Reliance on Assumptions and Estimates with Uncertainty
The study relies on multiple assumptions and estimates for key parameters like infection fatality rate (IFR), vaccine effectiveness (VE), and the proportion of the population infected at different stages. These parameters have considerable uncertainty and vary widely across studies and regions, especially for countries with limited data.
Retrospective Observational Design and Potential Biases
The study's retrospective nature and the lack of a randomized controlled trial introduce inherent biases that are difficult to fully adjust for. Factors like vaccine hesitancy and healthcare access may confound the relationship between vaccination and mortality.
Aggregation of Heterogeneous Data Across Countries
The study aggregates data from countries with vastly different pandemic experiences, vaccination strategies, and healthcare systems. This variation could make global estimates less representative and limit the generalizability of the findings.
Counterfactual Assumptions About Omicron Infections and IFR Estimates
The study assumes that unvaccinated individuals would have been infected during the Omicron period, and their IFR estimates are derived mainly from pre-vaccination data. This is likely a counterfactual and could be an oversimplification of the reality during the Omicron period.
Complexity and Uncertainty of Life-Year Calculation
The study's calculations for life-years saved rely on assumptions about life expectancy and the standardized mortality ratio, which can be complex and subject to interpretation, especially in older age groups.
Lack of Separation Between Deaths Averted and Deaths Due to Vaccine Harm
The study calculates "deaths averted" due to vaccine efficacy, without separating deaths from vaccination-related harms. While acknowledging they are likely small compared to the overall benefit, this distinction would provide a more complete picture of the vaccine impact.
Limited Consideration of Non-Fatal Outcomes and Adverse Events
The focus is mainly on mortality benefits, and non-fatal outcomes and adverse effects are not quantitatively assessed.