Evaluating Real-World COVID-19 Vaccine Effectiveness Using a Test-Negative Case-Control Design (preprint)

ABSTRACT

It is important to assess the extent to which the real-world effectiveness of marketed vaccines is consistent with that observed in the clinical trials, and to characterize how well vaccines prevent COVID-19 symptoms. We conducted a modified test-negative design (TND) to evaluate the RW effectiveness of three COVID-19 vaccines by leveraging data from an on-going, US community-based registry. Vaccine effectiveness was examined in two ways considering cases who (1) tested positive for COVID-19 (695 cases, 1,786 controls) and who (2) tested positive with at least one moderate/severe COVID-19 symptom (165 cases, 2,316 controls). Any vaccination (full or partial) was associated with a 95% reduction in the odds of having a positive COVID-19 test [adjusted odds ratio (aOR) = 0.05 (95% confidence interval (CI) 0.04, 0.06)]. Full vaccination was associated with an aOR of 0.03 (95% CI 0.03, 0.05) while partial vaccination had an aOR of 0.08 (95% CI 0.06, 0.12). Any vaccination was associated with a 71% reduction in the odds of testing positive and having at least one moderate/severe symptom (aOR=0.29 (95% CI 0.20, 0.40)). High effectiveness was observed across all three vaccine manufacturers both for prevention of positive COVID-19 test results and prevention of moderate/severe COVID-19 symptoms.