Prospective Eastern India study links pediatric microbial keratitis profiles to age

A prospective study of children treated at a tertiary eye-care center in Eastern India found distinct clinical and microbiological patterns in pediatric microbial keratitis according to age. Among 86 children with 99 affected eyes, those aged 5 years or younger were more likely to have systemic, nutritional and bilateral disease, while children older than 5 years more often had corneal injury, microbiologically confirmed infection and fungal keratitis.

The researchers recruited consecutive children with clinically diagnosed microbial keratitis at the cornea clinic and emergency services between April 2024 and September 2025. Eligibility required a corneal stromal infiltrate with an overlying epithelial defect, with or without hypopyon. The analysis compared children aged 5 years or younger with those older than 5 years and recorded clinical, microbiological, nutritional, treatment and outcome data through three months.

The cohort included 42 children aged 5 years or younger and 44 older than 5 years. The mean age was 5.9 ± 4.8 years, and the reported range extended from 8 days to 14 years. Fifty children were male and 36 female. Thirteen children had bilateral keratitis; all were aged 5 years or younger. Most participants lived in rural areas, 24 of 86 were undernourished, and 28 of 86 were incompletely immunized.

Systemic illness was reported as a predisposing factor in 57.1% of children aged 5 years or younger, compared with 9.1% of those older than 5. Vitamin A deficiency occurred at presentation in 42.9% of the younger group, including 73% of infants in that group. Undernutrition was also more frequent among younger children, reported in 42.9%, whereas corneal injury was the predominant risk factor among children older than 5 years, reported in 81.8%. Bilateral keratitis was more frequent among undernourished and incompletely immunized children than among their respective comparison groups.

Mean symptom duration before presentation, ulcer size and ulcer severity did not differ significantly between the two age groups. Perforation at presentation was numerically more frequent in children older than 5 years, at 31.8% versus 18.2%, but the difference was not statistically significant. Prior use of an antibiotic-corticosteroid combination was significantly more common in the older group. The mean duration of symptoms before presentation was nearly 12 days for the cohort as a whole.

Microbiological evaluation identified an organism in 51.5% of the 99 eyes. Smear and culture results were negative in 48.5%, which the investigators classified as unidentified or sterile. All microbiologically confirmed eyes were smear-positive, while 36.4% of all eyes were culture-positive. Confirmation was higher among children older than 5 years, at 68.2%, than among those aged 5 years or younger, at 38.2%. Fungal keratitis was also significantly more common in the older group. Among confirmed infections, Aspergillus was followed by Fusarium as the most frequently reported fungal organisms, while bacterial isolates were mainly gram-positive cocci, particularly coagulase-negative staphylococci.

Treatment initially consisted of broad-spectrum topical antibiotics, modified according to clinical response and microbiological results. Topical natamycin was added when fungi were identified. Corneal ulcers resolved with medical treatment alone in 59.6% of eyes. Overall, 40 of 99 eyes required surgical intervention: 20 received cyanoacrylate glue with a bandage contact lens or a Tenon’s patch graft, and 20 underwent therapeutic penetrating keratoplasty. In multivariable analysis, larger ulcers and longer keratitis duration were associated with surgery, while severe ulcer grade and longer duration were associated with perforation at presentation.

At three months, corneal opacity was the most common recorded outcome in both age groups. Severe ulcer grade and incomplete immunization were associated with a poor outcome at three months. Anterior staphyloma occurred only among children aged 5 years or younger, while phthisis bulbi and secondary glaucoma occurred in both groups without a statistically significant difference in the overall outcome profile. Quantitative visual acuity could be assessed only in children older than 5 years; among those assessed, mean best-corrected visual acuity improved from 1.49 ± 0.69 LogMAR at presentation to 0.75 ± 0.51 LogMAR at three months.

The findings are limited by the single-center referral setting, which may have enrolled a disproportionately severe and recalcitrant population, and by the three-month follow-up. Visual outcomes could not be compared directly between age groups because quantitative acuity testing was not possible in many younger children. Deprivation amblyopia and longer-term graft survival, scar progression and visual rehabilitation were also not adequately assessed. The study further lacked advanced microbiological methods such as in vivo confocal microscopy, polymerase chain reaction and next-generation sequencing, so organisms missed by conventional smear and culture techniques may have affected the reported microbiological profile.

Subject of Research: Age-related clinical, microbiological and outcome differences in pediatric microbial keratitis in Eastern India

Article Title: Age-related differences in the clinical and microbiological profile of pediatric microbial keratitis in Eastern India

Article References: Kusumesh, R., Panday, N., Arya, L. K., Ahmar, R., Ambasta, A., Kumari, A., Kumari, N., Sinha, B. P., & Mohan, N. (2026). Age-related differences in the clinical and microbiological profile of pediatric microbial keratitis in Eastern India. PLOS One, 21(10), e0359943. https://doi.org/10.1371/journal.pone.0359943

Image Credits: AI Generated

DOI: 10.1371/journal.pone.0359943

Keywords: pediatric microbial keratitis, corneal ulcers, age differences, vitamin A deficiency, undernutrition, fungal keratitis, corneal trauma, microbiological culture, therapeutic keratoplasty, Eastern India, Age-related, differences