Outcomes and Clinical Features Predictive of Fungal Endophthalmitis

Purpose: To analyze intravitreal antifungal injections performed at a tertiary center to determine (1) risk factors increasing fungal endophthalmitis likelihood at the time of patient presentation, (2) prognostic factors at presentation, and (3) validity of American Academy of Ophthalmology (Academy) ophthalmologic Candida septicemia (candidemia) screening guidelines. Design: Single-center, retrospective clinical case-control study. Methods: Clinical course, visual outcome, and final diagnosis were analyzed for 75 patients (81 eyes) receiving intravitreal antifungal injections between 2014 and 2021. Features were compared between fungal endophthalmitis and clinically similar diseases (masquerades). Results: Fungal endophthalmitis was more likely than masquerade based on injection in emergency department or inpatient setting (P =.0002) vs outpatient, greater visual acuity (P =.049), artificial indwelling line present (P =.0004), sepsis within past 6 months (P =.0002), prior/current hepatitis C diagnosis (P =.044), total parenteral nutrition (P =.0028), complicated diabetes mellitus (P =.035), actively treated cancer (P =.021), immunosuppressive medication within past year (P =.035), immunocompromising condition number (P =.031), and delayed pain (P =.0094) and vision loss (P =.020) onset. Visual acuity at presentation correlated with visual outcome (R² = 0.3524, P = 8.5 × 10^–7). No patients with fungal endophthalmitis were asymptomatic or identified by screening. Conclusions: Many conditions can mimic fungal endophthalmitis, but certain risk factors should raise its suspicion. Regardless of diagnosis, presenting vision correlates with final vision. Lastly, no patients with fungal endophthalmitis were asymptomatic or identified by screening, further supporting current Academy candidemia screening guidelines.