Purpose: To examine the discriminatory performance of donor corneoscleral rim cultures for predicting endophthalmitis after corneal transplantation.
Design: Systematic literature review.
Participants: Studies that reported the prevalence of donor rim cultures after refrigerated preservation of donor corneas distributed for keratoplasty.
Methods: Random-effects meta-analysis estimated pooled sensitivity, specificity, likelihood ratios, and diagnostic odds ratios. Meta-regression and stratification explored study-level reasons for diagnostic performance.
Main outcome measures: Incidence of postkeratoplasty endophthalmitis.
Results: Of 17,614 corneal grafts, 2459 (14%) had a positive donor rim culture and 31 (0.2%) developed endophthalmitis. Twenty-one had concordant donor and recipient isolates, including 10 with Candida species. The sensitivity of donor rim cultures among 10 studies reporting postsurgical endophthalmitis was 67% (95% confidence interval [CI], 47%-83%), and specificity averaged 85%. Endophthalmitis occurred 12 (95% CI, 5-29) times more often among recipients of a culture-positive donor cornea. With Bayesian analysis, a culture-positive donor cornea would raise the infection risk to 1%, whereas fungal isolation from the donor rim predicts a 3% probability of fungal endophthalmitis. Pooling of studies was limited by a significant discrepancy among reports that could not be explained by differences in antibiotic supplementation of the preservation medium, method of culture inoculation, or type of culture medium.
Conclusion: Endophthalmitis after penetrating keratoplasty is more likely with a culture-positive donor rim, notably candidal endophthalmitis from fungal contaminants, but better evidence is needed to determine the prognostic value and manner of routine microbiological screening.