Methods and Material
(For online publication only)
Pre-study considerations
1. Two women with E.coli cystitis delivered urine samples two days before starting therapy with antibiotics, and each day thereafter for three days. The women divided the morning urine in two aliquots. One of the aliquots was fixed in Carnoy solution, the other left native. Both aliquots were delivered the same day. Depending on the distance to the hospital and differences in their daily schedules, the time between voiding the urine and delivery to the laboratory could be between 2 and 6 hours. In the laboratory, the native left urine was then fixed with Carnoy solution. The comparison of 7 sample pairs demonstrated a marked deviation in bacterial concentrations of immediately and delayed fixed urine. In 5 samples that were fixed after delivery to the laboratory, the E.coli concentrations were 1 to 3 orders (mean Log 1.8) higher than in the immediately fixed samples. Only in 2 samples collected during antibiotic therapy (day 2 and 3) with low bacterial concentrations of less than 104, the difference between immediate and delayed fixed samples was negligible. Two of the delayed fixed urine samples (after 6 hours) contained fungi, while no fungi were detected in immediately fixed samples.
Obviously urine samples delivered native to the laboratory and fixed there are biased both in composition (fungi) and concentrations of the microorganisms.
2. The Carnoy fixed urine samples from 2 women with E.coli bacteriuria were investigated initially and 1,14, 28 days and 6 months later using FISH. The bacterial concentrations stayed unchanged within this time period.
Based on these data, we decided to use only immediately fixed samples of the morning urine and designed the number of samples and the collection frequency mainly on the expedience and convenience of the women and the personnel in the clinic and laboratory.
The cultural detection of bacterial concentrations is biased by vaginal flora. In mixed culture of vaginal and urine microbiota “irrelevant” bacteria can overgrow and suppress the pathogenetic important strains. For cultural diagnostics the middle portion of urine is therefore used in the hope, that contamination by vaginal microbiota will be reduced, while concentrations of uro-pathogenic bacteria will remain unchanged. Although reasonable, the urine portioning is not backed by any quantitative data, which would for example allow to predict time dependent wash-out effects of urine flow on adherent strains. Furthermore in pollakisurie, which is the most prominent clinical sign of cystitis, the collection of a middle portion is often not achievable, especially when urine samples are collected daily or three times a week over the period of cystic complaints.
Biases through growth interference play no role in FISH studies. We therefore asked women to use the first portion of urine, which is most convenient to collect and reproducible at home.