Clinical Question: Do antibiotics decrease symptoms in women with dysuria and frequency who have a negative dipstick test result for nitrites and leukocytes?

Setting: Outpatient (primary care)

Study Design: Randomized controlled trial (double-blinded)

Allocation: Concealed

Synopsis: The authors invited women 16 to 50 years of age to participate in this study if they presented to their New Zealand general practitioner with a history of dysuria and frequency but with a midstream urine specimen that was negative for nitrites and leukocytes using standard urine dipstick analysis. As a check on the validity of the dipstick, urine specimens also were cultured, although the results were not known until after the treatment and assessment had been completed. The 59 participants were randomized to receive placebo or trimethoprim (Proloprim) in a dosage of 300 mg daily for three days.

At the end of treatment, 76 percent of the women who received trimethoprim had resolution of dysuria compared with 26 percent of women who received placebo (P = .0005). By seven days, 90 percent of the treated women had resolution of dysuria compared with 59 percent of women receiving placebo (P = .02). One additional patient had resolution of symptoms by seven days for every four women who received trimethoprim instead of placebo (number needed to treat = 4; 95% confidence interval, 1.9 to 14.1). Urinary frequency was unaffected by treatment. It was not that the dipstick test failed to diagnose infection: culture of dipstick-negative urine grew organisms in only five of 59 women; therefore, the negative predictive value of the dipstick analysis was 92 percent.

Bottom Line: In women with dysuria and frequency who have a negative result on urine dipstick analysis for nitrites and leukocytes, three of every four will respond to antibiotic treatment compared with one in four who are taking placebo. In this study, the negative dipstick result correlated with culture 92 percent of the time. These results imply that some women have microbial infections that are not identified by dipstick or culture, or perhaps that the antibiotic is doing something other than killing bacteria. (Level of Evidence: 1b)