Rapid direct antibiotic susceptibility testing in endophthalmitis.
OBJECTIVE: To compare a new rapid antibiotic susceptibility test (RAST) to the conventional method in patients with endophthalmitis. DESIGN: Prospective nonrandomized comparative trial. PARTICIPANTS: Intraocular aspirates from 24 consecutive patients with endophthalmitis were tested. METHODS: Approximately 0.25 ml of vitreous or anterior chamber aspirate was obtained and tested for antibiotic sensitivity using the Kirby-Bauer agar disk diffusion method. Using this conventional testing method, the aspirates were cultured for bacterial growth before antibiotic susceptibility testing (AST). In contrast, for direct RAST, the aspirates were inoculated directly onto agar plates with antibiotic disks without first culturing for bacterial growth. RESULTS: Of the 24 endophthalmitis aspirates studied, 4 (17%) could not be tested by direct RAST because either more than one bacterial species was present or no bacteria were cultured. The remaining 20 cases provided 467 antibiotic-microorganism combinations. The two methods yielded identical susceptibility results in 409 (88%) of the 467 tests. In 42 tests (9%), a minor discrepancy occurred between the conventional method and RAST, in which one resulted in intermediate susceptibility, whereas the other was either resistant or sensitive. Significant discrepancy, defined as a bacterial sample reported as sensitive in one method yet resistant in the other, occurred with 16 samples (3%). Results of the RAST were available within 6 to 10 hours compared with more than 24 hours for the conventional method. Of clinical significance, the results of the RAST revealed that in 11 cases of epidemic Stenotrophomonas maltophilia endophthalmitis, the bacteria were resistant to the antibiotics prescribed, and change of antibiotic therapy was made on the basis of the RAST results. CONCLUSIONS: RAST provided accurate antibiotic susceptibility results in a much shorter time than did the conventional method. We suggest that conventional AST be performed for confirmation of the RAST results, but initial antibiotic therapy can quite reliably be based on the results of RAST.