Predictors of Post-treatment Visual Improvement After Bacterial Keratitis Infection at a Referral Center in Southern California.

PURPOSE: To evaluate visual outcomes after bacterial keratitis (BK) and identify predictive factors for poor prognosis at a tertiary referral center in Southern California.

METHODS

This is a cross-sectional retrospective review of patients' medical records with culture-positive BK at University of California Los Angeles from January 1, 2014, to December 31, 2019. Main outcome measure was change in best-corrected visual acuity (BCVA) at 12 weeks posttreatment. Statistical analyses included ANOVA for independent populations, Pearson correlation, and binary logistic regression. P -value <0.05 was considered significant.

RESULTS

A total of 317 patients with 374 culture-positive cases were identified. Mean age was 62.1 ± 19.9 years, 52% were female, and 68.1% of isolates were Gram-positive. A total of 114 cases occurred during the summer months (June-August), and 29.8% had a history of contact lens wear, trauma, or systemic disease. Median symptom-to-presentation time was 3.5 days (range 1-60). Mean BCVA improved from 1.84 ± 0.91 at presentation to 1.40 ± 0.96 log of the minimum angle of resolution at 12 weeks (mean difference = 0.44, 95% confidence interval, 0.34-0.53; P < 0.001). Increasing age correlated negatively with change in BCVA (Pearson β = -0.167, P < 0.008). Multivariate analysis identified age >60 years, presence of systemic comorbidities, and presenting BCVA >1.00 log of the minimum angle of resolution as predictors of poor visual outcome.

CONCLUSIONS

This study provides the largest contemporary analysis of BK in Southern California and establishes a clinically applicable risk stratification model. Older age, systemic comorbidities, and poor presenting vision were independently associated with worse visual outcomes. These findings support earlier identification of high-risk patients, tailored treatment strategies to improve visual prognosis and optimize care delivery in BK.