To evaluate the incidence of new retinal artery occlusion (RAO) and retinal vein occlusion (RVO) after the diagnosis of coronavirus disease 2019 (COVID-19) or vaccination against COVID-19 and compare the incidences with the population with neither. Nationwide population-based cohort study. From a nationwide population-based cohort, 8 418 590 patients were categorized into control (group 1), COVID-19 infection (group 2), and COVID-19 vaccination (group 3) groups. The cumulative incidence of RAO and RVO was calculated in groups 1, 2, and 3 using the Kaplan-Meier method. We calculated hazard ratios (HRs) and 95% confidence intervals (CIs) based on the Poisson distribution for RAO and RVO according to each group and subgroup using Cox proportional hazards models, with group 1 as the reference. We conducted univariable and multivariable analyses for the risk factors of RAO and RVO according to each subgroup. Cumulative incidence and risks of incidence of RAO and RVO from the index date to day60. In multivariable analysis, no significant increase in RAO and RVO risks after COVID-19 or COVID-19 vaccination were observed in either men or women. These results were observed consistently across various conditions in sensitivity analyses. In subgroup analysis, individuals who were vaccinated before infection showed no significant increase in RAO or RVO risks in both sexes compared with the control group. In the subgroup analysis of vaccinated patients, the HRs of RAO and RVO for different vaccine types did not show an increase compared with the control group; however, an exception was observed in women who received mRNA-1273 vaccines, who showed a higher RAO HR (4.65; 95% CI, 1.27-17.03; P= 0.021). Within 60 days of COVID-19 diagnosis or vaccination, RAO and RVO occurred rarely. We observed no increase in the HR of RVO and RAO relative to COVID-19 or COVID-19 vaccination except for a possible increase in the RAO HR in women who received mRNA-1273, for which the raw incidence was extremely low. Further investigation is required to validate this result. The author(s) have no proprietary or commercial interest in any materials discussed in this article.
Read full abstract