Detection of bacteriuria by bioluminescence: effect of pre-analysis centrifugation of specimens.

Three bioluminescence-based, rapid methods of detecting significant bacteriuria were applied in parallel to 514 urine specimens. The results were compared with those of a quantitative pour plate viable count method, defined as positive if greater than or equal to 10(5) c.f.u./ml of urine were observed. When adjusted to yield 21% falsely positive results the three rapid methods yielded 24%, 21% and 19% falsely negative results. If specimens with evidence of urethral or vaginal contamination were excluded (237 specimens remaining) the three methods yielded respectively 14%, 8% and 13% falsely negative results. A major source of disagreement between the bioluminescence-based methods and quantitative culture thus appeared to be contaminated urine specimens.