I diopathic uveitis comprises a heterogeneous group of intraocular inflammatory diseases that are often immune-mediated and can lead to significant visual morbidity. Although the precise triggers remain unclear in many cases, systemic immune activation following viral infections is recognized as a potential contributor to the onset or exacerbation of ocular inflammation through mechanisms such as molecular mimicry, bystander activation, and disruption of ocular immune privilege. ^,, Viral infections can also compromise the blood-retina barrier and alter local immune regulation, reducing tolerance to self-antigens and promoting a proinflammatory milieu.

Vaccination is a cornerstone of infectious disease prevention, primarily designed to elicit protective immunity against specific pathogens. While some studies have shown that coronavirus disease 2019 (COVID-19), ^,,,, human papillomavirus (HPV), ^,, and varicella ^,,, vaccines are not associated with a significantly reduced incidence of uveitis in the general population, most have focused on adverse events or the risk of recurrence in individuals with a prior history of uveitis. Few investigations have explored whether vaccinated individuals may have a reduced risk of developing new-onset noninfectious uveitis, and they report differing results. Indeed, the study of uveitis following HPV, varicella, and zoster vaccination primarily relies on case reports. ^,, More recently, the influence of COVID-19 vaccination on the development of uveitis has been explored with mixed results. ^, The potential protective benefit of vaccinations may stem from either a direct immunomodulatory effect or the prevention of systemic viral infections that could otherwise trigger immune dysregulation. By limiting exposure to immunogenic viral antigens, vaccines may indirectly reduce the risk of immune-mediated ocular inflammation.

In this study, we analyzed the incidence of idiopathic uveitis in a large, real-world cohort to explore whether vaccination against COVID-19, HPV, or varicella zoster virus, including both the primary varicella vaccine and the recombinant or inactivated herpes zoster vaccines, is associated with a lower risk of developing new-onset idiopathic uveitis (NIU) compared to unvaccinated individuals.

This retrospective study utilized the US Collaborative Network in the TriNetX platform. This network contains deidentified EHR data of over 120 million patients from over 60 healthcare organizations. The platform includes standardized EHR data using International Classification of Diseases (ICD) codes with records spanning from 2006 to 2025. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology reporting guidelines. The data reviewed is a secondary analysis of existing data, does not involve intervention or interaction with human subjects, and is deidentified per the deidentification standard defined in Section §164.514(a) of the HIPAA Privacy Rule. The process by which the data is deidentified is attested to through a formal determination by a qualified expert as defined in Section §164.514(b)(1) of the HIPAA Privacy Rule. This formal determination, updated in December 2020, replaces the previous waiver granted by the Western Institutional Review Board. This study adhered to the tenets of the Declaration of Helsinki.

To assess associations between vaccination and NIU, a cohort of patients with an encounter procedure code for each single respective vaccination was compared to a control cohort without an encounter procedure code of the respective vaccination. This study utilized International Classification of Diseases, 10th Revision (ICD-10), Anatomical Therapeutic Chemical (ATC), Vaccine Administered (CVX), RxNorm, and Current Procedural Terminology (CPT) codes. Included vaccinations included COVID-19 (CVX 213, CPT 91301-91309, CPT 91311-91322), HPV (CVX 137, RxNorm 798262, 798264, 798266, 798268, 1596930-1596934), Varicella (CVX 21, CPT 90716), Recombinant Herpes Zoster (CPT 90750), and Live Herpes Zoster (CPT 90736). Patients diagnosed with known infectious or noninfectious causes of uveitis prior to the index date and those diagnosed with infectious uveitis at any time were excluded. Exclusion criteria were based on prior large-scale EHR research on uveitis, with additional inflammatory conditions addressed through propensity score matching (PSM). As HPV is not regularly tested, we utilized several ICD-10 codes that capture disease attributed to HPVs. All codes utilized can be found in Table S1. All patients required at least two ophthalmology encounters and one general healthcare encounter at least 3 months prior to their ophthalmology encounters to adjust for bias, which may occur from patients not regularly following with a contributing institution.

Our primary outcomes included incidence of idiopathic uveitis at 3, 6, and 12 months after the respective vaccination code in patients without history of viral disease. A composite NIU outcome was also utilized to capture subgroups including anterior uveitis (H20.00-2, H20.04, H20.1, H20.81, and H20.9), intermediate uveitis (H30.2), posterior uveitis (H30.0, H30.1, H30.89, H30.9, H30.14, and H35.06) and panuveitis (H20.82, H30.81, and H44.11). Patients with history of any NIU subgroup before the index date were excluded for that component of the analysis. A secondary analysis was conducted to assess risk of NIU, excluding history of viral disease.

To minimize confounding, PSM was performed using the TriNetX analytic feature with greedy 1:1 matching without replacement and a caliper of 0.25 standard deviation (SD). Matching was conducted based on patient age at index, sex, race, ethnicity, nicotine dependence, alcohol dependence, obesity, vitamin D deficiency, asthma, ankylosing spondylitis, systemic corticosteroid use, and problems related to education, literacy, housing, and economic circumstances. Covariate balance between the matched cohorts was assessed using absolute standardized differences (std. diff.), where values greater than 0.10 indicated imbalance. All statistical analyses were performed through the TriNetX platform. Descriptive statistics are presented as mean ± SD or as count (%). Risk ratios (RR) and their 95% confidence intervals (CI) were reported. To account for multiple statistical comparisons, RRs were considered significant if their CIs fell below 0.90 or above 1.10. Forest plots were generated using R statistical computing software (Version 4.4.1, R Core Team 2024).