Failure of endophthalmitis prophylaxis: laboratory model comparing antibiotics and bacterial loads.

PURPOSE: To assess by an in vitro model the influence of antibiotic agent and bacterial load on the failure of intracameral endophthalmitis prophylaxis.

SETTING: Microbiology laboratory.

DESIGN: Bacterial growth comparison.

METHODS: Bacteria were incubated with moxifloxacin, cefuroxime, ceftazidime, or imipenem. Antibiotic concentrations were approximately twice those achieved by standard clinical intracameral doses. Bacterial loads were either low or high (0.0025 or 0.25 McFarland). The tested strains included 4 Enterococcus faecium , 7 Klebsiella pneumoniae , 8 methicillin-resistant Staphylococcus aureus , 3 Streptococcus pneumoniae , and 6 Pseudomonas aeruginosa . Samples were diluted by half every 2 hours to model the half-life of intracameral antibiotics. Bacterial broths were plated at 24 hours, and colonies were counted at 48 hours. Samples with >250 colony forming units /mL were deemed failures.

RESULTS: Antibiotics that successfully eliminated a pathogen often failed when the bacterial load was 100 times greater. Moxifloxacin had no failures against low-load pathogens. Moxifloxacin and ceftazidime both had significantly fewer failures than cefuroxime ( P < .01; P > .02). Ceftazidime was far more effective than cefuroxime against gram negatives, but less effective against low-load enterococci.

CONCLUSIONS: High bacterial loads can multiply uncontrollably in the presence of antibiotics which are effective when the bacterial load is lower; this underscores the importance of uncontaminated equipment and ocular surface antisepsis. Broadest coverage against common postoperative endophthalmitis pathogens was demonstrated by moxifloxacin and ceftazidime.