A total of 259 (24.9%) of the 1,042 eyes examined had vision impairment, or Snellen VA worse than 20/50. Vision impairment was found in 18.0% of better-seeing eyes and 32.0% of worse-seeing eyes. The median logMAR VA was 0.23 (Snellen VA 20/34) in all eyes, 0.19 (Snellen VA 20/30) in the better-seeing eyes, and 0.30 (Snellen VA 20/40) in the worse-seeing eyes. Multivariable analysis revealed that the presence of vision impairment was significantly associated with cataract (odds ratio [OR] = 7.68; p<0.001), uveitis (OR = 2.08; p = 0.007), corneal scar (OR = 4.23; p = 0.001), and optic neuropathy (OR = 6.32; p = 0.034) (Table 6).

This retrospective observational study included a large cohort of 521 EVD survivors who underwent comprehensive ophthalmic examination at long-term follow-up, ≈3.5 years after their initial ETU admission for acute EVD. Results of the study demonstrate a broad spectrum of ophthalmic findings that were associated with vision impairment, including cataract, uveitis, corneal scar, and optic neuropathy.

Uveitis remained one of the most common ophthalmic diagnoses found, affecting 8.4% of EVD survivors. The prevalence of uveitis in our study was lower than in other studies from Sierra Leone also conducted during the West Africa EVD outbreak, which reported uveitis in 18%-34% of EVD survivors a few months after ETU discharge (8,11,12). In those studies, EVD survivor status was determined on the basis of EVD survivor certificates, history of ETU admission, and laboratory diagnostics when available. Some EVD survivors in our cohort might have been treated for uveitis in the previous few years; 87% reported having >1 previous eye exam, which was also associated with the presence of uveitis (p = 0.05). Another explanation is that EVD-associated uveitis develops shortly after the acute illness; risk declines over time, potentially in relation to the viral load in ocular tissues. One previous study showed that a higher viral load at acute EVD illness was associated with the development of uveitis (11). Viable Ebola virus was detected from the ocular fluid of an EVD survivor at 14 weeks in association with panuveitis but was no longer detected at 27 months, when the uveitis had been inactive for 3 months (17). In the Ebola Virus Persistence in Ocular Tissues and Fluids (EVICT) study of survivors being evaluated for cataract surgery, 46 survivors with vision-impairing cataract tested negative for Ebola virus RNA in the aqueous humor at 19 months and 34 months after acute EVD (18,19). On the basis of those data, it is possible that EVD survivors have a greater risk for uveitis immediately after acute EVD as a result of viral persistence and that risk for uveitis decreases over time as the virus clears. Further studies will clarify the relationship between viremia, viral persistence in immune-privileged sites such as ocular tissues, and the development and progression of uveitis.

Nearly 20% of EVD survivors in this cohort had moderate vision impairment of 20/50 or worse in their better-seeing eye, which indicated ongoing ophthalmic disease at long-term follow-up. Ophthalmic findings associated with vision impairment in EVD survivors, including uveitis, cataract, corneal scar, and optic neuropathy, indicate a range of ocular disease that may contribute to visual illness and requires ongoing management. The EVICT study reported similar findings that vision-impairing cataract, posterior synechiae, optic neuropathy, and retinal detachment were strongly associated with worse logMAR VA in EVD survivors referred for vision impairment or cataract evaluation (20). In a population-based study conducted in 2021in Sierra Leone, 5.1% of participants were bilaterally blind and ≈16% had varying degrees of vision impairment (21). The most common causes of bilateral blindness in the general population of Sierra Leone were untreated cataracts (59.4%), glaucoma (21.7%), and nontrachomatous corneal opacity (8.4%). Of note, the participants in the population-based study were >50 years of age, whereas the median age in our cohort of EVD survivors was 29 years. In addition, uveitis was not reported as a major cause of blindness or vision impairment in the general population.

Cataracts were found in 11.2% of eyes in our cohort and were strongly associated with vision impairment. Our analysis showed that uveitis found in the study exam was associated with a previous diagnosis of cataract (p<0.001); ≈10% of newly diagnosed cataracts were uveitic in nature. Considering the relatively young age of this cohort, a substantial number of the cataracts are likely attributable to previous or ongoing inflammation or treatment with topical corticosteroids. Cataract development related to uveitis was reported in other studies. Tiffany et al. (12) showed that 7 out of 8 EVD survivors with cataracts had concurrent uveitis. Mattia et al. (11) found that 10% of EVD survivors with uveitis had concurrent early cataracts and were relatively young (median age 29 years), suggesting that that the cataracts were not related to age. Thus, early detection and treatment for uveitis in EVD survivors could reduce the frequency of cataract development and improve visual outcomes.

Corneal scar and optic neuropathy were less common in EVD survivors but were also strongly associated with vision impairment status. A post hoc analysis of the EVICT study corroborated findings of worse logMAR VA in survivors with optic nerve disease than those without optic nerve disease (22). However, the presence of ocular surface disease, including dry eye, band keratopathy, and corneal scar, were not strongly associated with worse visual outcomes (23). Further studies will determine whether these ophthalmic manifestations are more common in EVD survivors than control participants and are associated with worse vision. To understand the prevalence of uveitis, visual impairment, and disease pathogenesis of ocular inflammation in EVD survivors compared with close-contact control patients, a longer-term study is currently underway in Sierra Leone (24).

Limitations of our study include the retrospective study design and lack of a control group. Timing or directionality of associations cannot be determined because this was a cross-sectional study. In addition, Ebola serum IgG was not available, so EVD survivor status was determined by survivor certificates and a history of ETU admission. Finally, there is also a potential for selection bias for symptomatic survivors that could skew the results toward a higher prevalence of ophthalmic findings; however, we recruited for examination any EVD survivor in the targeted regions who could be reached.

In summary, this study identified a range of ophthalmic complications that were associated with vision impairment in EVD survivors >3 years after the resolution of their acute illness. Uveitis remained one of the most common ophthalmic findings in EVD survivors, but cataract, corneal scars, and optic neuropathy were also found to be associated with poor visual outcomes. Our findings highlight the need for long-term vision care and follow-up for uveitis as well as a range of ophthalmic conditions in EVD survivors to optimize their visual potential and associated quality of life.