Patients were recruited from urological clinics and controls from staff or volunteers. Female patients who described OAB symptoms according to International Continence Society (ICS) criteria , of urinary urgency with or without urge urinary incontinence (UUI) were included. Healthy female adults matched for age and menopausal status and with no urinary symptoms formed the control group. All participants provided written informed consent and completed the International Consultation on Incontinence Questionnaire (ICIQ) LUTS for urgency and pain. Women who were pregnant or planning pregnancy were not eligible for inclusion. Participants with other urinary tract disease, diabetes, immune disease or taking diuretics or other drugs influencing the urinary tract, which may have compromised data validity, were excluded.
Symptoms were recorded using three validated questionnaires: The ICIQ was selected to evaluate symptoms . The Whittington Urgency Score, a ten-item scale, was used to measure symptoms and degree of urinary urgency; the questionnaire has been validated [17, 18]. The Whittington Pain Questionnaire, a validated, eight-item scale, was used to record the most prevalent dysaesthetic/pain symptoms associated with the lower urinary tract .
Study visits and processes
Written informed consent was obtained at the first visit, prior to any study-related procedures, and eligibility was checked. Participants attended 12 study visits in total, scheduled every 4 weeks. During this time, patients were treated with antimuscarinics agents and antibiotics if pyuria implied infection.
MSU sample collection
Participants provided a midstream clean-catch urine sample. Patients were given verbal and written instructions on avoiding contamination . The urine was decanted into three 30-ml sterile universal, anonymised specimen tubes, blinding the researcher.
Inflammation and immune response: microscopy for pyuria and urothelial cell shedding
Immediate microscopy was performed on fresh, unspun, unstained urine samples. A disposable pipette was used to place a drop of urine in the filling chamber of a Neubauer haemocytometer and covered with a glass coverslip. Olympus CX41 light microscope (×200) (Olympus, Southend-on-Sea, UK) was used to analyse the sample. Leucocyte and epithelial cell count was enumerated using a standard operating procedure in triplicate. All three measures were recorded and mean value calculated.
Bacterial colonisation: urothelial clue-cell analysis
Urine samples were processed within 1 h of collection and refrigerated at 4 °C until assessment. A collection chamber consisted of a single-channel cuvette and retainer, a Shandon filter card (Fisher Scientific, Loughborough, UK) and a Superfrost Ultra Plus glass microscope slide; 80 μl of urine was transferred into the collection chamber for centrifugation and spun at 75 g for 5 min. Cellular components formed a visible deposit on the slide. Cells were then fixed with 4% formaldehyde (Thermo Scientific, Fisher Scientific) at room temperature for 15 min. Cell membranes were stained with wheat germ agglutinin (WGA) conjugated to Alexa Fluor 488 (Invitrogen). The cellular deposit was incubated for 15 min at room temperature. Alexa Fluor 488 excites at a wavelength of 495 nm and emits at 519 nm; hence, the cell membranes appeared green under fluorescent microscopy. The host and bacterial DNA were stained using the DNA stain 4′-6′-diamidino-2-phenylindole (DAPI) and cells immediately mounted with FluorSave reagent (Calbiochem). A coverslip was carefully applied ensuring no air bubbles, and the coverslip was fixed with nitrocellulose with ethyl acetate. DAPI gives mammalian nuclei and bacteria a blue appearance under fluorescent microscopy; it excites at a wavelength of 360 nm and emits at 460 nm. DAPI is able to label intracellular and extracellularly attached bacteria without the need for permeabilisation. Slides were examined under a fluorescent Olympus CX41 upright epi-fluorescence microscope. The proportion of clue cells—urothelial cells exhibiting adherent or intracellular microbes—were calculated by counting the total number of cells present and then the proportion of cells with associated bacteria. Counts were performed in triplicate, and an average clue-cell proportion was recorded.
Microbiological assessment: enhanced sediment culture
Enhanced sediment cultures were processed within 2 h of sample collection. Samples were refrigerated at 4 °C until processed. Five millilitres of fresh unspun, unstained, urine was centrifuged at 627 g for 5 min in a Denley BR401 centrifuge (RMAX 140 mm) (Denley, Heckmondwike, UK). The supernatant was removed, leaving the urinary sediment, which was resuspended in 400 μl of 1% sterile phosphate-buffered saline (PBS) solution. Four 1:10 serial dilutions were performed for accurate quantitative bacterial counting. Chromogenic CPS3 agar plates (bioMérieux, Basingstoke, UK) were used for culture. All culture plates were incubated aerobically for 24 h at 37 °C in a CO2-dependant incubator.
Each bacterial isolate was quantified. No threshold was used to discriminate positive or negative growth. The mean colony count from all sectors was calculated. The CPS3 chromogenic medium allows bacterial identification of uropathogens to genus or species level, dependent on the microbe. The growth of distinct bacteria is colour-specific to allow easy and fast enumeration. Colour identification was based on the manufacturer’s standardised colour guide and supplemented with Gram staining and rapid biochemical tests for further characterisation. Analytical Profile Index (API) testing was used when these methods were unable to identify an isolate conclusively.
Microbiological assessment: routine culture
All routine microbiological cultures were undertaken in the hospital microbiology laboratory. Thirty millilitres of urine in a sterile universal specimen tube was cultured immediately upon receipt or after overnight refrigeration at 4 °C. Trained blinded biomedical scientists undertook all analyses. One microlitre of the urine sample was inoculated onto a CPS3 agar plate using a sterile 1-μl loop. The culture plate was then incubated aerobically for 24 h at 37 °C. Bacterial colonies were identified by colour and morphologic characteristics. Rapid reagent testing (spot testing) was employed to supplement colour-based bacterial identification. Bacterial growth was estimated by visual assessment of colony density. A positive culture was defined as the growth of a single recognised uropathogen at ≥105 CFU ml–1. Polymicrobial growth above this threshold was reported as mixed growth. Any bacterial growth <105 CFU ml–1 was reported as no significant growth.
Primary and secondary outcome measure
The primary outcome measure was total log10 bacterial CFU of all isolates obtained from culture of the urinary sediment. The secondary outcome measures were:
Microscopic pyuria count
Urothelial cell count
Urothelial cells demonstrating associated bacteria (clue cells)
Routine urine culture in hospital laboratory
ICIQ-LUTS symptoms score
Whittington Urgency Score
Whittington Pain Score
A sample size of 20 participants in each group provided 83% power to detect a significant difference in log10 bacterial growth, with alpha of 0.05. This was calculated from pilot data in which the log10 bacterial count standard deviation (SD) was 2 and mean difference 0.5. The primary analysis was to determine the difference in total log10 CFU per ml−1 of bacterial growth between patients and controls. The independent variable was group (patient 1, control 0), which was entered as a fixed effect. The dependant variable was total bacterial growth on sediment culture (log10 CFU ml−1). Measure repetition was identified by visit number.
Secondary analyses were explored the relationship between bacterial growth, LUTS, pyuria and urinary urothelial cell shedding. With the independent variable being group (patients/controls), dependent variables were selected, in turn, as LUTS, urgency and pain scores, log10 pyuria and log10 epithelial cell count (μl−1). Measure repetition was identified by visit number.
For additional analysis, data were pooled into two sets, patients and controls, and pooled to compare the performance of routine culture methods against results of urinary sediment culture. Analyses of monthly data were achieved through the repeated-measures procedure of the generalised linear model (GLM) provided by SPSS. The nonparametric Mann–Whitney test was used to examine pooled data for differences in outcomes between patients and controls. Multinomial logistic regression was used to examine differences in microbial species dispersion between groups.