Treatment outcomes and clinicomicrobiological characteristics of a protocol-based approach for neonatal endogenous endophthalmitis

Abstract

Purpose: To describe the outcomes and clinicomicrobiological profile of 11 years of a protocol-based management in neonatal endogenous endophthalmitis.

Methods: This was a retrospective interventional study of endogenous endophthalmitis in 31 eyes of 26 neonates. The protocol for active infection included systemic antimicrobials, vitreous and/or aqueous tap, and intravitreal antimicrobials under topical or general anesthesia along with core vitrectomy in selected cases. Blood, urine, umbilicus, aqueous, and vitreous samples underwent microbiological evaluation. Retinopathy of prematurity screening and treatment were done when indicated. Primary outcome was anatomic status assessed by comprehensive eye examination and by fundus photography whenever possible.

Results: Twenty-one of 26 babies (81%) were preterm. Two types of presentations included those with a fulminant appearance (24 eyes) and those with focal retinitis detected during routine screening (7 eyes). Vitreous culture was positive in 12/20 eyes (60%). Pseudomonas aeruginosa (8) was the most common isolate. Incorrect initial diagnosis was common. Treatment included intravitreal injections in 26 eyes, 10 of which also underwent vitrectomy. Twenty-four of the 26 patients (92%) received parenteral antimicrobials and 17 had evidence of systemic infection. All eyes with a fulminant presentation developed phthisis, while all focal fungal cases were salvaged.

Conclusions: Neonatal endogenous endophthalmitis has 2 distinct presentations. Focal retinal infections have good visual and anatomical outcomes while fulminant nosocomial cases do poorly. Management under topical anesthesia can be an alternative strategy for sick babies that cannot undergo surgery under general anesthesia due to systemic morbidity. Awareness about early diagnostic signs may help early referral.