Definition of urinary tract infection and assessment of efficacy in drug trials - A laboratory perspective

1. There appears to be no difference between study groups of patients with symptomatic UTI with regard to bacteriuria and nitrite positivity. However, a full prospective study should confirm this. 2. Most methods currently used are not sufficiently sensitive to detect low numbers of bacteria and to distinguish between contamination and infecting organisms. 3. Screening methods are easy and cheap to perform and select a similar population to those included on the criteria of low counts. It would seem to be sensible to continue with present criteria and to perform studies to determine the practicalities of using low coliform counts. All treatment failures must be carefully evaluated.