Evaluation of Risk Factors of Endophthalmitis Requiring Evisceration/Enucleation among Patients with Endophthalmitis at KCMC from 2020–2024

Abstract

Aim: To evaluate the risk factors associated with endophthalmitis requiring evisceration/enucleation among patients managed at Kilimanjaro Christian Medical Centre (KCMC) in Northern Tanzania. Study Design: Hospital-based retrospective analytic study. Place and Duration of Study: Kilimanjaro Christian Medical Center (KCMC) Eye Department, Moshi, Tanzania, from January 2020 to December 2024. Methodology: A total of 211 patients diagnosed with and treated for endophthalmitis during the study period were included. Clinical records were reviewed, and relevant demographic, clinical, and treatment data were extracted. Patients were grouped by outcome into those with eyes preserved and those who underwent evisceration. Modified Poisson regression was used to assess associations between potential risk factors and endophthalmitis requiring evisceration, with adjusted prevalence ratios and 95%confidence intervals reported. A P-value <0.05 was considered statistically significant. Results: Of 15,171 admissions to the KCMC eye department between January 2020 and December 2024, 211 patients were diagnosed and managed for endophthalmitis, yielding a prevalence of 1.39%. Among these, 38 (18%) underwent evisceration. In multivariable analysis, two factors were independently associated with increased risk of evisceration: corneal ulcer-related etiology (APR: 3.18; 95% CI: 1.54–6.60; P=0.002) and poor presenting visual acuity of 5 logMAR (APR:3.19; 95%CI: 1.95- 5.21; P<0.001). Other variables, such as delayed presentation, lower IOP, and lack of vitrectomy, were significant in bivariable analysis but did not remain significant after adjustment. Conclusion: Corneal ulcer etiology and poor presenting visual acuity were the strongest predictors of evisceration among patients with endophthalmitis at KCMC. These findings underscore the need for early diagnosis, prompt referral, and timely surgical intervention. However, the interpretation of these results should consider the retrospective cross-sectional design, which limits causal inference, and the absence of endogenous endophthalmitis cases, which may reduce generalizability.