Point-of-care tests for pediatric urinary tract infections in general practice: a diagnostic accuracy study

Abstract

Abstract Background Early diagnosis of pediatrics urinary tract infections in the outpatient settings is challenging but essential to prevent hospitalization and kidney damage. Objective We aimed to evaluate the diagnostic test accuracy of a selection of point-of-care tests for pediatric urinary tract infections in general practice. Methods A prospective cross-sectional study in 26 general practices in Flanders, Belgium (clinicaltrials.gov, NCT03835104). Urine was sampled systematically from children between 3 months to 18 years presenting with an acute illness of maximum 10 days. Samples were analyzed at the central laboratory with a routine dipstick test, the Utriplex test, the Uriscreen test and the Rapidbac as index tests, and with urine culture showing more than 105 colony-forming units per milliliter of one pathogen as reference standard. For each test, we calculated sensitivity, specificity, positive and negative likelihood ratios, and predictive values with 95% confidence intervals. Results Three-hundred urine samples were available for analysis of which 30 samples were culture positive (10%). Sensitivities and specificities were 32% (95% CI 16%–52%) and 86% (95% CI 82%–90%) for the dipstick test, 21% (95% CI 8%–40%) and 94% (95% CI 91%–97%) for the Utriplex test, 40% (95% CI 16%–68%) and 83% (95% CI 75%–88%) for the Rapidbac test, and 67% (95% CI 38%–88%) with 69% (95% CI 60%–76%) for the Uriscreen test. Conclusion All 4 point-of-care tests were suboptimal for use in the broad range of children presenting with acute illnesses to general practice. General practitioners need novel methods for obtaining reliable urine samples during the time of the consultation, especially for children not yet toilet-trained.