RESUMO
OBJECTIVE: To evaluate a novel urine-collection device (UCD) that automatically collects a midstream urine (MSU) sample, and compare contamination rates to those of the conventional MSU sampling method, as the contamination of urine samples for microbiological analysis in women leads to diagnostic ambiguity and unnecessary costs, and may result in part from an incorrect collection procedure. PATIENTS AND METHODS: In all, 2823 women from four centres, most from antenatal clinics, were randomized to two urine-collection methods: conventional MSU collection and collection with a novel MSU UCD (the Whiz, JBOL Ltd, Oxford, UK). Semi-quantitative growth and user acceptability were compared between the collection methods. RESULTS: MSU samples collected with the UCD had significantly fewer mixed growth samples (9% vs 14%, P = 0.001; 36% relative reduction), significantly fewer heavy mixed growth samples (1.2% vs 3.0%, P = 0.004; 60% relative reduction) and required significantly fewer re-tests (11% vs 16%, P = 0.002; 31% relative reduction). There were more samples with clinically insignificant growth than the conventional MSU group (86% vs 82%, P = 0.005). Those using the UCD preferred it to the conventional method (67.5%) and experienced significantly less spillage during sample collection (27% vs 46%, P = 0.001; relative reduction 41%). CONCLUSION: The UCD reduced contamination rates in urine samples and improved the predictive value of the urine culture in a manner acceptable to patients and staff.