Ophthalmia neonatorum (ON) remains the most common ocular condition in the neonatal period, with the potential for significant morbidity if not diagnosed and treated promptly. Its incidence and causative microbial agents vary widely across regions, particularly in relation to socio-economic status, and over different time periods. These variations reflect underlying differences in the prevalence of maternal infections, the extent of prenatal screening, the quality of postnatal care, and local protocols for neonatal ocular prophylaxis [14].

Historically, Neisseria gonorrhoeae was a leading cause of ON in many regions. A study from Singapore (1983–1986) reported Neisseria gonorrhoeae in 68% of culture-positive cases [15]. However, with the decline in gonorrhoeal sexually transmitted infections in the industrialized countries during the 1970 s and 1980 s, its prevalence in ON has significantly decreased. In our study, only 6 out of 797 neonates (0.75%) had N. gonorrhoeae–related conjunctivitis.

In contrast, Chlamydia trachomatis emerged as the predominant pathogen in our 10-year cohort, detected in 140 cases (17.5%). Other organisms identified included Methicillin-sensitive Staphylococcus aureus (10.2%), MRSA (4.1%), Escherichia coli (3.5%), Corynebacterium macginleyi (2.3%), and mixed growths (2.3%). These findings are consistent with global literature, which typically reports either Chlamydia trachomatis or Staphylococcus aureus as the most common causes of neonatal conjunctivitis.

In a study from Hong Kong, Staphylococcus aureus was reported as the leading pathogen, accounting for 36% of culture-positive cases, followed by Chlamydia trachomatis (21%) [16]. In contrast, a study from Ireland identified Chlamydia trachomatis as the most frequent isolate, with Staphylococcus aureus being less common (20.4% versus 18.5%) [7] Five-year retrospective studies conducted in Finland and Korea also reported Staphylococcus aureus as the predominant organism, representing 16% and 52% of cases, respectively. Notably, Chlamydia trachomatis was not detected in either of these two cohorts [8, 17].

Data from a survey conducted among 449 global ophthalmologists members of American Association of Paediatric Ophthalmology and Strabismus (AAPOS) identified Chlamydia trachomatis as the most common cause of ON accounting for 35.4% followed by Staphylococcus aureus (20%) [18].

Across our 10-year study period, Chlamydia trachomatis consistently remained the leading cause of ON, with annual case counts ranging from 9 to 19. Methicillin-sensitive Staphylococcus aureus and MRSA were the second and third most common pathogens, respectively over the years. In the latter half of the study period, we observed an increase in Gram-negative organisms including Escherichia coli, Klebsiella pneumoniae, and Moraxella catarrhalis. Additionally, Corynebacterium macginleyi—a less commonly reported pathogen—was isolated exclusively in recent years. This shift in microbiological patterns is supported by data from Southern China, where a 15-year review noted a declining trend in Gram-positive isolates and a concurrent rise in Gram-negative pathogens in neonatal conjunctivitis [19].

In our study, only 35.39% (282/797) of samples were culture-positive, highlighting the limitations of conventional culture techniques. Newer technique of metagenomics with next generation sequencing allows high yield and unbiased characterization of the ocular surface microbiota, consisting of commensal bacteria, and can improve the diagnostic yield in infective cases as compared to conventional culture techniques [20].

Diagnosing Chlamydia trachomatis conjunctivitis in neonates remains challenging due to its non-specific symptoms which overlap with other bacterial or viral cause. Mucopurulent discharge, conjunctival redness, and swollen eyelids have been described as characteristic signs of chlamydial infection [9]. Chang et al. further identified blood-stained discharge as a highly specific clinical marker, with both specificity and positive predictive value reported at 100%. Periorbital edema was also found to be highly specific (82%) for Chlamydia trachomatis infection [16].

Our multivariate analysis identified bloody discharge as the strongest independent predictor of Chlamydia trachomatis infection. Other statistically significant predictors included eyelid swelling, conjunctival redness, and unilateral eye involvement. The frequent unilateral presentation may reflect the localized inflammatory response characteristic of early-stage chlamydial conjunctivitis. Moreover, a borderline association with later onset of symptom was observed, suggesting that Chlamydia trachomatis infections present slightly later than other microbial causes, though this trend did not reach statistical significance.

Additionally, vaginal delivery was found to be a significant risk factor for chlamydial conjunctivitis, likely due to vertical transmission during passage through an infected birth canal. This finding is consistent with a recent meta-analysis, which reported a higher incidence and prevalence of ON in infants delivered vaginally compared to those born via Caesarean Sect. [14].

In view of difficulty in clinically distinguishing between various aetiologies of ON, Gram stain, immunofluorescence (DIF), PCR and bacterial culture remains the standard to establish a definitive diagnosis. A recent study from the United States involving 1,870 neonates highlighted significant variability in diagnostic practices. Gram stains were performed in only 13.1% of cases, while bacterial cultures were obtained in 21.3% of which 51.3% of these cultures yielded positive isolates [21]. Gram staining also has limited utility in the detection of Chlamydia trachomatis due to its intracellular nature. In our study, only 4 out of 140 (2.8%) confirmed Chlamydia trachomatis cases were positive on Gram stain, further reinforcing the need for targeted diagnostic methods. Direct immunofluorescence assays were used to identify chlamydia infection for the first 7 years of our study. It was replaced by PCR providing a rapid, highly sensitive, and automated molecular method, hence improving the diagnostic yield [22].

Given the high prevalence of Chlamydia trachomatis and the difficulty in confirming the diagnosis clinically, the identification of reliable clinical predictors is crucial. The clinical predictors identified in our study, particularly bloody discharge and mode of delivery are highly relevant in diverse healthcare environment. In resource rich centres, these predictors will improve efficiency by directing towards targeted PCR testing. In settings with limited or no access to molecular diagnostics, they may serve as practical clinical surrogates to guide early initiation of treatment and reduce the risk of complications.

Our large sample size and 10-year follow-up provide a robust overview of the local.

epidemiological trends and causative burden of ON, highlighting the changing microbial patterns over time. Awareness of the predominant pathogens can guide empiric antibiotic choice and provide better counselling to parents about prognosis and treatment expectations. From a public health perspective, monitoring shifts in causative organisms supports updates to screening protocols, infection control measures, and the rational use of molecular diagnostics.

The main limitation of this study is its retrospective design, which inherently limits the ability to control for confounding factors and introduces the possibility of incomplete or missing clinical data. There may be variability in the recognition and documentation of clinical signs, as multiple physicians at different stages of training were involved in patient assessment. In addition, the evolution of diagnostic testing over the study period—from immunofluorescence (DIF), which is an operator dependent test requiring expertise to visually read the slide to highly sensitive and specific automated polymerase chain reaction (PCR)—may have influenced pathogen detection rates.

Finally, as this study was conducted at a single tertiary paediatric centre in Singapore, the findings may not be fully generalizable to other settings with differing population demographics, healthcare access, and diagnostic practices.

In the future, metagenomic approaches to studying the ocular microbiota in neonates and children may provide a more comprehensive understanding of both normal and pathogenic flora. Such insights may facilitate more accurate diagnosis of ON and support the development of targeted therapies for neonatal conjunctivitis, ultimately improving patient outcomes.