Trial design and methods

Parts of this chapter are reproduced with permission from Wiseman et al.1 and McClinton et al.37 These are Open Access articles distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) licence, which permits others to distribute, remix, adapt and build upon this work, even for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The text below includes minor additions and formatting changes to the original text.

We prospectively registered PUrE on a publicly available website on 11 November 2015 as International Standard Randomised Controlled Trial Number (ISRCTN) ISRCTN98970319.

Trial design

We have published the PUrE study protocol37 in an open access journal.

PUrE comprised two separate pragmatic, multicentre, individual patient-randomised, open-label superiority RCTs, with an internal pilot phase, to determine which of ESWL, FURS and PCNL offers the best treatment outcomes in terms of clinical effectiveness and cost-effectiveness for people with lower pole kidney stones, seeking treatment within the UK NHS.

We compared the care pathways that started with two active interventions in each RCT.

RCT1: This compared FURS with ESWL for patients with lower pole kidney stones of maximum dimension ≤ 10 mm.

RCT2: This compared FURS with PCNL for patients with lower pole kidney stones of maximum dimension > 10 and ≤ 25 mm.

Participants

Our participants were adults (≥ 16 years), requiring treatment for a stone in the lower pole of either kidney. Adults with a stone ≤ 10 mm, confirmed by non-contrast computed tomography of the kidneys, ureters and bladder (CTKUB), were considered for RCT1. Adults with a stone > 10 and ≤ 25 mm, confirmed by CTKUB, were considered for RCT2. Our aim was to recruit 522 participants to RCT1 and 522 participants to RCT2.

Outcomes

Our primary clinical outcome was health status AUC calculated from multiple measurements of EQ-5D-5L38 up to 12 weeks post intervention. Our primary economic outcome was the incremental cost per QALYs gained at 12 months post randomisation. The trial design is summarised in Figure 1.

FIGURE 1 PUrE RCTs trial design. SF-12, 12-Item Short Form Survey; EQ-5D-5L, EuroQol-5 Dimensions, five-level version; NRS, Numeric Rating Scale; RCT, Randomised Controlled Trial. (Figure not reproduced here; see the original article.)

Patients were recruited to PUrE and were followed-up to 12 months post randomisation or 12 weeks post intervention, whichever was later.

Population

The PUrE population was adults (≥ 16 years old), with a stone ≤ 25 mm in the lower pole of the kidney confirmed by CTKUB, who presented to NHS urology departments and were considered suitable for intervention.

Potential participants and clinicians had to agree that active intervention was appropriate to remove the LPS and that the potential participant was suitable for either of the interventions for their specific stone size.

Patients with multiple renal tract stones were eligible to participate in one of the PUrE RCTs provided the stone in the lower pole measured ≤ 25 mm in maximum dimension and that the LPS was the index stone for treatment and follow-up.

The following inclusion criteria were used to identify eligible participants.

Inclusion criteria

• Adults ≥ 16 years of age.
• LPS ≤ 25 mm in maximum dimension with decision to treat that stone.
• Presence of stone previously confirmed by CTKUB.
• Able and willing to undergo either treatment for specified stone size.
• Capacity to give informed consent to participate in trial, which includes adherence to trial requirements.

Exclusion criteria

• Pregnancy.
• Patients with co-existing stone that takes precedence in deciding treatment modality (such as obstructing ureteric stone or large upper pole stone).
• Patients with health or other factors that are absolute contraindications to an intervention that they may be allocated.
• Patients unable to understand or complete trial documentation.

Identification

Sites

United Kingdom NHS hospital sites were identified via the British Association of Urological Surgeons (BAUS) from urology centres with a high rate of LPSs and an ability to deliver all trial interventions. These sites were identified using a feasibility questionnaire, which was reviewed and approved by the PUrE Chief Investigator (CI) and the Trial Office Team. Prior to starting recruitment, a site initiation visit was conducted in person at each site where the local research teams, including the principal investigator (PI), were trained in site processes and procedures by members of the Trial Office team.

Participants

The local research teams at the participating sites identified eligible participants. There were numerous routes to identifying potentially eligible patients, for example, eligible patients could be referred to urology departments because their LPS had been found opportunistically by abdominal imaging, or during investigation of urinary tract symptoms or they could present as an emergency with loin pain or infection.

The PUrE study was publicised at sites by the local research team using posters approved by the Research Ethics Committee (REC).

The local research team sent a flyer to eligible patients with information about the PUrE RCT, for which they were eligible, and at an appropriate point during their initial assessment gave patients the patient information leaflet (PIL) for RCT1 or RCT2 as appropriate.

The PILs explained the purpose and need for the RCT for which the patients were eligible, the RCT interventions and their risks, and the information the patients would be asked to complete during their participation. We developed the two PILs in conjunction with the BAUS Section of Endourology patient group.

The local research team gave patients adequate time to consider participating in the study. If the patient chose to, they could consent to participate in PUrE at the time of their initial assessment. Alternatively, they could choose to take the study information away with them for further consideration, in which case the local research team contacted the patient after at least 24 hours to determine their interest. If the patient was willing to participate in the trial, the local research team arranged with them to complete the appropriate PUrE (RCT1 or RCT2) consent form.

The local research team recorded brief details of the patients they assessed for inclusion on their site screening log. The local research team completed, and entered on the database, brief anonymised details, with reasons if given, for those patients who were assessed for inclusion but not recruited.

Recruitment and consent

All staff involved in recruitment and consent had evidence of up-to-date good clinical practice (GCP) training and had received the appropriate PUrE study training. The local research team ensured that any questions the patient had about the study were answered satisfactorily and asked them to complete the PUrE consent form if they confirmed that they were willing to participate.

Participants who could not give informed consent (e.g. due to incapacity) were not eligible to participate in PUrE. The consent form included consent to inform their general practitioner (GP) that they were taking part in PUrE, and consent for the participant to be approached for further studies on kidney stones, and for long-term follow-up through their local and central NHS clinical records after their active trial participation had finished.

Randomisation and concealment of allocation

Eligible and consenting participants were assigned to one of two treatment groups using simple randomisation (computerised random numbers). A delegated member of the local research team used the telephone 24-hour interactive voice response randomisation application or the web-based application hosted by the fully registered UK Clinical Research Collaboration, Clinical Trials Unit (CTU) at the Centre for Healthcare Randomised Trials (CHaRT), and Health Services Research Unit (HSRU) at the University of Aberdeen to assign the participant to their treatment group. Patients with a stone ≤ 10 mm, entered into RCT1, were randomised 1 : 1 to either FURS or ESWL. Participants with a stone > 10 and ≤ 25 mm, entered into RCT2, were randomised 1 : 1 to either FURS or PCNL.

The local research team informed the participants of their treatment allocation. The participant then followed the standard care pathway for the allocated intervention. In this pragmatic study, the only trial-specific interventions apart from randomised treatment allocation were the participants’ completion of outcome questionnaires.

Interventions evaluated

Experimental

• FURS as the first treatment option in a treatment pathway.

Standard

• RCT1 – ESWL as the first treatment option in a treatment pathway
• RCT2 – PCNL as the first treatment option in a treatment pathway.

At the time of funding, all three study interventions were (and still are at the time of reporting), in general use by Urology departments throughout the UK NHS. This study aimed to evaluate the interventions in a standard NHS setting for the results to be generalisable to current routine care in the UK. In line with this aim, all procedures were delivered in NHS facilities and supervised by NHS staff trained and competent in the procedures. All participants were under the care of a named consultant urologist.

The surgical interventions, FURS and PCNL, were carried out by a trained urologist, or by a trainee urologist under the supervision of a senior urologist, supported by the standard team of ward and theatre staff and radiographers. In some centres, a specialist uro-radiologist also assisted with the procedures, particularly with access to the stone.

The ESWL intervention was delivered using any device approved for this purpose, including both fixed-site and mobile lithotriptors. Delivery of the treatment was according to local practice by staff trained in the procedure; typically, a radiographer and nurse, and supervised by a urologist.

The specific technique and equipment used for FURS, ESWL and PCNL differed in fine detail between different surgeons and departments. Our PUrE protocol did not mandate the use of any specific detailed technical method for each intervention under study but as part of trial initiation of each site the standard procedure including equipment used for FURS, ESWL and PCNL for that site was recorded on a trial proforma.

For FURS, a thin (3 mm diameter) flexible endoscope (ureteroscope) is passed into the kidney via the natural urinary passages (urethra, bladder and ureter) and is used to directly see the stone. A laser fibre is then passed through the working channel of the ureteroscope and laser energy used to fragment the stone within the kidney. Larger fragments can be retrieved with a wire basket device passed through the working channel, while smaller fragments (< 2 mm) may be left to pass spontaneously. Generally, the patient will pass remaining fragments in the urine during the weeks following the procedure. The procedure is usually performed as a day case (or single overnight stay) and usually requires general or spinal anaesthesia. Patients often receive a single dose of antibiotic, at the start of the procedure to reduce the risk of infection. The duration of the operation depends on size of stone but is typically 0.5–1.5 hours. A temporary ureteric stent may be placed at the end of the procedure to protect against blockage of the ureter caused by postoperative reactive oedema or from the passage of residual stone fragments. The operating surgeon monitors the progress and degree of stone clearance during the procedure and this may be checked with imaging. Possible harms of the procedure are urinary tract infection, bleeding and damage to the urinary system, which may require a more prolonged period of stenting. The stent itself can cause pain and urinary symptoms, such as increased urinary frequency and haematuria.

For the purposes of the PUrE study, FURS treatment was expected to be a single procedure in the great majority of cases. However, an additional FURS procedure was considered part of the original FURS treatment strategy in cases of technical complexity or larger stones as long as it took place with 6 weeks of the initial FURS procedure. Any additional procedures beyond this timeframe were recorded separately for study purposes. Once stone clearance was confirmed, the ureteric stent was removed as an outpatient procedure with local anaesthetic. Placement and removal of the stent was considered part of the FURS treatment strategy for the purposes of the PUrE study.

Extracorporeal shockwave lithotripsy involves the generation of an external acoustic (sound) pulse, called a shockwave, outside the body, which is then focused onto the kidney stone, through the patient’s flank skin, causing it to fragment. Stone fragments pass down the urinary tract spontaneously, which may take a few weeks. It is routinely performed in an outpatient setting with analgesia, with or without sedation as required. Patients may be given a single dose of antibiotic at the start of the procedure if there is thought to be a higher-than-normal risk of getting an infection afterwards. Each session delivers a set number of shocks (this varies depending on the size of the stone) and stone fragmentation is monitored during the procedure and then by a plain X-ray (or other imaging as standard) taken at a follow-up appointment at approximately 3 weeks.

For the PUrE study up to two separate ESWL treatments were considered as part of the initial ESWL treatment pathway. These were to take place within an 8-week period and each episode was to be recorded separately for trial purposes. The treating urologist could however have decided that further ESWL was not appropriate if stone fragmentation was insufficient following the first or second session. The first session of ESWL was taken as the initial treatment point for the purposes of timing of outcome assessments. If the participant had additional ESWL treatments, the third and any subsequent treatments were not considered to be part of the initial ESWL treatment pathway. Possible harms of ESWL include urinary tract infection, visible bleeding in the urine and blockage of the ureter by the stone fragments. There is also a small risk of bruising surrounding the kidney.

Percutaneous nephrolithotomy is a surgical procedure to remove stones from the kidney by a direct approach. A small (10 mm) incision is made on the skin overlying the kidney, through which a needle is passed into the urine collecting tube system of the kidney. This can be guided either by simultaneous ultrasound imaging of the kidney, or fluoroscopy (X-ray) or both. This can be assisted by the preliminary telescopic placement of a tube through the urethra, bladder and ureter via which contrast fluid can then be injected into the collecting system to guide the needle passage through the skin and into the kidney. Placement of the needle is planned using available imaging (typically a CTKUB) in order to give the best access to successfully remove the stone. For a stone in the lower pole of the kidney, this is usually into the lowermost part of the collecting system. Once the needle is satisfactorily placed, a flexible guide wire is passed into the collecting system of the kidney and used to guide stretching (dilatation) of the needle track to make it wide enough for a hollow rigid access sheath to be passed, creating a channel between the skin and the urine-collecting system of the kidney. A rigid metal telescope (nephroscope) can then be inserted down this channel into the kidney’s collecting system to see the stone, which can be retrieved whole using graspers or fragmented using a variety of energy delivery devices; most commonly an ultrasonic probe or pneumatic device or combination device incorporating both energies. After the operation, the kidney may be drained for a period by a nephrostomy tube which is placed either through the access site, or internally using a stent, which is a drainage tube that lies between the kidney and the bladder. In addition, a urinary catheter may be inserted to drain the bladder for a short period after the procedure. The operation is performed under general anaesthesia with a typical duration of one to three hours depending on complexity, and patients may need to stay in hospital for up to a few days. Patients frequently receive antibiotic treatment at the start of the procedure, and this may be continued for a few days after, if there is active infection. The nephrostomy and catheter drainage tubes are usually removed after 24–48 hours without need for further anaesthesia; the stent, typically after 1–2 weeks using a flexible telescope passed into the bladder. Stone clearance is monitored during the procedure and, if necessary, by a plain X-ray (or other imaging) before discharge from hospital. Possible harms include urinary infection, bleeding (which may be severe), and inadvertent puncture of other organs. For the PUrE study, a single PCNL treatment was expected to completely remove stones up to 25 mm.

Apart from the randomised allocation of the initial intervention and participant completion of questionnaires, PUrE did not seek to change or impose any specific protocol regarding the clinical management of participants recruited at each trial site. It was anticipated that at 8 to 12 weeks post intervention, participants would receive imaging to assess stone clearance in accordance with usual standard of care.

We did however record relevant aspects of the participant care during their involvement in the trial up to 12 months after randomisation (or up to 12 weeks post intervention if this time point occurred later) and obtain patient-reported outcome measures (PROMs). In particular, trial participants undergoing any of the stone treatments may require further interventions either to correct harms arising from the initial intervention or because of incomplete stone clearance by the initial intervention. The circumstances, nature and outcome of these additional procedures were recorded, and PROMs obtained where possible before and after each additional intervention to assess effects on health status.

Blinding

This PUrE study was open label. We could not blind participants, clinical staff or the central trial team due to the differing nature of the interventions being tested. Sites were instructed to collect baseline measures (particularly those reported by study participants), before randomisation.

Data collection

We collected outcome data throughout the trial for each participant from consent until 12 months following randomisation or until 12 weeks post intervention, whichever was later. See Table 1 for the source and timing of measures.

Baseline

Participants were asked to complete the baseline questionnaire [the self-report measures; EQ-5D-5L38–44 Short Form questionnaire-12 items (SF-12);45,46 pain score on a numeric rating scale (NRS) and use of analgesics] by the local research team prior to randomisation.

The local research team also completed a baseline clinical case report form (CRF) for each participant. The baseline CRF recorded the participant’s name and full contact details, date of birth, and gender of the participant, stone dimensions, skin-to-stone distance and stone density in Hounsfield units. The local research team entered the data from the baseline questionnaire and CRF into the study database (a web-based trial management and data collection platform).

After randomisation the planned treatment date was added to the CRF and database.

Intervention and follow-up

Participants were asked to complete the questionnaires which included the EQ-5D-5L, pain (NRS) and use of analgesia, at fixed and variable time points during their trial participation.

The fixed time points were: just prior to initial intervention (pre intervention), weekly up to 12 weeks after initial intervention on their treatment pathway (first session of either ESWL, FURS or PCNL) and at 12 months post randomisation.

The variable time points after 12 weeks were just prior to, and 1 week after any additional intervention, including planned additional sessions of ESWL, removal of stent, or during any other hospital admissions related to treatment of their lower pole kidney stone, such as admissions for pain control or infection. Participants were also provided with the additional EQ-5D-5L questionnaire in a take-home study pack to complete at their discretion through their participation in the trial.

In addition, at 12 weeks post initial intervention, and at 12 months post randomisation, participants were asked to complete the SF-12 and questions relating to their primary and secondary care use and their associated travel.

We offered participants a choice of methods of questionnaire completion and reminders including postal, e-mail, web-based, telephone, SMS text with a link and in-person with a member of the local research team when in hospital. We used questionnaire reminders to facilitate questionnaire return. For the earlier time points, this was a text message or e-mail on the day that the questionnaire was due to be completed. For the later time points, 12 weeks and 12 months, this reminder was sent approximately 2 weeks, and again if required 4 weeks, after the questionnaire was due. Participants were sent a £10 multi-retailer shopping voucher as a token of appreciation for completion and return of the questionnaires at 12 weeks post intervention and 12 months post randomisation.

The local research teams completed intervention CRFs on the details and outcome following the initial and second intervention (if part of the treatment strategy). Capturing the date of the first treatment in a treatment pathway in a timely fashion was key to triggering the participants fixed-point weekly questionnaires. For participants where conservative management was agreed on their pathway, we used the date on which conservative management or the care plan was agreed to start their 12-week post-intervention period (‘nominated treatment date’). The local research team captured additional treatment sessions and stent removal at 8 to 12 weeks post intervention, after any additional stone-related treatments and admissions due to pain or infection up to 12 months post randomisation.

To measure the secondary clinical outcome of stone clearance, participants had kidney imaging between 8 and 12 weeks after treatment according to clinical need and participant convenience. We also asked local site clinical teams (radiologist/urologist) to state whether there was complete clearance of the target stone from the urinary tract (defined as no further action or observation required for that stone); acceptable clearance, where observation is required but no intervention planned; and unacceptable clearance, where further intervention would be required. We stated a preference for imaging by CTKUB during site set-up and initiation, but renal ultrasound or plain X-ray were acceptable according to patient preference, safety, and local practice.

Data management

The local research team entered the data that were collected locally onto the study database. The Trial Office worked closely with the local research teams to ensure data was as complete and accurate as possible, both through chasing missing data and by querying inconsistencies in the data.

The local research team administered the baseline and the pre-intervention participant questionnaires and the CRFs. The Trial Office administered the weekly post intervention and the 12-month postrandomisation participant questionnaires and reminders. Initially, the local research team provided participants with the 1- and 2-week questionnaires after their intervention to take home, complete and return to the Trial Office. However, it became evident that some sites were unable to provide this for the participants and therefore the Trial Office administered these questionnaires. The local research team and the Trial Office administered the variable time point questionnaires. The Trial Office entered the data from the paper questionnaires that participants returned to the Trial Office.

The Trial Office carried out data accuracy checks on a sample of data entered by sites and by the trial office as per the trial data-monitoring plan.

Participant change of status

Participants were free to withdraw their consent or change their mind about participation at any time. We collected outcome data from participants’ medical records (local research team via CRF completion), unless the participant specifically withdrew from such data collection. All data collected up to the point of participants declining further follow-up were retained and used in the analysis. All participants were followed up as per the trial protocol even if they did not get their allocated treatment.

Outcomes

All the EQ-5D-5L scores were converted to the EuroQol-5 Dimensions, three-level version (EQ-5D-3L) scale using the crosswalk as approved by NICE47 – herein referred to as the EQ-5D score.

Primary outcome

The primary outcome (PO) was patient-reported health status measured using EQ-5D score AUC from intervention to 12 weeks post intervention. The AUC was based on EQ-5D-5L completion at fixed time points: prior to initial intervention (first session of ESWL, FURS or PCNL), and weekly up to 12 weeks after initial intervention. For participants who had conservative management, we used the date when conservative management or the care plan was agreed, to start their 12-week post-intervention period (‘nominated treatment date’).

Economic outcome

The primary economic outcome was incremental cost per QALY gained at 12 months post randomisation based on the estimated NHS costs and participant responses to the EQ-5D-5L.

Secondary outcomes

Patient-reported

• Patient-reported health status measured using EQ-5D score AUC from randomisation to 12 weeks post intervention.
• Severity of pain (NRS); weekly at 1 to 12 weeks post intervention and at 12 months post randomisation.
• Generic health profile as measured by the SF-12, at 12 weeks and 12 months.
• Use of analgesia weekly at 1 to 12 weeks post intervention and at 12 months post randomisation.

Clinical

• Stone clearance measured at between 8 and 12 weeks post initial intervention using renal imaging (CTKUB preferred but plain X-ray or ultrasound acceptable).
• Maximum dimension of the largest fragment of the treated stone in mm.
• Need for additional treatment (carried out or planned) at 12 weeks post initial treatment and 12 months post randomisation.
• Complications during initial intervention.
• Post-intervention complications at 12 weeks (categorised by Clavien–Dindo classification) post treatment and up to 12 months post randomisation.

Economic

• NHS primary and secondary care resources used and their costs.
• Patient costs (out of pocket) including time off work up to 12 months post randomisation.

Stone clearance was further defined as complete clearance of the target stone from the urinary tract (defined as no further action or observation required for that stone); acceptable clearance, where observation was required but no intervention planned; and unacceptable clearance, where further intervention would be required.

Safety and breaches

PUrE involved procedures for treating LPSs which are all well established in current NHS clinical practice. Adverse effects may occur during or after any type of surgery. We did not record adverse events unless they met the criteria for a serious adverse event (SAE). SAEs were monitored and assessed by the local PI or their delegate against the criteria for seriousness, causality (related/unrelated) and expectedness and reported to the Trial Office via the trial database.

Within PUrE, ‘related’ was defined as an event that occurs as a result of a procedure required by the protocol, whether it was either (1) the specific intervention allocated at randomisation, or (2) it was administered as an additional intervention as part of normal care.

We did not consider hospitalisations for treatment planned prior to randomisation or for elective treatment of a pre-existing condition as SAEs, although complications occurring during such hospitalisation were reported as appropriate.

Hospital visits (planned or unplanned) and SAEs associated with further interventions or complications of trial treatment (e.g. expected SAEs) due to the LPS were recorded as an outcome, but not reported as an SAE. Planned primary care or hospital visits for conditions other than those associated with the LPS were not collected or reported.

All deaths for any cause (related or otherwise) were recorded on the SAE form.

Expected adverse events

The following events were considered as potentially expected for each intervention:

Flexible ureterorenoscopy: Mild burning pain passing urine; visible bleeding on passing urine; pain and frequent urination associated with a stent; kidney pain; urinary tract infection needing antibiotic treatment; damage to the ureter; leakage of urine into the body; scarring or stricture of the ureter.

Extracorporeal shockwave lithotripsy: Visible blood in urine; pain in the kidney; urinary tract infection; bruising of the skin; stone fragments stuck in ureter; kidney bruising; damage to the pancreas or lungs by the shockwaves requiring further treatment.

Percutaneous nephrolithotomy: Visible blood in the urine; fever (high body temperature); pain and frequent urination associated with a stent; severe kidney bleeding; damage to adjacent organs, such as lung, bowel, spleen or liver; damage to kidney or infection needing further treatment; leakage of urine into the body; leakage of irrigating fluids into the body.

It was a requirement to report any SAEs that were deemed related and unexpected to the sponsor within 24 hours of receiving the signed SAE notification. Such SAEs were also reported to the REC within 15 days of the CI becoming aware of the event. All related SAEs were summarised and reported to the appropriate authorities within their regular progress reports.

Breaches

Sites were asked to report potential breaches of trial protocol or GCP to the trial office. Trial office staff could also report potential breaches. Five breaches were recorded within the study and these are summarised in Appendix 1, Table 32. The CI and sponsor assessed these breaches as non-serious.

Sample size

The PO was the AUC measured from multiple completion of the EQ-5D-5L by each participant up to 12 weeks post initial intervention (first of allowed session of FURS, ESWL or PCNL). Each RCT was designed to detect a target difference of 0.3 standard deviations (SDs), with 90% power and 5% two-sided alpha. This required outcome data on 235 participants per group (470 total). Such a difference in generic health status was considered clinically relevant and in terms of treatment effect size, in the small to medium range as observed in other clinical studies.48,49 To allow for completely missing outcome data for approximately 10% of participants, and therefore for whom the AUC could not be calculated, we planned to randomise 522 participants in both RCT1 and RCT2, giving a total trial population of 1044 participants.

Statistical analysis

The trial analysis followed a statistical analysis plan (SAP). The SAP was drafted by the trial statistician in consultation with the Project Management Group (PMG). It was reviewed by the Trial Steering Committee (TSC) and Data Monitoring Committee (DMC) before analysis started. RCT1 and RCT2 were analysed separately. Full definitions of the derived outcomes are given in Appendix 2 (see Tables 33–35).

We based the main statistical analyses on all participants as randomised, intention to treat (ITT), irrespective of subsequent compliance with the treatment allocation. We used a treatment policy approach for our PO estimand which ignores any intercurrent events because the primary question of interest in each trial was a comparison of treatment pathways. In RCT1 there was a longer waiting time for one of the treatments. We tried to account for this by adjusting for waiting time, using a hypothetical strategy approach that attempts to answer the question: which pathway is best if there is no difference in waiting times.

We conducted a single analysis that took place following completion of trial follow-up.

We summarised baseline and follow-up data using appropriate descriptive statistics and graphical summaries.

Primary outcome derivation and analysis

The PO was generated for each participant using the trapezoidal rule. Data for participants who missed a scheduled time point was estimated using a multiple imputation approach to make use of partial outcome data. See Appendix 2 for the derivation of AUC in general using validated Stata code (package called eq5dmap) approved by NICE47 in the UK.

To be included in the primary analysis, participants needed responses at three specific time epochs described in Appendix 2, along with notes for handling intervening missing data.

The PO was analysed using linear regression with adjustment for centre (using robust variance estimates) and baseline EQ-5D score. Additional analysis adjusted for pre-intervention EQ-5D score and waiting time between randomisation and initial intervention.

Secondary outcome analysis

Secondary outcomes were analysed using the appropriate generalised linear models with adjustment for baseline variables as appropriate. For example, categorical variables were analysed using ordinal regression with robust error variance accounting for centre. Binary variables were analysed using a log-binomial model with robust error variance accounting for centre. Secondary outcomes that were measured repeatedly were analysed using mixed models, with a random effect for participant and a fixed (nominal) effect for time, treatment and design covariates, and modelled with robust error variance accounting for centre.

All treatment effects sizes were summarised by between-group estimates and 95% CIs from the appropriate models. All analyses were carried out using Stata (version 17, StataCorp. Stata Statistical Software: Release 17; 2021) (StataCorp LP, College Station, TX, USA).

Subgroup analysis

Subgroup analyses explored the possible modification of treatment effect by important factors; [participant body mass index (BMI), stone size (maximum dimension), stone density on CTKUB (Hounsfield units) and skin-to-stone distance]. Analysis was done by including a subgroup-by-treatment interaction in models. The CIs were set at 99% for this exploratory assessment.

Within group exploratory analysis

We explored within each allocated group whether technical factors modify the treatment effect [access sheath vs. no access sheath and digital vs. non-digital instrument (FURS); fixed site versus mobile device (ESWL); calibre of access tract (PCNL)]. This used treatment-by-factor interactions in the model and were classified as exploratory analyses.

Economic evaluation

We conducted a ‘within-trial’ and a model-based economic evaluation. We describe these in detail in Chapter 5.

Ethics approval and monitoring

PUrE received favourable ethics opinion from the North of Scotland REC on 12 November 2015 (REC reference number 15/NS/0113, IRAS ID 188563) and the Health Research Authority on 9 June 2016. Each participating NHS organisation reviewed the study and gave confirmation of capacity and capability [or research and development (R and D) management approval] before the study commenced at site.

Sponsorship

The University of Aberdeen and NHS Grampian co-sponsored PUrE.

Management of the trial

The trial management team (the Trial Office), consisting of the Trial Managers, a Data Co-ordinator and the CI, were based within CHaRT. The trial management team took responsibility for the day-to-day transaction and coordination of trial activities and processes, including approvals, site set-up and close down, plus day-to-day support for the recruiting sites, provision of approved trial paperwork, administration of participant questionnaires and data entry, chasing and checking using the trial web data entry portal. This trial management team worked with the wider PUrE team, the CHaRT programmers, the trial statistician, health economist and grant holders to deliver the PUrE study.

The PUrE web portal for data collection and trial management was developed and maintained by the programming team within CHaRT to a specification agreed with the trial management team.

Each recruiting site was led by a local PI. The PIs in most cases were supported by research nurses, trial co-ordinators or dedicated staff (the local research team), who were responsible for all aspects of the local organisation, including recruitment of participants, delivery of the interventions and notification of any problems or unexpected developments during the study period.

Oversight of the study

Project Management Group

The PUrE PMG was responsible for overseeing the management of the study. This group consisted of the representatives from the trial management team, grant holders [including the patient and public involvement (PPI) representative] plus the trial statistician. The members of the PMG are listed in the acknowledgements.

Trial Steering Committee

One independent TSC was responsible for monitoring and supervising the progress of the PUrE Study (both RCT1 and RCT2). The TSC defined and agreed their role and responsibilities in the TSC Charter and met nine times between February 2016 and December 2022. The TSC consisted of three independent experts, including a PPI representative, and the CI. Key members of the PMG also attended TSC meetings. The members of the TSC are listed in the acknowledgements.

Data Monitoring Committee

One independent DMC was responsible for monitoring safety and data integrity for both RCT1 and RCT2. The DMC consisted of three independent experts. It was independent of the PMG and TSC. The DMC members defined and agreed their roles and responsibilities, and meeting frequency via the DMC Charter at the outset of the PUrE study. The DMC met six times plus had one e-mail update, between May 2016 and December 2022. The trial statistician provided the data and analyses requested by the DMC prior to each meeting. The DMC members are listed in the acknowledgements.

Patient and public involvement

Our PPI grant co-applicant attended PMG meetings and TSC meetings throughout the study. He provided the other members of the team with his individual perspectives on stone disease, reviewed and commented on the study design, protocol and all study documentation, including patient-facing documents and the regular funder progress reports.

In addition, our PPI TSC member attended TSC meetings and provided review and comment on the study and the TSC papers.

Both were involved in discussions of the study results with the TSC and the trial investigators and contributed to the preparation of the plain language summary. They continue to be involved in developing dissemination materials for participants and contribute to academic papers. The PPI partner on the TSC will be asked to co-produce the participant results letter. They have reflected on their input and made suggestions for future research, which is included in the discussion.

Protocol amendments

There were four protocol amendments. These are summarised in Appendix 3, Table 36. All were reviewed by the sponsor and the study funder before being submitted to REC as appropriate.

Important changes to the methods after the start of the trial

Extension to recruitment and follow-up phase

Recruitment was slower than anticipated in both RCT1 and RCT2, so we requested an extension to the recruitment phase of 18 months, which was granted by the National Institute for Health and Care Research (NIHR) Health Technology Assessment (HTA) programme. Due to the COVID pandemic, recruitment was paused for 4 months, on the direction of the sponsor.  The follow-up phase of the study was then extended for 9 months to allow participants who were waiting for their intervention, delayed due to the pandemic, to receive their allocated treatment and complete their 12-week post-intervention follow-up. There were no other important changes to the methods after the start of the trials.

Studies within a trial and satellite studies

Christmas card study within a trial

The PUrE study, alongside seven other host trials, was involved in the Christmas Card Study Within a Trial (SWAT). The aim was to determine the effect on retention of sending Christmas cards to RCT participants and to explore the feasibility of a SWAT across multiple trials simultaneously. Full details of the methods and results have been published.50,51

Theoretically informed leaflet study within a trial

The PUrE study was involved in a SWAT to explore the benefit of adding a theoretically informed leaflet to the take-home pack with the 1- and 2-week questionnaires. This SWAT aimed to increase the response rates to the participant-reported outcomes (self-completed questionnaires) for the PUrE study by incorporating techniques based on behaviour change theory into a leaflet, developed from qualitative patient-reported studies and PPI input. Appendix 4, describes the methods and results as presented in a poster at the 6th International Clinical Trials Methodology Conference, October 2022.

Cambridge Renal Stone Patient Reported Outcome Measure validation study

We asked a sub-set of participants to complete the Cambridge Renal Stone Patient Reported Outcome Measure (CReSP)52 a disease-specific health-related QoL measure in development, as part of the validation studies for the CReSP. The validation studies will be reported separately by the CReSP developers and are not included in this report.