[Skip to Content]
Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address Please contact the publisher to request reinstatement.
[Skip to Content Landing]
Citations 0
September 11, 2002

Clinical Diagnosis of Urinary Tract Infection

Author Affiliations

Stephen J.LurieMD, PhD, Senior EditorIndividualAuthor

JAMA. 2002;288(10):1229-1231. doi:10.1001/jama.288.10.1229

To the Editor: In their Rational Clinical Examination article, Dr Bent and colleagues1 evaluated the accuracy and precision of the clinical history and examination for the diagnosis of urinary tract infection (UTI). Clinicians often rely on a single symptom or sign to establish the diagnosis,2 so it is important that they be aware of the pretest probability. However, the pretest probability of infection that Bent et al cite (48%) seems very high to us. One of us (T.F.) has calculated a 25% to 29% pretest probability of UTI in patients attending 8 primary care practices in the United Kingdom.2 We are also aware of another study that found a pretest probability of 35%.3 We therefore reexamined the studies that Bent et al used, added the study by Dobbs and Fleming,3 and performed a meta-analysis on only those articles that exclusively studied women with symptoms suggestive of UTI.46 We excluded studies if they included men2,7,8 or women with suspected vaginal infection.9 There was no heterogeneity (P = .12), so a fixed-effects model was used. This analysis yielded a combined pretest probability estimate of 35.6% (95% confidence interval, 32.1%-39.0%).

First Page Preview View Large
First page PDF preview
First page PDF preview