Outcomes with Endogenous Endophthalmitis from Candidal versus Bacterial Sepsis

To compare 1-year ophthalmic outcomes in endogenous endophthalmitis after candidal versus bacterial sepsis using a large, multicenter electronic health record database.

A retrospective propensity score-matched cohort study with a global, federated database.

Patients in the TriNetX Analytics Network who were diagnosed with endophthalmitis and had antecedent candidal or bacterial sepsis within 7 days were identified.

Patients were matched 1:1 by propensity score for age, sex, race, ethnicity, type 2 diabetes mellitus, human immunodeficiency virus infection, malignant neoplasms, liver disease, and transplanted organ or tissue status.

One-year risk ratio (RR) of incident retinal detachment, vitreoretinal intervention, and all-cause mortality. A secondary analysis assessed composite vision or globe loss without exclusion of prior diagnoses. The number of intravitreal injections administered was also compared between cohorts.

In total, 318 patients with candidal, and 1173 patients with bacterial sepsis and endophthalmitis were identified. After matching, each cohort included 304 patients with well-balanced baseline characteristics. Compared with bacterial sepsis, candidal sepsis was associated with significantly lower rates of retinal detachment (3.8% vs. 7.8%; RR, 0.49; 95% confidence interval [CI], 0.24-0.99), vitreoretinal intervention (4.1% vs. 14.2%; RR, 0.29; 95% CI, 0.15-0.53), and composite vision and globe loss in the secondary analysis (4.3% vs. 26.6%; RR, 0.16; 95% CI, 0.09-0.28). The mean number of intravitreal injections administered per patient did not significantly differ between cohorts (mean ± standard deviation, 2.25 ± 4.03 candidal vs. 1.23 ± 0.65 bacterial; P = 0.209). One-year mortality was significantly higher with candidal sepsis (35.4% vs. 25.3%; RR, 1.40; 95% CI, 1.09-1.79; P = 0.007).

In this large-scale, propensity score-matched analysis, candidal sepsis was associated with significantly lower rates of retinal detachment and vitreoretinal intervention compared with bacterial sepsis, despite higher all-cause mortality. These divergent findings indicate that endogenous Candida endophthalmitis, despite its greater systemic morbidity, may carry lower ophthalmic risk than bacterial disease. Multicenter, prospective studies with visual acuity data are needed to further characterize pathogen-specific differences in ophthalmic risk.

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.