Pilot studies play an important role in health research. A well-conducted pilot study with clear aims and objectives encourages methodological rigour, ensures that the work is scientifically valid and publishable, and results in higher quality RCTs [
21]. Lancaster et al. produced a framework and addressed some of the methodological reasons for undertaking pilot studies related to RCTs [
21]. Using this framework as a guide we summarise below our justifications for undertaking a pilot study and its methodology.
The randomised external pilot trial rehearses the methodology proposed for the definitive RCT. A run through of the trial procedures on a smaller scale will practise and identify any logistical issues that might be modified. The number of eligible patients and rates of recruitment and randomisation will confirm whether the required sample size for a definitive trial could be identified and recruited. Recruitment has implications for both timing and funding of a definitive trial and poor recruitment is one of the major reasons for abandoning trials early [
21,
38]. A RCT to address whether IUT alters treatment decisions and treatment outcome in women suitable for surgical treatment of SUI or stress predominant MUI is likely to require large numbers of participants and would therefore be expensive and time-consuming. Prior to allocating resources to such a large trial feasibility must be determined to reduce the risk of the definitive trial not being able to recruit to time and target.
Data collection for the proposed definitive trial consists of both clinical records data and patient self-completed questionnaires. The outcome questionnaires rely on patient recall over a six month period and analysis of completion rates and quality of data will not only test comprehensibility but guide timing of future data collection. Several secondary outcome measures have been chosen with a view to refining the choice on the basis of data yield and quality. The demands of a trial are an identified barrier to patient participation and refinement of the data collection tools will ensure these are kept to a minimum [
38].
Qualitative interviews with the women will explore their reasons for participating, or not, and their experiences of the trial with the potential benefit of guiding protocol amendments such as the consent procedure and questionnaire format, to improve the recruitment process and questionnaire responses.
The acceptability of the protocol is an important aspect of the trial design. As mentioned previously IUT is widely established in clinical practice and the absence of this investigation prior to offering surgery is a potential barrier to clinicians' willingness to randomise their patients. Clinician surveys in other urogynaecology research have identified clinicians' views of perceived benefit and complications of the two arms as impacting participation [
14]. Women may also not see the benefit of testing a test, believing that the results of any available test must guide their clinician and themselves to the most appropriate management. Prior to embarking on a definitive RCT we need to ensure that not only are the participants willing to be randomised but that the clinicians are also willing to randomise their patients. A UK national survey of clinicians' views on IUT in this particular context and their willingness to randomise patients within a definitive trial will help establish if sufficient clinicians are in equipoise. Qualitative interviews with a sub group of clinicians will help contextualise the responses. During the qualitative interviews with the women willingness to be randomised will also be explored.
Attempts to estimate sample size for a definitive trial were unproductive because of uncertainty in the distribution of the key variables. The pilot study aims to collect data on the primary outcome measure, the ICIQ-FLUTS combined symptom score, to calculate the sample size for a definitive trial.
The decision to proceed with the definitive trial as it is currently envisaged will only be made if all aspects of feasibility, including adequate patient recruitment and clinician engagement are established. The combination of quantitative methods (e.g. recruitment and retention rates, percentage of clinicians in clinical equipoise) and qualitative methods (e.g. an assessment of whether any identified barriers to recruitment and retention could be overcome by modifications to trial design or procedures) will inform the decision and ensure its effectiveness in achieving the research aim.
Currently there are two other on-going trials looking at the clinical utility of urodynamics in similar patient groups; the VaIUE trial (Value of Urodynamic Evaluation)[
39] and VUSIS-2 trial (Value of Urodynamics prior to Stress Incontinence Surgery) [
40]. Similar to INVESTIGATE-I, the ValUE study randomises women with a clinical diagnosis of SUI or stress predominant MUI to either no further assessment or IUT. In the VUSIS study all women undergo invasive urodynamic testing. Those with discordant clinical and urodynamic findings are randomised between surgical treatment as dictated by their clinical assessment and individualised treatment dictated by a combination of clinical and urodynamic results and therefore addresses a different clinical question. Both of these definitive trials use a non-inferiority design [
41]. Whereas VUSIS does not define a non-inferiority margin VaIUE defines a margin of 11% which we consider somewhat high and a difference that might potentially influence the decisions of both clinicians and patients. The primary outcome of both is based on the Urogenital Distress Inventory (UDI) score at 12 months. Although our primary outcome is the ICIQ-FLUTS combined symptom score we have chosen to include the UDI as an additional secondary outcome. If we subsequently proceed to the definitive trial, INVESTIGATE-II, assuming the other studies complete recruitment and publish their results, this will allow easier comparison of results. While we are encouraged to see that others look on this topic as being an important clinical uncertainty, we remain of the opinion that a feasibility study is an important step before embarking on a definitive trial.
1
This article has been