[Screening for bacteriuria in diabetic patients. Is it possible to stop systematic urine cytobacteriological testing?].
OBJECTIVES AND METHOD: The discordance between test by urine dipstick (nitrites and leucocyte-esterase) and analysis in laboratory, with urinary culture on the same sample was studied in diabetic patients, from October 2000 to May 2002, to eventually stop systematic laboratory test. The dipstick result (Clinitek 20 Bayer) was classified as "possibility of bacteriuria" if one of the two tests was positive. Bacteriuria was considered significant if the laboratory test result gave, at least 10(5) bacteria per mL, (one strain), and at least 10(4) leucocytes. The out point was the dipstick negative predictive value (NPV). RESULTS: The study included 683 patients. The dipstick result was "possibility of bacteriuria" in 153 cases (22.4%). Thirty-nine bacteriuria (5.7%) were reported, including 2 dipstick false negatives. The NPV was 99.6% [IC 95% : 99.1-100]. CONCLUSION: The systematic laboratory tests were stopped.