Epidemiology of Microbial Keratitis on Uganda: A 7-year update

Abstract

<ns3:p> Purpose To provide the latest update on the epidemiology of microbial keratitis in Uganda. Methods We prospectively recruited patients presenting with microbial keratitis (MK) at three eye units in Uganda between November 2023 and June 2025. Data collected included demographic characteristics, clinical history, presenting features, and laboratory findings. Participants were specifically asked about prior ocular trauma and use of traditional eye medicines (TEM). All centres performed microscopy and culture, while in vivo confocal microscopy (IVCM) and polymerase chain reaction (PCR) were additionally performed at Mbarara University and Referral Hospital Eye Centre. Results We recruited 369 patients with MK. The median age of the participants was 45 years (range: 18 — 88); and the majority were male (60.7%). The median presentation time was 17 days from onset (IQR 8–32) with only 4.9% of the patients presenting early (within 3 days). At presentation, most (60.8%) of the patients were blind in the affected eye. The overall diagnostic yield was 52.2%, fungal keratitis being the most common. The factors that were associated with a poor presenting visual acuity included distance from the hospital (p &lt; 0.001), use of traditional eye medicine (p &lt; 0.048) and delayed presentation (p &lt; 0.001). These results were comparable to the baseline findings from our previous work ( <ns3:italic>Epidemiology of microbial keratitis in Uganda – A cohort study</ns3:italic> ) seven years ago. Clinical and laboratory diagnoses aligned moderately for fungal keratitis but poorly for bacterial disease. Conclusion There has not been a significant change in the epidemiology of MK in Uganda. MK is still very common in different regions in the country. Fungal keratitis remains the leading cause of MK in Uganda. The use of advanced diagnostic tests like PCR and IVCM can help improve yields in the identification of the causative microorganisms in patients with MK. </ns3:p>