Rapid semiautomated screening and processing of urine specimens

A rapid urine culture procedure was evaluated in which positive urines were detected by using light-scatter photometry (Autobac). Specimens were analyzed at 3, 5, and 6 h. Specimens detected as positive at 3 h were then further evaluated by a direct 3-h susceptibility procedure (Autobac) and by a 4-h identification procedure (Micro-ID). Of 949 specimens, 175 had >10(5) colony-forming units per ml by colony count. Of these latter specimens, 75.4% had been detected by 3 h, and 95.4% were detected by 6 h. Of specimens positive by Autobac at 3 h, 96% (95.7%) had >10(5) colony-forming units per ml. If pure by Gram stain, those positive specimens were inoculated to direct susceptibility and identification systems. When direct Autobac susceptibilities were compared with the standard Autobac method done from the plate the following day, discrepancy rates were 1.3% very major, 2.1% major, and 7.4% total. The direct identifications were 94% (94.2%) correct when using the Micro-ID manual and a collection of octal patterns unique to this system, in which urine/broth culture inoculum was employed instead of the usual organism colony suspension. Those urine specimens negative after screening at 3 h were evaluated at 5 and 6 h, and an additional 126 specimens were detected as positive. These were then processed by routine plate inoculation, due to the limitations of the work day. By 6 h, 95.4% of specimens with >10(5) colony-forming units per ml were detected. The 4.6% false-negative results consisted of patients on antibiotics, or slowly growing bacteria suspected of being distal urethral contaminants. Thus, 83.5% of the urine cultures received by 9:00 a.m. (10.6% 3-h positives and 72.9% negative at 6 h) could be evaluated and reported within one 8-h work day.