthermore, while the study aimed to explore long-term mortality risk, inherent complexities in studying such populations, including the potential for survivorship and collider bias, may influence the observed associations. Importantly, a wealth of evidence from other studies has demonstrated the significant role of COVID-19 vaccination in preventing severe illness and death.

REFERENCES

• 1 – Pinheiro Rodrigues & Kramer de Noronha Andrade (2024) Evaluation of post-COVID mortality risk in cases classified as severe acute respiratory syndrome in Brazil: a longitudinal study for medium and long term. Frontiers in Medicine.
• 2 – Jia et al. (2023) Estimated preventable COVID-19-associated deaths due to non-vaccination in the United States. European Journal of Epidemiologia.
• 3 – Zhong et al. (2022) Estimating Vaccine-Preventable COVID-19 Deaths Under Counterfactual Vaccination Scenarios in the United States. medRXiv (preprint).
• 4 – Pinheiro Rodrigues et al. (2024) Mortality risk of Severe Acute Respiratory Syndrome cases classified as COVID-19: A longitudinal study. PLoS One.
• 5 – Xu et al. (2024) Vaccine Mortality risk after COVID-19 vaccination: A self-controlled case series study. Vaccine.
• 6 – Xu et al. (2021) COVID-19 Vaccination and Non–COVID-19 Mortality Risk — Seven Integrated Health Care Organizations, United States, December 14, 2020–July 31, 2021. Morbidity and Mortality Weekly Report.
• 7 – Neto et al. (2023) Clinical-epidemiological characteristics and survival of cases of severe acute respiratory syndrome (SARS) due to COVID-19, according to the COVID-19 vaccination schedule in Brazil, 2021-2022: a prospective study. Epidemiologia e Serviços de Saúde.
• 8 – Thygesen et al. (2022) COVID-19 trajectories among 57 million adults in England: a cohort study using electronic health records. The Lancet Digital Health.
• 9 – Seo et al. (2022) Impact of prior vaccination on clinical outcomes of patients with COVID-19. Emerging Microbes & Infections.
• 10 – Griffith et al. (2020) Collider bias undermines our understanding of COVID-19 disease risk and severity. Nature communications.
• 11 – Dahabreh & Kent (2011) Index event bias: an explanation for the paradoxes of recurrence risk research. JAMA.