Abstract
Four hundred and twenty-two urine samples were screened for significant bacteriuria using bioluminescence and microscopy of uncentrifuged urine. A smaller number of false-negatives were seen with bioluminescence (10%) than with microscopy (40%) while both techniques gave a similar number of false-positives (18%). The kit required a large amount of manual preparation, largely pipetting. With this and the short shelf-life of the reconstituted reagents, it is not suitable for small numbers of urines. At 45p per urine, the cost of bioluminescence is too high.