An observational diagnostic accuracy study comparing the urine dipstick with a consensus-based reference standard for...

The urine dipstick is widely used to screen for urinary tract infection (UTI) in daily practice. However, in older adults, especially women, its reliability is limited by the high prevalence of asymptomatic bacteriuria and leukocyturia. Recently, a consensus-based reference standard was introduced to classify patients based on the likelihood of UTI. Using this standard, we assessed the diagnostic accuracy of the urine dipstick in older adults.

We performed an observational study using prospectively collected data from two ongoing studies. We included adults aged ≥ 60 years with clinically suspected UTI who underwent dipstick testing, urine flow cytometry, and urine culture during the same episode. The Roche COMBUR-7 urine dipstick served as the index test, and episodes were classified using the consensus-based reference standard. We evaluated all possible dipstick thresholds and their combinations, calculating sensitivity, specificity, predictive values, and likelihood ratios using Generalized Estimating Equations to account for multiple episodes per participant. Receiver operating characteristic analysis was performed to determine the optimal balance between sensitivity and specificity.

We included 368 episodes from 199 participants (mean age 73.8 years; 77.9% women). Dipstick sensitivity was 93.6% (95% CI 90.1%-95.9%) and specificity was 58.6% (95% CI 45.4%-70.6%) for the conventional threshold of nitrite + and/or leukocyte esterase ≥ 1 + . A higher threshold, "total of ≥ 2 + " (leukocyte esterase ≥ 2 + or nitrite positive with leukocyte esterase ≥ 1 +) improved performance, yielding a sensitivity of 82.2% (95% CI 77.0%-86.4%) and a specificity of 75.7% (95% CI 61.6%-85.8%). With this threshold, the positive likelihood ratio improved from 2.3 (95% CI 1.76-3.06) to 3.4 (95% CI 2.0-6.1).

In older adults with suspected UTI, urine dipstick testing demonstrates limited diagnostic performance. Current thresholds yield high sensitivity but poor specificity, which may contribute to unnecessary antimicrobial treatment. Given the high prevalence of asymptomatic bacteriuria in older adults, a higher threshold of "total of ≥ 2 + " (either LE ≥ 2 + , or nitrite + combined with LE ≥ 1 +) may improve diagnostic accuracy and be considered when assessing patients ≥ 60 years old for antibiotic therapy. Prospective studies are needed to confirm this recommendation.