Use of an antimicrobial removal device in endophthalmitis cultures.

Abstract

This study sought to determine whether the use of an antimicrobial removal device (ARD) to process intraocular fluids increases microbial detection compared with conventional cultures. The authors retrospectively reviewed all cases of endophthalmitis submitted to their laboratory from January 1982 through December 1996. Aqueous or vitreous specimens or both that were cultured on conventional media (blood agar, chocolate agar, anaerobic blood agar, and thiol broth) and by ARD processing were included in the study. Specimens were inoculated into tubes with ARD for 5 to 10 minutes. The fluid was then withdrawn and cultured using conventional media; thioglycolate broth was added to the tube containing the resin beads. The conventional and ARD-processed cultures were incubated at 35 degrees C for at least 7 days. Of the 338 endophthalmitis cultures processed using both conventional cultures and a parallel ARD, 166 (49.1%) yielded positive microbial growth on one or more media. Of the 166 culture-confirmed cases, 127 (76.5%) were positive in both the ARD-processed and direct cultures, 17 (10.2%) were positive by conventional culture only, and 22 (13.3%) were positive by the ARD-processed sample alone (P = 0.52). The spectrum of microorganisms was similar among all culture groups. The detection of coagulase-negative staphylococci and micrococci by ARD alone was slightly better than detection by conventional culture only (P = 0.06). Of 93 positive cultures from 179 patients in whom prior antibiotic use was documented, 75 (80.6%) were positive by both methods, 8 (8.6%) by conventional cultures only, and 10 (10.8%) only by the ARD-processed specimen (P = 0.81). Use of an antimicrobial removal device does not significantly increase the microbial yield of endophthalmitis cultures compared with conventional culture techniques, whether or not antimicrobial therapy is being used.