To investigate the risk of ocular adverse events after COVID-19 mRNA vaccination.
Matched cohort and self-controlled case series (SCCS) studies.
We used a population-based database of medical claims and vaccination records in a large city of Japan. In the matched cohort study, we identified individuals who received COVID-19 vaccination (BNT162b2) from February 2021 to September 2021. One control was selected from non-vaccinated individuals by matching time, date of birth, sex, Charlson comorbidity index, and the enrollment period for health insurance. In the SCCS study, we analyzed individuals who developed ocular adverse events from February 2021 to the end of follow-up.
In the matched cohort study, we applied the Kaplan-Meier estimator to estimate the cumulative incidence of ocular adverse events over 21 days after the first dose and 84 days after the second dose. In the SCCS method, we used conditional Poisson regression to estimate the incidence rate ratio of ocular adverse events during the risk periods (0-21 days after the first dose and 0-84 days after the second dose) compared to the remaining periods.
MAIN OUTCOME MEASURES
Composite outcome of uveitis, scleritis, retinal vein occlusion, and optic neuritis.
There were 99,718 pairs eligible for the matched cohort study after the first dose (mean age, 69.3 years; male, 44%). The vaccinated and control groups developed 29 and 21 events, respectively, over 21 days after the first dose, and 79 and 28 events, respectively, over 84 days after the second dose. The differences in cumulative incidence (reference, the control group) were 2.9 (95% confidence interval, -14.5 to 19.1) events/100,000 persons and 51.3 (16.2 to 84.3) events/100,000 persons, respectively, for the first and second dose. The SCCS study showed the incidence rate ratios of 0.89 (0.62 to 1.28) and 0.89 (0.71 to 1.11) for the first and second doses, respectively.
The matched cohort analysis found an increased risk for the composite outcome after the second dose; however, the SCCS analysis showed no increased risk. Considering that the SCCS can cancel out time-invariant confounders, the current results suggest that COVID-19 vaccination is unlikely to causally increase the risk of ocular adverse events.