Microbial keratitis in north-western Spain: a review of risk factors, microbiological profile and resistance patterns.

Abstract

Purpose: To review the risk factors, clinical characteristics, and microbiological profiles of microbial keratitis cases, as well as the antibiotic resistance patterns of bacterial isolates in the region of Galicia, Spain.

Methods: This retrospective case series includes patients with culture-positive non-viral microbial keratitis between 2010 and 2020, treated at nine hospitals within the region of Galicia, North-West Spain. The standard protocol involved Gram staining for bacterial infections and calcofluor white staining for fungal or amoebal infections, identification by MALDI-TOF mass spectrometry or microscopy, and antimicrobial susceptibility interpreted according to EUCAST or CLSI guidelines.

Results: 780 microorganisms were isolated from corneal scraping cultures from 654 patients. 36.9% resided in urban areas, and 63.1% in rural areas. Isolates were more frequently collected in spring and summer. The median time to corneal scraping was 0 days (IQR 0-2), and the median time to epithelialisation was 24.0 days (IQR 11-49). Most cases had a single corneal infiltrate (509 cases; 77.8%) and affected the stroma (432; 66.1%), with small (< 3 mm) epithelial defects (347; 53.1%). Significant risk factors included contact lens wear (24.2%) and exposure to organic matter (4.9%). The most frequent bacteria was CoNS (207; 26.4). Fungi (77; 9.9%) and amoebae (6; 0.8%) were less common. Steroid use and eyelid disease increased resistance in CoNS species. An increase in the percentage of MRSA (compared to MSSA) was detected over the study period (p = 0.045).

Conclusions: In Galicia (Spain), microbial keratitis was mostly attributed to CoNS. An increase in MRSA keratitis was observed. Analysis of risk factors may help in suspecting antibiotic resistance. Surveillance programs for detecting the development of antimicrobial resistance are necessary to provide treatment guidelines based on local data.