Predictors of Chlamydia trachomatis conjunctivitis in neonates: a 10-Year retrospective study.

Abstract

Background: Ophthalmia neonatorum (ON) is a common neonatal ocular condition with potentially serious ocular and systemic complications. The spectrum of causative organisms varies by geographical regions, maternal health practices, and over time. Chlamydia trachomatis remains a significant pathogen with non-specific symptoms that overlap with other infections. This study aims to assess local burden of Chlamydia trachomatis and identify clinical predictors.

Methods: We conducted a 10-year retrospective review (2014-2023) of neonates presenting with suspected ON at a tertiary paediatric eye centre in Singapore. Clinical and microbiological data were analysed to determine etiological trends and identify predictors of C. trachomatis conjunctivitis. Diagnostic methods included Gram stain, culture, immunofluorescence, and PCR testing. Multivariate logistic regression was used to assess associations.

Results: A total of 797 neonates were included; 140 (17.5%) tested positive for Chlamydia trachomatis followed by Staphylococcus aureus (10.2%) and MRSA (4.1%), while Neisseria gonorrhoeae was rare (0.75%). Ten-year trend showed a consistent pattern with Chlamydia trachomatis been the highest causative organism followed by Methicillin-sensitive Staphylococcus aureus. The presence of bloody discharge was the strongest predictor for Chlamydia infection (OR = 20.99, 95% CI: 6.09-72.29, p < .001), followed by vaginal delivery (OR = 17.63, 95% CI: 5.59-55.51, p < .001). Additional significant predictors included eyelid swelling (OR = 3.04, 95% CI: 1.58-5.83, p < .001) and conjunctival redness (OR = 2.73, 95% CI: 1.73-4.29, p < .001).

Conclusions: Chlamydia trachomatis remains the leading cause of ON in our cohort. Key clinical predictors-bloody discharge, eyelid swelling, conjunctival redness, and unilateral involvement-can assist clinicians in early recognition and timely initiation of targeted testing and therapy. These findings support the development of predictive diagnostic frameworks in settings with limited access to molecular diagnostics.