Ophthalmology · Medical Literature
[Articles] Regional and temporal trends in antimicrobial susceptibility among isolates from bacterial keratitis: a systematic review and meta-analysis
Abstract
Regional differences exist in antimicrobial resistance for some clinically relevant drug–pathogen combinations. A standardised reporting template would reduce heterogeneity and facilitate future research.
Clinical relevance
Do antimicrobial susceptibility patterns among bacterial isolates from culture-positive keratitis differ by geographic region, and have they changed over time from 1975 to 2024? __SNAPSHOT__ {"studyDesign":"Systematic review and meta-analysis","setting":"Global; 47 countries; led by the International Centre for Eye Health (London School of Hygiene & Tropical Medicine). Data from tertiary and community ophthalmology settings as reported in included primary studies.","sampleSize":"n = 129,388 episodes of culture-positive bacterial keratitis across 340 included studies","population":"Patients with culture-confirmed bacterial keratitis; isolates collected between 1975 and 2024 from studies spanning high-income countries, South Asia, East Asia and the Pacific, sub-Saharan Africa, and other regions. Specific patient-level inclusion criteria not further detailed in the available abstract.","intervention":"Characterisation of in vitro antimicrobial susceptibility of corneal bacterial isolates by drug-pathogen combination and by geographic region and calendar period","comparator":"Not applicable (epidemiological synthesis; regional and temporal comparisons performed within the meta-analysis)","primaryOutcome":"Antimicrobial susceptibility rates for clinically relevant drug-pathogen combinations, assessed globally and stratified by region and time period","followUp":"Not applicable (cross-sectional microbiological data from each included study; study data span 1975–2024)","fundingSource":"Not reported in the available abstract; institutional affiliation with LSHTM and Moorfields Eye Hospital NHS Foundation Trust noted"}
Key findings
This is the largest synthesis of antimicrobial susceptibility data for bacterial keratitis to date. Across the full dataset, many first-line antibiotics retained high aggregate activity: among Gram-negative isolates, ciprofloxacin susceptibility was 96.3% and gentamicin susceptibility was 95.8% overall; among Gram-positive isolates, chloramphenicol susceptibility was 90.8% and vancomycin susceptibility approached 100%. However, marked regional heterogeneity was identified. South Asia consistently demonstrated lower susceptibility for several clinically important drug-pathogen combinations: Pseudomonas aeruginosa susceptibility to ciprofloxacin was 80.2% in South Asia versus 98.3% in high-income countries and 96.8% in East Asia and the Pacific — a clinically meaningful difference of approximately 18 percentage points relative to high-income settings (calculated). Regarding temporal trends, the review found no conclusive evidence of a widespread global decline in antibiotic susceptibility over the study period. Some analyses suggested a gradual reduction in Staphylococcus species susceptibility to ciprofloxacin over time, but overlapping confidence intervals prevented firm conclusions. Specific p-values, confidence intervals for pooled estimates, and NNT/NNH values are not reported in the available abstract. No adverse event data are applicable to this epidemiological study design.
Practice implications
This review establishes that globally aggregated susceptibility data for bacterial keratitis are substantially misleading as a basis for regional empirical treatment decisions. Clinicians in South Asia, in particular, should not rely on high-income country susceptibility benchmarks when choosing empirical topical therapy for Pseudomonas keratitis, as ciprofloxacin susceptibility in that region (~80%) is notably lower than in high-income settings (~98%). Vancomycin remains highly active against Gram-positive corneal pathogens across regions and should be considered for severe or refractory cases. The study does not establish that resistance is worsening globally over time — this is a reassuring but cautious finding, as temporal analyses were constrained by heterogeneous reporting across primary studies. The findings do not apply to patients in sub-Saharan Africa, a region severely underrepresented despite carrying a substantial burden of microbial keratitis blindness (only 8 of 340 included studies originated from this region). The study was a synthesis of published in vitro susceptibility data; it does not establish a causal relationship between in vitro resistance rates and clinical treatment failure, and it cannot determine optimal empiric regimens. Importantly, the authors note that resistance patterns from hospital-acquired or bloodstream infections cannot be assumed to apply to ocular isolates. Only an abstract and institutional press release were available for this summary; full data tables, forest plots, and meta-regression coefficients were not reviewable. Practice should not uniformly change based on this evidence alone, but it strongly supports the adoption of region-specific treatment protocols and investment in local corneal microbiology surveillance. __LIMITATIONS__ Only an abstract and institutional press release were available; full methodology, risk-of-bias assessments, and complete statistical results could not be independently reviewed.,Extreme geographic underrepresentation: approximately half of included studies originated from high-income countries, while only 8 studies came from sub-Saharan Africa — a region with high keratitis-related blindness burden — severely limiting generalizability to low-income settings.,High expected between-study heterogeneity: variability in laboratory methods (disk diffusion vs. MIC-based testing), breakpoint criteria (CLSI vs. EUCAST), local antibiotic prescribing pressures, patient populations, and reporting periods across 340 studies from 47 countries likely inflates or obscures true susceptibility estimates.,Temporal trend analyses were inconclusive due to overlapping confidence intervals, limiting the ability to detect or rule out gradual resistance emergence with certainty.,The review assessed in vitro susceptibility only; corneal pharmacokinetics of topical antibiotics, which can produce drug concentrations far exceeding systemic breakpoints, were not accounted for — meaning standard MIC-based susceptibility classifications may not reflect clinical treatment response.,Publication bias is likely: centers with unusual resistance findings or outbreaks are more likely to publish microbiological data, potentially overestimating resistance rates.,Isolates reflect culture-positive cases only; culture-negative or empirically treated cases are excluded, introducing selection bias toward more severe or atypical presentations.,Funding source not reported, precluding assessment of industry sponsorship bias. __BOTTOM_LINE_CATEGORY__ potentially_practice_informing __BOTTOM_LINE__ This landmark meta-analysis of 129,388 bacterial keratitis episodes across 47 countries confirms important regional differences in antimicrobial susceptibility — most critically, a ~18 percentage point lower ciprofloxacin susceptibility for Pseudomonas keratitis in South Asia compared with high-income countries — while finding no conclusive evidence of global temporal resistance worsening. Clinicians should not extrapolate global or high-income susceptibility benchmarks to South Asian or other low-resource settings, and local surveillance data should guide empirical topical antibiotic selection. A standardised global reporting framework for corneal microbiology data is urgently needed to reduce heterogeneity and enable future actionable surveillance.