Pragmatic evaluation of a midstream urine collection technique for infants in the emergency department.

Abstract

Our objective was to examine the performance characteristics of a bladder stimulation technique for urine collection among infants presenting to the emergency department (ED). This prospective cohort study enrolled a convenience sample of infants aged ≤ 90 days requiring urine testing in the ED. Infants were excluded if critically ill, moderately to severely dehydrated, or having significant feeding issues. Bladder stimulation consisted of finger tapping on the lower abdomen with or without lower back massage while holding the child upright. The primary outcome was successful midstream urine collection within 5 minutes of stimulation. Secondary outcomes included sample contamination, bladder stimulation time for successful urine collection, and perceived patient distress on a 100-mm visual analog scale (VAS). We enrolled 151 infants and included 147 in the analysis. Median age was 53 days (interquartile range [IQR] 27-68 days). Midstream urine sample collection using bladder stimulation was successful in 78 infants (53.1%; 95% confidence interval [CI] 45-60.9). Thirty-nine samples (50%) were contaminated. Most contaminated samples (n = 31; 79.5%) were reported as "no significant growth" or "growth of 3 or more organisms". Median bladder stimulation time required for midstream urine collection was 45 seconds (IQR 20-120 seconds). Mean VAS for infant distress was 22 mm (standard deviation 23 mm). The success rate of this bladder stimulation technique was lower than previously reported. The contamination rate was high, however most contaminated specimens were easily identified and had no clinical impact.