This study was conducted over an 8-week period in a phase-2, parallel double-blinded, randomised controlled design. Adults over the age of 18 years with symptoms of UI and/or OAB were recruited via a variety of advertising media including newspapers advertisements and notices posted at community centres. Self-identified participants were initially screened for suitability via telephone by research clinicians, based on definitions outlined by the Standardization Committee of the International Continence Society.
Ethics, consent and permissions
The study was approved by the Ethics Committee of Endeavour College of Natural Health (Queensland, Australia; approval number HREC #12/030). All participants provided written informed consent.
Inclusion criteria, based on an adult only population, included those who experienced in the most recent six months, symptoms such as: urinary day frequency (≥10/day), nocturia (≥2/night), urgency (≥2/day), and incontinence (≥1/day). To be eligible, participants needed to have a minimum of 2 of these symptoms. Urodynamics were not performed, patients were recruited solely on the basis of their symptoms, as the former is invasive and provides only a brief snapshot of bladder function under artificial conditions [41]. Participants with comorbidities such as controlled hypertension, osteoarthritis, controlled diabetes, anxiety, chronic obstructive pulmonary disease, etc., were included in the study. These diseases/disorders were not expected to confound the results.
Exclusion criteria included: recent (≤1 year) relevant surgeries such as hysterectomy, prolapse repair, prostate surgery, childbirth/currently pregnancy; current use of any natural therapies for bladder symptoms or prescribed medication for UI or OAB; unregulated doses of diuretics; undergoing treatment for mental health issues or psychiatric disturbances; other concomitant health conditions, including uncontrolled diabetes mellitus, heart disease, pancreatic, hepatic or renal disease, neurologic disease, recurrent urinary tract infections, benign prostatic hypertrophy, continual leakage, menstrual cycle-related incontinence, and chronic inflammatory conditions.
Randomisation
Participants meeting the above criteria, provided written informed consent and were randomised via the block of four method (using Microsoft Excel® command “Rand”) by a third party, into either treatment or placebo as indicated by either blue or yellow stickers on identical product bottles and allocated patient files. Both participants and researchers remained blinded to treatment allocation until after completion of statistical analyses, to ensure no risk of bias for the entire duration of the study and into completion. Treatment allocation was provided by a 3rd party external to the researchers and clinicians and was not disclosed until after statistical analysis, also minimising the chance of bias.
Interventions
Each capsule contained 420 mg of a proprietary blend of Urox® (Seipel Group, Brisbane, Australia) containing Cratevox™ (Crataeva nurvala L.; Capparidaceae; Varuna) stem bark extract standardised for 1.5% lupeol; non-standardised Equisetum arvense L. (Equisetaceae; horsetail) stem extract; and, non-standardised Lindera aggregata Sims. (Lauraceae; Japanese evergreen spicebush) root extract. The placebo contained a colour-matched vegetarian capsule containing colour-matched cellulose. Identity of each plant was confirmed prior to manufacture of the finished capsule via high performance thin layer chromatography (HPLC) for Crataeva and Equisetum (Southern Cross University, Lismore, New South Wales, Australia) and Lindera (Alkemist Labs, Costa Mesa, California, USA). Crataeva and Equisetum were wild-crafted and grown without the use of pesticides. Lindera was cultivated and was tested for pesticide residue by gas chromatography mass spectroscopy (GC-MS).
Capsules were manufactured in a Therapeutic Goods Administration licensed facility according to the PIC/S Guide to Good Manufacturing Practice for Medicinal Products, PE 009–9-15 January 2009. Capsules underwent microbiological and heavy metal testing to ensure they complied with product specifications. The dosage was two capsules per day taken once daily with food. The dosage regimen was determined based on earlier research with Crataeva and Equisetum alone [29], pharmacopeia and traditional herbal textbook dosage recommendations, and earlier, unpublished research with the Crataeva, Equisetum and Lindera blend [25,26,27,28].
Initial consultations were held at two Brisbane outpatient centres. Follow-up interviews were conducted via telephone. Participants were telephoned approximately one week before the scheduled interview and were intermittently called during the week thereafter if they were unable to be reached to schedule an alternative interview time. The clinicians conducting interviews were herbalists with additional qualifications in naturopathic medicine, acupuncture and/or traditional Chinese medicine. In Australia, herbalists are regulated either by the National Herbal Association of Australia (NHAA) or, if the herbalist has multiple qualifications which is often the case, they are regulated by other professional association bodies (e.g., the Australian Acupuncture and Chinese Medicine Association, the Australian Naturopathic Practitioners Association, etc.).
The week prior to initial consultation participants were requested to complete a micturition diary and relevant health related quality of life surveys. The attending clinician completed a clinical data sheet, containing a range of questions including demographics, exercise, health history and personal habits. Any incomplete quality of life surveys at outset were completed during the initial interview. Micturition diaries were collected by post at 2, 4 and 8 week intervals, along with the completion of quality of life surveys and follow up clinical data sheet via telephone. Participants were asked to keep the micturition diary for 3 days prior to each consultation and were provided reply paid envelopes to return the surveys and any unused capsules to assess compliance.
The primary outcome measure was urinary frequency defined as the number of voluntary diurnal and/ or nocturnal micturitions, self-reported via a validated urinary diary method per Wyman [42]. These diaries are designed to assist with the collation of urinary symptoms for 3 consecutive days, where participants indicate the number of times each symptom occurs. Relevant symptoms include urinary frequency, urgency, urgency incontinence or stress incontinence. Long-term diaries decrease patient compliance while a 3-day diary is reported as reliable to that of a 7-day record [43]. Secondary outcome measures included number of urinary urgency episodes/day, and number of incontinence episodes/day, using the same diary technique as for the primary measure.
Health-related quality of life (HR-QOL) was measured using condition specific, previously validated standardised instruments outlined elsewhere [44,45,46,47]. These surveys included the short versions of the Overactive Bladder Questionnaire (OAB-SF; maximum score for each question = 6), which was administered if symptoms of urgency and/or frequency existed, while the Urinary Distress Inventory (UDI; maximum score for each question = 4) and Incontinence Impact Questionnaire (IIQ; maximum score for each question = 4) were utilised in cases of urinary incontinence.
Information regarding number and type of incontinence pad/diaper usage per week was collected from all participants at each time point. At the completion of the trial participants were asked if they found any benefit from the treatment and whether they would be willing to continue treatment with this medication.
Analysed outcome data was collected via patients’ self-evaluation and returned by patients via mail in order to minimise clinicians and researchers chance to elicit any hypothetical bias and influence outcomes. Once received, diary responses were filed until data entry and subsequent analyses. Participants were not patients of the interviewing clinicians and their postal outcome results were not considered or evaluated for the duration of the trial. These follow up interviews consisted of quality of life questions as well as questions to ascertain any changes in participants’ lifestyle, which might otherwise confound results such as changes in exercise, water intake, diet or extraneous stress, or newly presenting medical conditions and medication changes.
Sample size calculation and statistical analysis
For the primary outcome (day frequency), based on earlier research using a Crataeva and Equisetum combination [29], it was calculated that 45 participants in each of the 2 groups were required to detect a difference of 1.6 (±2.7) urinary frequency episodes per day between treatment and placebo groups, with a two-tailed alpha of 0.05 and a power of 80%. For total incontinence, 53 participants were required to detect a difference of 1.2 (±2.2) incidents/day, while for urgency 54 participants were required in each group to detect a difference of 2 (±3.7) incidents/day. To account for potential drop-outs and variations in presenting symptoms, a total of 150 participants (75/group) were recruited. Participants were enrolled if they presented with at least two of the four assessable symptoms: day frequency, urgency, nocturia, and/or incontinence.
One person was responsible for data transfer from paper copies to the electronic database. Although double data entry was not performed, 10% of the data was randomly checked until no mistakes were found. Less than 1% changes were made to the data set following this procedure. All data except for day frequency was non-normally distributed. Residuals (difference between predicted and actual observed value) of linear regression model for all variables did not meet the assumptions of linear regression. Hence, all comparisons for the two treatments (as change from baseline) were made using mixed effects ordered logistic regression adjusted for repeated measures (Stata, v. 13.1, StataCorp, TX, USA) as this model does not require the assumptions of linear regression to be true. Holm estimation test was used to adjust p-values for repeated measures. Analyses included only participants who were symptomatic at baseline for each parameter investigated. Data (day frequency, nocturia, urgency and incontinence) were evaluated by per-protocol and intention-to-treat analysis, with the last result brought forward for participants who dropped out or were lost to follow-up. Backward stepwise regression (p < 0.22 for covariate inclusion) indicated small effect of age, gender, water intake and diuretic use on the outcome model. However, result for adjusted and non-adjusted data were similar, hence unadjusted analyses is presented.
Responses to questionnaires were grouped for ease of reporting. The OAB-SF was analysed if the participant experienced urgency at baseline. The change from baseline is reported for the total of all OAB-SF questions, as well as sub-grouped according to classifications of symptom bother (questions 1–6; maximum score 36); difficulty coping (questions 7, 10, 11, 14, 19; maximum score 30); concern/worry (questions 8, 12, 16; maximum score 18); interference in sleep (questions 9, 13, 17; maximum score 18) and affects social interaction (questions 15, 18; maximum score 12). The IIQ questions 1–7 (maximum score 28) are reported as a total as are the responses to UDI questions 8–13 (maximum score 24).