Estimating the Causal Effect of Two-Dose COVID-19 Vaccination on Hospitalization Rates

Background and aims: COVID-19 remains a global health challenge, with vaccination crucial for reducing severe cases. This study evaluated a two-dose COVID-19 vaccine’s effectiveness in lowering hospitalization rates using advanced statistical techniques. This study evaluated the efficacy of a two-dose COVID-19 vaccination regimen in reducing hospitalization rates by employing advanced statistical techniques to control confounding variables in the observational data. Methods: A retrospective cohort study was conducted among individuals tested for COVID-19 at Mashhad University of Medical Sciences from March 21, 2021, to March 20, 2022. The study population comprised all individuals who underwent polymerase chain reaction testing for COVID-19 during this period. A census sampling method was employed, resulting in a final sample size of 306630 individuals. The participants were classified as "vaccinated" if they received both doses and "unvaccinated" if they received none. Hospitalization was defined as COVID-19-related admissions occurring at least two weeks post-vaccination. The required data were collected from three databases, including the Sina Health Information System, the Healthcare Services Monitoring System, and the Hospital Information System. To create comparable groups, propensity score (PS) matching and weighting were utilized, and a logistic regression model was utilized to estimate the average treatment effect (ATE) of vaccination on hospitalization outcomes. Results: Among the 306630 patients included in the study, 104115 (33.95%) were unvaccinated, while 202515 (66.05%) were vaccinated. Overall, 29458 patients (9.61%) were hospitalized, comprising 28,244 unvaccinated and 1214 vaccinated individuals. Vaccinated individuals exhibited significantly lower odds of hospitalization. The adjusted odds ratio for hospitalization was 0.72 (95% confidence interval [CI]: 0.68–0.76) when using PS weighting, 0.32 (95% CI: 0.30–0.34) with matching, and 0.34 (95% CI: 0.33–0.35) after adjusting for extreme weights. Conclusion: The findings underscore the protective effects associated with a two-dose COVID-19 vaccination regimen and emphasize the significance of employing robust statistical methods in evaluating real-world data.