Objective: To assess the reliability for biochemical and microbiological analysis of urine collection from disposable diapers in elderly women with severe urinary incontinence.
Design: Cross-sectional comparison of two methods of collection.
Patients and measurements: Urine was sampled from 52 women inpatients (aged 68-98 years) in a geriatric hospital ward by pressing a diaper which the patient had worn for 3 hours. Just after this collection, another sample was obtained by retrograde catheterization. Both samples were analyzed for sodium, potassium, chloride, proteins, urea, creatinine, calcium, and phosphate and for cell counts, the presence of bacteria, and bacteria culture.
Results: For all the biochemical parameters, the urinary concentrations obtained by the two methods were strongly and significantly correlated. The following differences (mean and range, in mmol/L) were found between the values obtained by the two methods: sodium: 6.05 (-26 to 27), potassium: (-16 to 14), chloride: -1.13 (-24 to 23), urea: 6.85 (-33 to 37), creatinine: 0.24 (-0.95 to 1.45), calcium: -0.22 (-1.27 to 1.70), and phosphate: 2.17 (-2.5 to 13.4). For diagnosis of urinary tract infection, agreement between the two methods was good (kappa = 0.84), and bacteriological agreement was obtained in 25 out of 28 cases (89%). However, for diagnosis of microscopic hematuria, agreement was poor (kappa = 0.50), probably due to the overestimation of the true urinary red cell count in the samples collected by catheterization.
Conclusion: For routine microbiological and biochemical analysis, urine extraction from disposable diapers is a simple and reasonably reliable method of sampling urine from elderly women with severe incontinence.