Types of urethral catheter for reducing symptomatic urinary tract infections in hospitalised adults requiring short-term catheterisation: multicentre randomised controlled trial and economic evaluation of antimicrobial- and antisepticimpregnated urethral catheters (the CATHETER trial)

Urethral catheter design

Tubes that can be inserted through the urethra to drain the urinary bladder have been used for centuries. The current standard single-use indwelling catheter design was developed by a urologist, Frederic Foley, in the mid-1930s and marketed by the American company CR Bard Inc. The catheter consists of a drainage channel, open both at the tip positioned in the bladder and at the other end outside the body, which can be connected to a drainage bag. 1 A second channel allows inflation, through a port with a non-return valve next to the drainage outlet, of a 10-ml retention balloon, which is positioned in the bladder and prevents the catheter coming out (Figure 1). The catheter is inserted into the urethra by a trained health professional, using an aseptic technique and lubricating anaesthetic gel, and is left indwelling for as long as is necessary; removal requires deflation of the balloon and simple withdrawal of the catheter. 2

FIGURE 1.

A urethral catheter positioned in the female and male bladder.

The calibre, length, number of channels and material of manufacture of the catheter can all be varied. 3 Natural extruded rubber (latex) continues to be used as the standard material of manufacture, although usually with an added internal and external water-resistant coating of polytetrafluoroethylene (PTFE) to ensure that smoother surfaces are in contact with the urethral mucosa and urine. In 2011, these standard catheters in the UK NHS cost approximately £0.914 (US$1.48; €1.03) each. More recently, catheters made out of moulded plastics, such as silicone, have become available. These have the advantage of consistent smooth surfaces and hypoallergenicity but with higher costs of material and manufacture with a unit cost to the UK NHS in 2011 of £2.074 (US$3.41; €2.37).

Use of urethral catheters in hospitals

Urethral catheters are one of the most commonly applied medical devices, with an estimated 96 million used worldwide in 1999. 5 Within the UK, approximately 15–25% of the 14.5 million patients admitted to NHS hospitals each year will receive catheterisation at some stage during their stay. 6–9 The most common usage is for short term (which we have arbitrarily defined as up to and including 14 days) perioperative bladder drainage during and immediately after surgical or other interventional procedures. This patient group is the primary focus for our trial. 10 Other reasons for urethral catheterisation include prolonged unconsciousness or immobility, monitoring of urine output for management of fluid balance in critically ill patients, acute urinary retention and longer-term care of urinary incontinence. For patients undergoing interventions that require temporary catheterisation, the catheter is generally inserted just prior to starting the procedure and is removed when the patient is sufficiently recovered in terms of bladder function and mobility to safely re-establish normal micturition. The duration of catheterisation for these purposes is highly variable, with a recent study recording a mean [standard deviation (SD)/median] of 3.5 (4.8/2) days. 11 The type of catheter used depends on local purchasing policies and indication for use; for example, a recent audit in a large north-east England acute care hospital, which is likely to be typical of current NHS practice, showed that of the 15% of inpatients with an indwelling catheter, 50% were fitted with a silicone catheter and 36% a PTFE-coated latex catheter; in 14% the catheter type was unknown. 12

Definition of catheter-associated urinary tract infection

Normally, the urethral antimicrobial barrier is closed, preventing infection. The presence of an indwelling urinary catheter disrupts this barrier and allows colonisation of the urethra and subsequently the bladder with commensal and externally acquired organisms. These may result in bacteriuria, symptomatic urinary tract infection (UTI), and, rarely, bloodstream infection; these are collectively referred to as CAUTI. 13,14 For epidemiological and health protection purposes, CAUTI continues to be predominantly defined worldwide through consensus policy documents produced by the US government Centers for Disease Control and Prevention (CDC).

The CDC definition of symptomatic UTI associated with urethral catheterisation at the time of trial inception (2007)15 required the presence of appropriate symptoms together with either microbiological confirmation of significant bacteriuria or that a clinical diagnosis of UTI had been made and treatment instituted (Box 1).

The CDC definitions were updated in 2009 while our trial was in progress. 16 The 2009 specification states that symptomatic infection can be deemed catheter-related only if an indwelling urinary catheter had been present within 48 hours of diagnosis. In addition, there must be evidence of concurrent symptoms and urinary abnormality – either pyuria or bacteriuria. Criteria 2f and 2g concerning physician diagnosis and treatment initiation have been removed in the updated 2009 version. This more restricted definition was designed for reporting of CAUTI from US hospitals as part of national surveillance of health-care-associated infections. It is less appropriate for our pragmatic trial design where we wished to additionally capture the community impact of CAUTI following discharge from hospital and avoid reliance on submission and microbiological analysis of urine specimens. For the primary outcome of the trial we therefore based definition of CAUTI on criterion 2g of the 2004 CDC definition. We do recognise, however, that other definitions of CAUTI have been used in previous trials. We therefore defined microbiologically proven antibiotic-treated UTI and the presence of bacteriuria irrespective of symptoms or treatment as tertiary outcomes.

The impact of catheter-associated urinary tract infection on hospital-based care

It is estimated that individuals with an indwelling catheter are faced with a daily risk of 5% of developing bacteriuria,6 with the proportion affected after 7 and 14 days of indwelling catheterisation being approximately 35% and 70%, respectively. It has been estimated that symptomatic UTI occurs in 20% of patients with bacteriuria,17,18 and while bloodstream infection occurs in < 1%,19,20 it is associated with a high (30%) mortality rate. An important possible consequence of development of CAUTI in an individual is the prolonging of hospital stay. Estimates of the average duration of this extra stay vary from 0.55 to 5 days. 21 CAUTI can also adversely affect health-related quality of life (QoL). 5,22 Both of these factors are important in health economic terms, as they impact on both QoL reduction and additional treatment costs, which, for the UK in 1995, were estimated to amount to a mean [95% confidence interval (CI)] of £1327 (£1140–1465) per episode, which extrapolates to a cost of £125M per annum to the NHS. 21 The presence of bacteriuria in hospital patients with an indwelling catheter is also a potential source of cross-infection, particularly in critical care units, with an estimated risk per episode of 15%. 23 In addition to CAUTI risk, short-term indwelling urethral catheters, although useful to avoid the need for voluntary bladder emptying, also contribute to postoperative discomfort, loss of dignity and delayed discharge from hospital.

The increasing realisation that hospital-acquired infections account for significant morbidity and mortality, together with associated financial and personal costs, led health-care providers to develop strategies to reduce the burden of these events. UTI, in general, is one of the two most common hospital-acquired infections, accounting for between 20% and 40% of cases. 24–26 Between 56% and 80% of these cases can be attributed to the use of indwelling urethral catheters. 20,26,27 CAUTI is therefore a major focus of these preventative strategies. The UK government Department of Health set up the Saving Lives initiative in 2007 and, subsequently, the High Impact Actions for Nursing and Midwifery initiative in 2009, which made reduction in CAUTI a key aim for the NHS through High Impact Action number 6. 28 This was based on guidelines for urethral catheter insertion and subsequent care developed from a systematic review of the evidence. 29

Pathogenesis of catheter-associated urinary tract infections

The development of bacteriuria associated with an indwelling catheter is thought to occur in stages. Microbes gain entry to the normally sterile upper urethra and bladder either by physical introduction during catheter insertion or by migration along the interface between the outer catheter surface and the urethral mucosa, or by upward migration through the internal catheter channel lumen following colonisation of the drainage system. 30 The disruption of the normal filling and emptying cycle of the bladder, and the position of the catheter drainage channel inlet above the catheter balloon resulting in a urine residue in the bladder, prevent physical removal of invading bacteria and so facilitate urinary colonisation. 31 Bacteria then adhere to the urinary tract epithelium or the catheter surface and excite an inflammatory reaction resulting in local and systemic symptoms. 32–34 Bacteria expressing more aggressive virulence factors can migrate further into the upper urinary tract or bloodstream, resulting in worsening sepsis. 35 As the duration of catheterisation increases, bacteria begin to attach to the catheter surfaces, where they multiply and produce polysaccharides, forming a biofilm. 36 Subsequently, mineral precipitation due to increased urinary pH causes catheter blockage and urinary stagnation, further facilitating bacterial growth. 37

Microbiology

Urinary tract infection associated with short-term catheterisation typically involves a single organism, in contrast with long-term catheterisation, where polymicrobial infection is frequent (Table 1). 38 Although a variety of microorganisms may be associated with CAUTI, enteric Gram-negative bacilli are the most frequently isolated. 40 Escherichia coli is the most frequently isolated single species, but other species such as Klebsiella spp., Proteus spp. and Enterobacter spp. are also commonly identified. Enterococci, Pseudomonas aeruginosa and Candida spp. are also important causes of CAUTI, particularly in patients within critical care settings. 40 Staphylococci and other Gram-negative bacilli are isolated less frequently. 38

Risk factors for catheter-associated urinary tract infection

A number of patient characteristics are associated with increased risk of CAUTI, including female sex, older age, impaired immunity and illness severity. Care process factors include lack of antibiotic use, longer duration of catheterisation, insertion by poorly trained personnel, and deviation from catheter care protocols. 38

Measures and strategies aimed at reducing catheter-associated urinary tract infections

There are a number of existing recommendations that can reduce the risk of CAUTI (Box 2) and the evidence for each of these has been summarised in recent reviews. 29,38,41–44

A range of UK guideline and best practice documents have encouraged implementation of some of these measures into the NHS with performance monitoring of individual NHS organisations to ensure compliance. These can be broadly divided into eight distinct actions. 29,45–48

• education of patients, their care-givers and health-care personnel, in terms of hand hygiene and steps in preventing spread of infection

• assessing the need for catheterisation – to avoid or consider alternatives

• selection of appropriate type of catheter for individual and clinical context

• use of strict aseptic technique for catheter insertion

• use of antibiotic prophylaxis in selected high-risk groups at insertion

• use of a closed drainage system or catheter valve

• maintenance of a sterile closed drainage system by obtaining urine specimens from the sampling port, positioning of drainage bag above floor level and below bladder level; frequent emptying of drainage bag to maintain urine flow and prevent reflux; daily washing of meatus

• reduction in duration of catheterisation: regular review of need for catheterisation and aim for early removal of catheter.

Development of urethral catheters containing antimicrobials

The most researched technical innovation in catheter design over the past 10 years has been the introduction of antimicrobial coatings applied to catheter surfaces or impregnated into the catheter material. Several manufactures have marketed either antimicrobial-impregnated or silver-coated catheters as representing a technology to reduce CAUTI risk and it is this development that is the focus of this trial. 23

Evidence for clinical and bacteriological effectiveness

Several systematic reviews have investigated the effectiveness of antimicrobial catheters in reducing CAUTI, including pooled data from up to 13,000 patients. The results of five systematic reviews23,42,56–58 suggest that silver alloy-coated catheters reduce the incidence of bacteriuria in hospitalised patients catheterised for < 2 weeks in comparison with standard catheters. The magnitude of relative risk reduction varied in each of the analyses due to different inclusion criteria, ranging from 16% (95% CI 6% to 47%; absolute risk reduction of 2.0%)58 to 46% (95% CI 33% to 57%; absolute risk reduction of 11.3%). 42 The pooled results for nitrofurazone-impregnated catheters showed a relative risk reduction of up to 48% (95% CI 32% to 66%; absolute risk reduction 8.6%);42 however, the benefit of this type of catheter appeared to be limited to the initial 7 days of catheterisation. It should be noted that these event rates and associated risk reductions apply to CAUTI diagnosis based solely on microbiological identification of bacteriuria without any patient-driven or clinician-defined contribution to the primary outcomes used. Overall, the methodological quality of these reviews was poor to moderate according to the AMSTAR quality assessment tool,59 with the exception of that by Schumm et al. ,42 which was methodologically more robust. [K Schumm (now K Gillies) and T Lam, authors of the updated Cochrane review, were both members of the trial steering and project management groups and are named authors of this monograph.]

In summary, these systematic reviews have concluded that silver alloy-coated and nitrofurazone-impregnated catheters do show promise for the reduction of CAUTI. The reviews did, however, highlight a number of uncertainties regarding the evidence base, in particular the clinical and health economic relevance of the outcome measures used.

Problems with current evidence

Inclusion criteria

Adults (≥ 16 years of age) requiring urethral catheterisation (which was expected to be required for a maximum of 14 days), from selected hospital wards with a high volume of short-term catheterisation.

Exclusion criteria

• Patients for whom urinary catheterisation was expected to be longer-term (defined as > 14 days).

• Patients who had urological intervention or instrumentation within the 7 days preceding recruitment (e.g. catheterisation, cystoscopy, prostatic biopsy and nephrostomy insertion).

• Patients who required non-urethral catheterisation (e.g. suprapubic catheterisation).

• Patients who had a known allergy to any of the following: latex, silver salts, hydrogel, silicone or nitrofurazone.

• Any patient who had a microbiologically confirmed symptomatic UTI, at time of randomisation.

• Patients who were unable to give informed consent or retrospective informed consent.

Participants were equally allocated to one of the three trial interventions using a web- or telephone-based system managed by the Centre for Healthcare Randomised Trials (CHaRT), University of Aberdeen. The inclusion criterion for participation of centres was a high volume of short-term catheterisations, principally as part of elective surgical activity. Twenty-four centres took part in the trial: Aberdeen Royal Infirmary; Royal Blackburn Hospital & Burnley General Hospital; Blackpool Victoria Hospital; Bristol Royal Infirmary; Edinburgh Royal Infirmary; Guy’s Hospital; Harrogate District Hospital; Hillingdon Hospital; Hinchingbrooke Hospital; Raigmore Hospital, Inverness; Liverpool Women’s Hospital; Newcastle upon Tyne Hospitals (Newcastle General Hospital, Freeman Hospital, and Royal Victoria Infirmary); Norfolk and Norwich University Hospital; North Tyneside General Hospital; Nottingham City Hospital; Royal Preston Hospital; Queen Alexandra Hospital, Portsmouth; Southampton General Hospital; Sunderland Royal Hospital; Musgrove Park Hospital, Taunton; Torbay Hospital and Yeovil District Hospital. The trial recruited from a wide range of clinical settings including cardiovascular, obstetrics and gynaecology, orthopaedics and neurosurgery as detailed in Chapter 4.

Planned trial interventions

There were two experimental groups of equal importance:

• nitrofurazone-impregnated silicone urethral catheter (N), sourced from Rochester Medical, UK (product reference number 95214)

• silver alloy-coated latex hydrogel urethral catheter (S), sourced from CR Bard Ltd, UK (product reference number 236514UKS).

The control group was managed with a PTFE-coated latex urethral catheter (P), sourced from CR Bard Ltd, UK (product reference number 1254S14UK).

All catheters used in the trial had an external circumference of 14 mm, termed 14 French (Fr) or 14 Charrière (Ch), with equivalent luminal calibre, length, recommended balloon volume (10 ml), drainage ports and external connection fittings. The silver alloy-coated and control catheters were manufactured from latex, whereas the nitrofurazone catheters were made from silicone. The choice of a PTFE catheter as the ‘standard’ control was based on the results of an audit of short-term catheter use in all secondary care wards in Newcastle upon Tyne Hospitals and Aberdeen Royal Infirmary, which confirmed that the PTFE-coated latex urethral catheter was the most commonly used in both hospitals (> 70%).

Proposed duration of intervention

The period of urethral catheterisation was expected to be between 1 and 14 days.

Primary outcome measure

Incidence of symptomatic UTI treated with antibiotics at any time up to 6 weeks post randomisation (number of participants with at least one occurrence). This was defined as any symptom reported up to 3 days after catheter removal, at 1 or 2 weeks post catheter removal, or at 6 weeks post randomisation combined with a prescription of antibiotics, at any of these times (see Table 4).

Subgroup analyses of the primary outcome examined possible interaction with participant age, sex, comorbidity, duration of catheterisation, indication for catheterisation and antibiotic use prior to trial enrolment.

Details of which data contributed to the primary outcome are shown later in this chapter (see Table 4), and Chapter 3 (see Important changes to methods after trial commencement) details how the attributes of the primary outcome were strengthened during the course of the trial.

Secondary outcome measures

Tertiary outcome measures

Individual analysis of the components of the definition of the primary and secondary outcomes is shown below.

Trial flow

The trial process is detailed in Figure 2.

FIGURE 2.

The CATHETER trial recruitment processes and procedures.

Identification of patients

Patients were identified either by a member of the local research team or by ward staff. In order to publicise the trial to ward staff and patients, laminated recruitment posters were placed at prominent locations at each site as appropriate.

Recruitment process

Once a patient was identified as being eligible for the trial, he/she was approached by a member of the local research team and given the trial patient information sheet (see Appendix 2). Once the patient had been given time to consider and understand the implications and requirements of the trial, and was happy to take part, he/she was asked to sign the trial consent form (see Appendix 2).

The only exception to this was for patients in unplanned situations (e.g. non-elective admissions and catheterisations). The local research team was made aware of this and, once the patient’s condition had stabilised, he/she was provided with the patient information sheet and given an opportunity to consent retrospectively to take part in the trial; this methodology for recruitment was approved as part of our ethics submission. The decision to catheterise such patients was based solely on clinical need by the local clinical care team. As the antiseptic/antimicrobial-impregnated catheters have previously been reported to lessen the risk of infection compared with the usual standard catheter, the inclusion of these patients in the trial prior to informed consent did not pose a clinical risk, did not lessen their standard of care and was not thought to be otherwise disadvantageous to the participant. Patients who subsequently gave informed consent then completed the normal trial processes. Patients who declined to give informed consent retained the trial catheter as appropriate but no trial data were collected.

Randomisation and allocation to intervention

Eligible participants were randomly allocated 1 : 1 : 1 to one of the three interventions using a computer-generated system that was concealed and remote from the users. The local research team or ward staff performed randomisation using either the automated interactive voice recognition (IVR) telephone randomisation application or the CATHETER trial website (www.charttrials.abdn.ac.uk/catheter), both managed by CHaRT, University of Aberdeen. Both methods of randomisation were available 24 hours a day, 7 days a week. Compliance with the allocated intervention was recorded. No stratification or minimisation was used.

Blinding of trial interventions

The nitrofurazone catheter was easily identifiable because of its bright-yellow colour. To guard against bias in this respect, although the recruiter knew the allocated intervention, as far as was practicable participants and clinicians making decisions regarding the participant’s catheter care were not told. General practitioners (GPs) were also not informed by trial staff of the catheter type the participant received and therefore were unlikely to be influenced by knowledge of the catheter type. Urine samples taken at baseline and 3 days after catheterisation were analysed by staff at a local laboratory, who were blind to the intervention.

Interventions

The calibre of urinary catheters is measured as the external circumference in millimetres, Fr or Ch gauge, and they can be supplied in differing lengths to suit the male (38–46 cm) or female (24–28 cm) urethra. The uniform calibre and length of the three catheters tested in this trial was specified as 14 Fr and 40 cm, respectively. 68 This was chosen as 14 Fr is the standard calibre for short-term monitoring use. A longer urethral catheter does not cause problems when used in women, but it is not possible to use a shorter catheter in men. They were purchased direct from the manufacturers (CR Bard Ltd UK, Rochester Medical, UK) by the trial office at a unit price fixed in 2007 for the duration of the trial and distributed by the trial office in Aberdeen to the sites as needed. Each catheter had a detachable sticker attached to the outer packaging. This sticker displayed the ‘CATHETER’ logo and either ‘N’, ‘S’ or ‘P’ to denote catheter type (N = nitrofurazone, S = silver, P = PTFE). These stickers were then placed directly on to the participants’ consent forms to permit verification that they were given the catheter to which they were randomised.

Data collection

Questionnaires and diaries were designed to obtain information on symptomatic UTIs as well as other catheter-associated problems (e.g. urethral discomfort), QoL, and any health economic implications, such as costs to the participants and the NHS (see Appendix 3). Clinical data at baseline and throughout the participant’s hospital stay were collected by the local research team (see Appendix 4). An amended version of the UTI Symptom Assessment Questionnaire69 was used to assess UTI symptoms 3 days post catheterisation.

Following informed consent, the participants were asked to complete a baseline questionnaire before randomisation (for participants undergoing unplanned catheterisation, randomisation occurred before informed consent and the baseline questionnaire was then collected after consent was obtained). At baseline, a sterile mid-stream specimen of urine (MSSU) was collected and sent for microbiological analysis, in an accredited laboratory [Clinical Pathology Accreditation UK (CPA)] according to local diagnostic protocols, immediately prior to catheterisation (if one had not been sent within the preceding 48 hours). In situations where this was not possible, a specimen of urine was obtained during the process of catheter insertion, i.e. catheter specimen of urine (CSU) using standard aseptic techniques.

Three days post catheter removal, participant-reported outcome data were recorded by a questionnaire given to the participant by the local research team. When possible, these data were collected while the participant was hospitalised, but in situations where the participant was discharged before this time point, they were asked to complete this at home and return it to the trial office. An MSSU was collected for culture within or at 3 days of catheter removal (analysed in CPA-accredited laboratories according to local diagnostic protocols). For the purposes of the trial we defined a positive urine culture as the presence of at least 10⁴ (CFUs)/ml.

If a clinical diagnosis of symptomatic UTI was made at any stage, including during the period of catheterisation, either a CSU or MSSU was obtained according to normal clinical practice.

Participants were asked to complete diaries at 1 and 2 weeks post catheter removal and at 6 weeks post randomisation. Table 2 describes the outcome data collected.

Diaries and questionnaires completed prior to hospital discharge were collected by the local recruitment co-ordinator. On discharge, patients were supplied with a pack containing the relevant 1- and 2-week diaries and pre-paid addressed envelopes to return the diaries to the trial office. At 6 weeks post catheterisation, participants were sent the follow-up questionnaire from the trial office and asked to return it using the pre-paid addressed envelope provided. Where necessary, participants who did not return their diaries or questionnaires were telephoned either by the local research co-ordinator or a member of the trial office and asked to return their paperwork. If possible, primary outcome data were collected by phone at this point.

Where the period of catheterisation was longer than initially anticipated (i.e. the catheter was still in place after day 14), trial data were collected as if the catheter had been removed on day 14 (e.g. the 3-day post removal questionnaire was completed 17 days post catheterisation). The date of catheter removal was subsequently recorded. The rationale for this was that it would not be appropriate to exclude data from these participants but it also would not be justified to prolong follow-up for any participant beyond the planned 6 weeks after randomisation. Given the likely small numbers of participants affected, we chose to use the same documentation but we instructed local trial staff to aid participant completion, as the majority of patients affected were likely to remain in hospital during the extended period of catheterisation. This ensured that we captured any UTI event during the 6-week period and ensured effective use of limited resources.

Collection of information to describe urinary tract infections

The primary outcome was derived using information collected from the sources at the time points described in Table 3 and was determined using the algorithm described in Appendix 5.

Participants met the definition of the primary outcome if they fulfilled the criteria described in Table 4. This is expanded in Appendix 5.

General practitioner confirmation of antibiotic prescription

Where participants reported antibiotic use at 1, 2 or 6 weeks or failed to return any trial paperwork at these time points, confirmation of these details, or a request for this information, was sent to the participant’s GP. This included a brief letter explaining the need for this information and a table for the GP to complete asking whether or not the participant had presented to them with a UTI during the period of their participation in the trial, and if so whether or not they had given the participant a prescription for antibiotics for UTI (see Appendix 4).

Change of status/withdrawal

The status of some participants changed during the trial for a number of reasons. These included post-randomisation exclusions; participants deciding they no longer wished to be a part of the trial; and decisions by medical staff that it was not appropriate for the participant to remain in the trial.

Participants were free to decline further follow-up from the trial at any point without giving a reason. Participants could also be withdrawn for medical reasons. In such cases, primary outcome data were still collected if the participant consented to the use of their relevant hospital and general practice records.

In addition, some patients were classed as post-randomisation exclusions for one of the following reasons:

• patients who were randomised but received a suprapubic catheter

• patients who were randomised but did not have any catheter inserted

• emergency patients who were randomised but subsequently declined to participate in the trial.

The justification for excluding participants who did not receive a urethral catheter from the analysis was that they did not fulfil intention-to-treat criteria, as the ultimate decision-maker, the responsible clinician in the operating suite, determined that an alternative urine drainage option was preferred.

Data management

Data collected at site were input into the electronic CATHETER database through the trial web portal (www.charttrials.abdn.ac.uk/catheter) by the local research team; those received by the trial office were entered by the trial office staff.

At the end of the trial, a random 10% sample of all of the trial data were re-entered by the trial office to verify correct data input. Any discrepancies between originally entered data and re-entered data were reviewed and checked against the original paper copy by an individual who was not involved in entering either data set. Incorrectly entered data were corrected at the time of checking. An initial data entry error rate of > 5% would have triggered a requirement to re-enter the entire data set from that questionnaire. This was not found to be necessary.

Trial oversight committees

A Trial Steering Committee (TSC), consisting of an independent chairperson, two further independent members and the grant holders, provided oversight of the trial. The TSC met six times over the course of the trial (at least annually).

The independent Data Monitoring Committee (DMC) met early in the trial and agreed its terms of reference and other procedures. They then met six times over the course of the trial (at least annually). The DMC reported any recommendations to the chairperson of the TSC.

Important changes to methods after trial commencement

Timing and frequency of analyses

The data monitoring committee considered confidential interim inspection of the data on five occasions. At the first meeting, the DMC recommended refining the definition of algorithm used to define the primary outcome; this was due to a higher than anticipated rate of UTI (for further details, see Chapter 3, Important changes to methods after trial commencement).

Planned secondary subgroup analyses and sensitivity analyses

Subgroup analyses of the primary outcome examined possible effect modification of the following:

• age (< 60 years, ≥ 60 years)

• sex

• comorbidity (pre-existing urological disease, diabetes, immune suppression)

• indication for catheterisation (incontinence, urine retention and monitoring purpose)

• antibiotic use prior to randomisation

• duration of catheterisation.

Modification of the treatment effect was explored using tests for interaction (all at stricter levels of significance; p < 0.01), and results are presented as forest plots with 99% CIs to reflect the exploratory nature of these analyses. Two further post hoc effect modification sensitivity analyses were carried out treating (1) age and (2) duration of catheterisation as continuous variable interactions with treatment allocation to explore any potential differential effects across catheters. All subgroup and treatment effect modification analyses were analysed using the same generalised linear modelling framework as the main analyses. A further sensitivity analysis was carried out to assess any potential impact of centre on the precision of estimates of treatment effects by including a random effect for centre in the model for the primary outcome.

Introduction

Two types of economic analyses were planned. The first was a ‘within-trial’, economic evaluation undertaken using data collected as part of the trial, and the second was based on a modelling exercise aiming to address the uncertainty (background noise) introduced by the heterogeneity in terms of underlying illness and type of operative procedure undergone by trial participants. The question addressed by both economic evaluations was: what is the cost-effectiveness of antimicrobial-impregnated (nitrofurazone) or antiseptic-coated (silver alloy) catheter versus standard PTFE-coated catheter? The methods and analysis for both the within-trial analysis and the modelling analysis are described below. The perspective of study was that of the UK’s NHS.

Within-trial analysis

Derivation of costs

Unit costs were based on study-specific estimates and data from standard sources. A summary of unit costs is presented in Table 6. Unit costs for the interventions were obtained from the manufacturers of the products through personal communication or from published price lists. The unit cost per day in hospital for each level of care was obtained from Information Services Division (ISD) of NHS Scotland. 71 This source does not give a cost per inpatient-day for all hospital services but rather presents data as a total cost per average case for each clinical specialty along with an average length of stay. We used these data to calculate ‘cost per day’ for each level of care. The total cost per case for each specialty provided by ISD includes both theatre costs and allocated costs (representing the costs of running the hospital). The former were omitted from our analyses, as we were not concerned with the costs of surgery undergone by trial participants (apart from the catheter unit cost) and we omitted the infrastructure costs, as they represent a fixed cost that will not vary with length of stay. The unit cost per case with these two elements removed was divided by the trial estimate of average length of stay for each specialty to give a cost per day. For example, the total gross cost per case in urology given by ISD was £2019 and the direct theatre cost per case was £479 (22%). We estimated the unit cost per stay by taking the theatre cost per case away from the unit cost per stay and then removing the allocated costs. This figure was then divided by the trial estimate of average length of stay for this specialty (3.3 days) to give the trial estimate of cost per day of £321 for participants treated under this specialty group.

Total patient NHS costs were derived by combining information on resource use with information on the unit costs of those resources. For each area of resource use, estimates of resource utilisation were combined with unit costs to derive total costs for each item of resource use for each patient. Averages were then calculated to give estimates for each item of resource use. The patient-level costs for each item of resource were summed to produce a total cost for each patient and allow an estimate of average total cost per patient to be calculated.

Effectiveness outcomes for cost-effectiveness analysis

Effectiveness was measured in terms of number of catheter-associated symptomatic UTIs treated with antibiotics up to 6 weeks after randomisation (trial primary outcome) and QALYs at 6 weeks.

Quality-adjusted life-years were derived using data from participant completion of the EQ-5D, a generic health status measurement tool, at baseline, 3 days after catheter removal, 1 and 2 weeks after catheter removal and at 6 weeks post randomisation as part of the main study questionnaires. The EQ-5D measure divides health status into five dimensions (mobility, self-care, usual activities, pain/discomfort and anxiety/depression). Each of these dimensions can have three levels, giving 243 possible health states. 72 These responses were converted into health–state utilities using UK population tariffs. 73 The utility scores were used to estimate the mean QALY score for each of the three trial groups. The estimation of QALYs took into account the death of any study participants with allocation of a utility score of zero from date of death to the date of 6 weeks after randomisation for participants who died during their involvement in the trial. QALYs were estimated using linear extrapolation between the QALY scores at baseline and all available EQ-5D data.

Incremental cost per infection avoided and quality-adjusted life-year gained

Average costs per patient for each intervention were calculated and the described comparisons were made. Similarly, the average number of QALYs were calculated for each intervention and compared with the control. Data collected on costs and effects of the interventions were combined to obtain an incremental cost-effectiveness ratio (ICER). This was performed by dividing the mean difference in costs by the difference in effect between the interventions and control group. This provides the incremental cost per infection avoided or incremental cost per additional QALY gained for the new interventions relative to standard practice, i.e. ΔC/ΔE = ICER (where C = cumulative costs at 6 weeks and E = cumulative effects over 6 weeks).

Measures of variance for these costs, infections and QALYs were derived using bootstrapping. 74 From the results of the bootstrapping, cost-effectiveness acceptability curves (CEACs) were created. CEACs are used to represent whether or not the two novel interventions are cost-effective at various threshold values for society’s willingness to pay for an infection avoided or additional QALY. CEACs present results when the analysis follows a net benefit approach. This approach utilises a straightforward rearrangement of the cost-effectiveness decision rule used when calculating ICERs (see below) to create the net monetary benefit (NMB). The NMB of the interventions in question is as shown below:

where λ represents the decision-maker’s willingness to pay for an infection avoided or for a QALY gained. If the above expression holds true, the intervention is considered cost-effective. As society’s willingness to pay is unknown, the NMB will be calculated for a number of possible λ values, including the threshold value of £20,000–30,000 for a QALY that is often adopted by policy-makers within the NHS. 75 The estimates of NMB at various threshold values for society’s willingness to pay for a unit of outcome are used to produce graphical and tabular representations of the CEAC.

Data analysis

As trial data were collected over a 6-week period no discounting was carried out. The number of missing data for variables used in the cost analysis was low, and data that were missing were therefore considered to be missing completely at random.

Patient flow

The flow of participants through the trial is summarised in Figure 5, in line with recommendations of the Consolidated Standards of Reporting Trials (CONSORT) statement. 77 Included in the total of 7102 randomised participants were 905 participants randomised following unplanned catheterisations, of whom 385 subsequently refused consent to be involved in the trial and were thereby excluded. A further 45 participants withdrew consent before catheterisation and were excluded from the trial. In total, there were 708 post-randomisation exclusions, equally distributed throughout the groups, for reasons such as use of a suprapubic catheter or no catheterisation (e.g. due to cancellation of procedures or catheterisation not being deemed necessary by treating clinical staff), as well as the 385 who refused consent following unplanned catheterisation (Table 8); this reduced the number of participants included in the trial to 6394 – 90% of those randomised.

FIGURE 5.

Consort diagram. a, Generally participants underwent catheterisation in the operating suite. In 272 (4.3%) of cases an alternative catheter to that allocated by randomisation was inserted mainly due to error by clinical staff. b, We achieved verification of the primary outcome for all participants. A total of 907 (14%) participants did not complete the planned 6-week post-randomisation follow-up (see Table 12).

Baseline characteristics

As one would expect with a large sample size, the randomised groups were well balanced at baseline across all measured covariates (Table 9). Participants were drawn from a wide age range (17–92 years) with a median of 61 years. Women accounted for 62% of the trial population. Subgroups were pre-specified to identify those at a greater risk of developing a UTI. A total of 10% of participants had pre-existing urological disease or diabetes and 7% were receiving immunosuppressive therapy. In the baseline urine sample taken prior to catheterisation (MSSU 84%) or at the time of catheterisation (CSU 16%), 8% of participants had ≥ 10⁴ CFU/ml and 27% had a pyuria with white blood cell (WBC) count of > 10/mm³ but were asymptomatic; 18% of participants had received antibiotics in the 7 days prior to randomisation for a variety of indications (Table 10).

Intervention received and descriptors of care

Table 10 contains details on the intervention received and aspects of in-hospital care. The main reason for catheterisation was for monitoring purposes during and after surgery. Duration of catheterisation and length of hospital stay were similar across all three groups, as was the proportion of participants receiving antibiotics for surgical prophylaxis.

Over 93% of participants in each arm received the catheter to which they were allocated. In the majority of cases where the allocated catheter was not used, this was due to theatre staff inserting a standard PTFE catheter in error. Analyses were by catheter allocated (intention to treat).

Descriptive data for duration of catheterisation showed a mean of 2.78 days in the nitrofurazone group, 2.95 days in the silver alloy group and 2.85 days in the PTFE control group, whereas the median was the same in all groups at 2 days. A total of 219 (3.43%) participants experienced a duration of catheterisation of > 14 days with 79 (3.67%) in the nitrofurazone group, 73 (3.48%) in the silver alloy group and 67 (3.13%) in the control group. Hospital stay varied widely, with a markedly skewed distribution; the median was 6 days in each group, but the upper extreme of the range varied, with the longest stay being 159 days (see Table 10).

Allocation of participants to the three trial groups was equal across all centres (Table 11).

Response rates

Participant response rates are shown in Table 12. Response rates were similar across the three trial arms, with overall response rates of 88% for the 3 days post catheter removal questionnaire, 72% and 71% for the 1 and 2 weeks post catheter removal diaries and 84% for the questionnaire at 6 weeks post randomisation.

Denominators reflect the number of participants that were included in the analysis for specific aspects of the data and reflect the response to the individual question.

Incidence of bacteriuria at, or within, 3 days of catheter removal

The rate of bacteriuria at 3 days post catheterisation was 13.5% (249/1846) in the nitrofurazone group, 17.4% (310/1785) in the silver alloy group and 17.5% (321/1839) in the PTFE control catheter group. Estimates of ORs and CIs from adjusted and unadjusted logistic regression models are presented in Table 16. The estimated absolute risk reduction between the nitrofurazone and PTFE catheters was 3.5% (97.5% CI 0.8 to 6.1). Between the silver alloy and PTFE this was estimated as an increase of 0.4% (97.5% CI –3.2 to 2.4).

Symptomatic urinary tract infection at 3 days of catheter removal

The rate of symptomatic UTI with a prescription of antibiotics at 3 days post catheter removal was 4.9% (106/2153) in the nitrofurazone group, 6.6% (137/2079) in the silver alloy group and 5.9% (127/2144) in the PTFE control catheter group. Estimates of ORs and CIs from adjusted and unadjusted logistic regression models are presented in Table 17 below. The estimated absolute risk reduction between the nitrofurazone and PTFE catheters was 1.0% (97.5% CI –0.6 to 2.5). Between the silver alloy and PTFE this was estimated as an increase of 0.6% (97.5% CI 1.1 to –2.3). Table 17 contains the estimated ORs and 97.5% CIs from unadjusted and adjusted logistic regression models.

Three-day symptom score

The three-day symptom score was the sum of ratings from ‘0’ (no symptoms) to ‘3’ (severe symptoms) for seven symptoms of UTI. The minimum score was therefore ‘0’ and the highest score was ‘21’, with a higher score indicating more severe symptoms. The proportion of participants within each study group who completed this score was similar, averaging 87%.

There was no statistically significant difference between either nitrofurazone or silver alloy catheters compared with control in terms of self-reported symptoms at 3 days (Table 18). A sensitivity analysis using multiple imputations did not change these results.

Catheter discomfort

There were no significant differences in self-reported discomfort between each of the interventional catheters and the control catheter reported at the stage of catheter insertion (Table 19).

Participants randomised to nitrofurazone catheters were 5.3% (97.5% CI 2.1% to 8.4%) less likely to report no discomfort during the period of catheterisation (Table 20). Those randomised to silver catheters were more likely to report no discomfort: 3.3% (97.5% CI 0.3% to 6.3%).

At catheter removal, participants in the nitrofurazone group were 12.3% (8.8% to 15.8%) less likely than participants in the PTFE group to report having no discomfort (Table 21). They were more likely to rate discomfort during catheter removal as mild, moderate or severe. There was no significant difference in discomfort during catheter removal between the silver alloy and PTFE catheters, 2.3% (97.5% CI –1.1% to 5.6%).

There were no significant differences between interventional and control catheter groups in self-reported discomfort in the period after the catheter was removed (Table 22). Multiple imputation of missing responses did not change any results of analyses of self-reported discomfort data.

Secondary economic outcome measures

Quality of life

Health-related QoL was measured using the EQ-5D. Response rates were low, at 63%, for diaries completed 1 week after catheter removal, which participants took home from hospital to return by post. However, response rates improved at the 6-week post-randomisation time point, at which 80% of trial participants returned the questionnaire containing the EQ-5D questions.

Table 23 and Figure 6 show that EQ-5D scores followed the same pattern in each group, with no evidence of a difference between the groups. Incorporating a time by treatment interaction showed that treatment effects were not different over time and these terms were dropped from the model for ease of interpretation. The mean (97.5% CI; p-value) difference in EQ-5D score between the nitrofurazone and PTFE groups was –0.001 (97.5% CI –0.022 to 0.003; 0.15), and for the silver alloy group compared with the PTFE group was –0.012 (97.5% CI –0.025 to 0.001; 0.07).

TABLE 23 - European Quality of Life-5 Dimensions scores over time

Cell contents are valid n, mean (SD); higher score is better.

Follow-up time point	Nitrofurazone (n = 2153)	Silver alloy (n = 2097)	PTFE (n = 2144)
Baseline	2127, 0.717 (0.292)	2076, 0.723 (0.291)	2123, 0.722 (0.299)
3 days	1860, 0.592 (0.274)	1801, 0.579 (0.281)	1871, 0.593 (0.271)
1 week	1363, 0.619 (0.272)	1308, 0.602 (0.292)	1366, 0.614 (0.270)
2 weeks	1405, 0.696 (0.259)	1328, 0.687 (0.270)	1398, 0.695 (0.248)
6 weeks	1705, 0.777 (0.243)	1665, 0.783 (0.240)	1721, 0.795 (0.234)

FIGURE 6.

Plot of EQ-5D scores over time. Bars are mean ± 1 SD. Note that the relative timing of responses compared with baseline varied for the 3-day, and 1- and 2-week time points and labels are notional only, for illustration.

The EQ-5D visual analogue scale (VAS) followed an almost identical pattern (Table 24 and Figure 7).

TABLE 24 - European Quality of Life-5 Dimensions VAS

Cell contents are valid n, mean (SD); higher score is better.

Follow-up time point	Nitrofurazone (n = 2153)	Silver alloy (n = 2097)	PTFE (n = 2144)
Baseline	2132, 70.46 (21.00)	2118, 70.94 (20.15)	2076, 70.61 (21.30)
3 days	1866, 65.09 (18.59)	1881, 65.20 (18.40)	1810, 64.94 (18.60)
1 week	1389, 68.80 (17.48)	1366, 68.06 (18.26)	1317, 68.21 (18.20)
2 weeks	1416, 75.01 (17.12)	1403, 74.53 (18.05)	1338, 74.62 (17.40)
6 weeks	1761, 78.47 (17.37)	1768, 78.30 (16.95)	1717, 78.53 (17.25)

FIGURE 7.

Plot of EQ-5D-VAS over time. Bars are mean ± 1 SD. Note that relative timing of responses compared with baseline varied for the 3-day, and 1- and 2-week time points and labels are notional only, for illustration.

Significant clinical events

Significant clinical events are presented in Table 25. There were two cases of septicaemia in the trial; neither was due to catheterisation, with both being due to infections acquired from the participants’ intravenous line.

Subgroup analysis

The following subgroups analyses were performed to establish whether or not any risk or protective factors influenced rates of UTI in the trial groups:

• sex

• older age

• comorbidity (immune suppression)

• indication for catheterisation

• antibiotic use prior to randomisation.

For most outcomes there was no apparent differential treatment effect between subgroups (Figures 8 and 9). The only potential exception to this was antibiotic use in the 7 days prior to randomisation, where there was marginal evidence that participants subsequently randomised to nitrofurazone who had received antibiotics in the previous 7 days had a greater incidence of UTI at any time up to 6 weeks after randomisation.

FIGURE 8.

Forest plot of subgroup analyses, UTI at any point up to 6 weeks post randomisation for nitrofurazone vs PTFE comparison.

FIGURE 9.

Forest plot of subgroup analyses, UTI at any point up to 6 weeks post randomisation for silver alloy vs PTFE comparison.

It is clear from Figure 10 the probability of suffering a catheter-associated UTI increased with age. The interaction between age and catheter type was not significant; there was no evidence that treatment effects were modified by age. The p-value for age-by-nitrofurazone and age-by-silver was 0.53 and 0.65, respectively.

FIGURE 10.

Plot of predicted probability of UTI at any point up to 6 weeks post randomisation by age (in years) and allocated catheter.

As expected the probability of UTI increased with duration of catheterisation (Figure 11). The interaction between duration and catheter type was not significant; there was no evidence that treatment effects were modified by duration of catheter. The p-value for duration-by-nitrofurazone and duration-by-silver was 0.19 and 0.87, respectively.

FIGURE 11.

Plot of predicted probability of UTI at any point up to 6 weeks post randomisation by duration of catheterisation (in days) and allocated catheter.

Centre effect

Adjusting for centre by including a random effect for centre (Table 26, third column) had negligible effect on fixed-treatment-effect estimates (see Table 26, second column) and CIs for both the marginal adjusted and unadjusted models.

Estimation of NHS costs

Table 28 is a conversion of Table 27 into costs to the NHS. As indicated by the SD, the cost data were highly skewed to the right; while most of the participants incurred low costs, some of them incurred very high costs. In terms of NHS costs incurred after trial participants were catheterised, the mean total cost per patient was £3259 (SD £3152) in the nitrofurazone group, £3438 (SD £3270) in the silver alloy group and £3390 (SD £3405) in the PTFE group. The main driver of total costs was length of stay. The cost of catheters themselves did differ (although the use of catheters was very similar) but the magnitude of differences was comparatively small. The reason for this was that the unit costs were higher for nitrofurazone and silver alloy catheters than for PTFE catheters.

Base-case analysis based on risk of infection as defined for the primary trial outcome

For the model-based analysis data inputs to the model relate to:

• The absolute risk of infection for participants randomised to PTFE (control).

• The absolute risk difference for infection for nitrofurazone compared with PTFE and silver alloy compared with PTFE.

• The cost of care over the 6-week study period for those participants who did not suffer a UTI.

• The difference in costs over the 6-week study period for those who suffered an infection compared with those who did not.

• The number of catheters used per participant and the unit cost of each type of catheter.

• The estimate of QALYs for the 6-week study period for those who did not suffer an infection.

• The difference in QALYs for the 6-week study period for those who suffered an infection compared with those who did not.

Data on the risk of infection for PTFE and the absolute risk differences were based on trial primary outcomes results reported in Chapter 5. Tables 30 and 31 report the NHS resource use and NHS costs for those who have a UTI and those who did not. As expected, the results of the comparison showed that participants who suffered a UTI had statistically significantly higher resource use (see Table 30) and hence costs (see Table 31). The main driver of the difference in total costs (excluding the unit cost of the catheter) appeared to be differences in length of stay. The total costs when the cost of the catheter was excluded for those both with and without a UTI were used to populate the economic model. It was also assumed that for each arm of the model (each arm representing one of the three catheters) one catheter per person would be used.

Table 32 reports the EQ-5D scores for those who suffered a UTI and those who did not. From these data it was estimated that the mean QALYs were 0.075 (SD 0.03) for the group that suffered a UTI and 0.081 (SD 0.02) for the group that did not. The mean difference in QALYs after adjusting for baseline EQ-5D and other characteristics was 0.006 (97.5% CI –0.009 to –0.003) lower for the participants who suffered a UTI.

A summary of the parameters, their values, source and the distribution used in the probabilistic sensitivity analysis is reported in Table 33.

The results of the model using the absolute risk reduction recorded as the primary outcome for trial data suggest that nitrofurazone is, on average, the least costly trial intervention, followed by PTFE and then silver alloy (Table 34). It is also, on average, the most effective catheter, with PTFE and silver alloy having similar effectiveness. This latter finding was expected, given the very small mean difference in risk of an infection between silver alloy and PTFE catheters. Overall, there is a 70% chance that nitrofurazone would be the least costly option and over an 80% probability that it would be considered cost-effective when society is willing to pay a maximum of £30,000 per QALY. Silver alloy has virtually no chance of being considered cost-effective when compared with the other two catheters. For nitrofurazone the reduction in the subsequent cost of care caused by a likely reduction in the risk of infections more than compensates for the increased cost of the catheter. However, the same is not true for silver alloy.

What the probabilistic analysis reported in Table 34 does not portray is the magnitude and variation in cost. The difference in cost between nitrofurazone and PTFE is graphically illustrated in Figure 12 and that for silver alloy and PTFE in Figure 13. For the comparison of nitrofurazone against PTFE, the 97.5% CI was –£36.19 to £11.45 and for the comparison of silver alloy against PTFE the 97.5% CI was £4.13 to £5.92.

FIGURE 12.

Distribution of health-care cost differences for participants randomised to nitrofurazone compared with PTFE. Based on 1000 Monte Carlo simulations.

FIGURE 13.

Distribution of health-care cost differences for participants randomised to silver alloy compared with PTFE. Based on 1000 Monte Carlo simulations.

Analyses for more ‘homogeneous’ patient groups

Three participant groups believed to represent more homogeneous patient populations were defined. These were:

• Those admitted to an obstetrics and gynaecology specialty ward only. For this analysis the costs and QALY for those with and without a UTI were based on this subgroup. We used the same risk of a UTI as in the base case, as these data were assumed to be robust and precise.

• Those with an EQ-5D score of 1 at 3 days after catheter removal. Again the costs and QALY for those with and without a UTI were based on this subgroup and the risk of a UTI was taken to be the same as the base case.

• Those participants who were recorded as having a symptomatic CAUTI treated with antibiotics at 3 days post catheter removal. Costs and QALY risks were defined in the same way as for the two previous subgroup analyses.

Tables 35–37 describe the costs and QALYs for each of these analyses. The first subgroup considered consisted of women who were admitted to an obstetrics and gynaecology specialty ward (n = 1736; UTI 336; no UTI 1400). For this subgroup, the adjusted difference in total costs between those who had an infection and those who did not was £141 (97.5% CI –117 to 400) and the QALY difference was –0.007 (97.5% CI –0.011 to –0.003) (see Table 35). The second subgroup we analysed, based on data from participants who reported that they had an EQ-5D score of ‘1’ (full health) at 3 days post catheter removal (n = 436; UTI 33; no UTI 403), resulted in a £988 (97.5% CI –64 to 2052) difference in costs and –0.002 (97.5% CI –0.012 to 0.007) difference in QALYs (see Table 36). Finally, Table 37 describes costs and QALYs for all trial participants, based on their infection status using the secondary trial outcome of symptomatic antibiotic-treated UTI at 3 days post catheter removal (n = 6394; UTI 370; no UTI 6023). Participants who suffered a UTI had, on average, £1417 higher cost (97.5% CI 982 to 1925) and –0.004 fewer QALYs (97.5% CI –0.009 to 0.000) compared with those who did not suffer a UTI.

TABLE 35 - Costs (£ sterling) from results of exploratory analysis based on participants who were treated in obstetrics and gynaecology specialty area

All differences adjusted for sex, age, reason for catheterisation, comorbidities, antibiotic use at 7 days, antibiotic use at catheterisation and CI based on bootstrapped data.

Resource type	UTI (n), mean (SD)	No UTI (n), mean (SD)	Adjusted difference,a mean (97.5% CI)	Unadjusted difference, mean (97.5% CI)
Intervention
Catheter	336, 4.21 (3.03)	1400, 4.15 (3.01)	–0.01 (–0.43 to 0.41)	0.06 (–0.34 to 0.48 )
Primary care
GP doctor visit	273, 31.78 (4.81)	1077, 3.07(12.81)	29.31 (24.28 to 34.35 )	28.71 (25.76 to 31.65 )
GP nurse visit	285, 1.37 (4.81)	1085, 0.14 (1.89)	1.24 (0.55 to 1.94 )	1.23 (0.82 to 1.64 )
Secondary care
Length of stay	335, 1958.21 (2003.18)	1370, 1805.98 (1870.84)	57.49 (–190.71 to 305.68 )	152.24 (–107.18 to 411.66)
Outpatient visit	287, 8.18 (33.73)	1088, 1.55 (14.45)	6.79 (2.20 to 11.37 )	6.63 (3.65 to 9.62 )
Visit to other providers	288, 5.07 (24.41)	1078, 0.30 (8.66)	4.91 (1.56 to 8.25 )	4.77 (2.75 to 6.79 )
Inpatient readmissions	286, 41.28 (78.50)	1082, 2.12 (51.79)	40.43 (4.10 to 76.75 )	39.16 (18.98 to 59.35 )
Medications
Antibiotics	311, 3.58 (2.56)	765, 0.63 (1.74)	2.95 (2.59 to 3.31 )	2.95 (2.65 to 3.26 )
Total	336, 2033.38 (2084.74)	1400, 1777.32 (1871.29)	141.46 (–116.76 to 399.68)	256.06 (–4.83 to 516.96)
Total without cost of catheter (used in model)	336, 2029.17 (2084.76)	1375, 1805.41 (1872.72)	127.93 (–136.78 to 392.64)	233.76 (–37.3 to 485.36)
QALYs	219, 0.080 (0.02)	832, 0.086 (0.02)	–0.007 (–0.011 to –0.003 )	–0.006 (–0.010 to –0.003)

TABLE 36 - Costs (£ sterling) from exploratory analysis based on participants reporting an EQ-5D score of ‘1’ (full health) at 3 days

All differences adjusted for sex, age, reason for catheterisation, comorbidities, antibiotic use at 7 days, antibiotic use at catheterisation and CI based on bootstrapped data.

Resource type	UTI (n), mean (SD)	No UTI (n), mean (SD)	Adjusted difference,a mean (97.5% CI)	Unadjusted difference, mean (97.5% CI)
Intervention
Catheter	33, 4.20 (2.61)	403, 4.20 (2.43)	–0.30 (–1.40 to 0.80)	0.01 (–0.99 to 1.00)
Primary care
GP doctor visit	27, 20 (23.06)	333, 1.08 (6.15)	18.63 (8.41 to 8.85)	19.92 (15.05 to 22.79)
GP nurse visit	28, 2.14 (5.68)	332, 0.27 (3.05)	1.96 (–0.51 to 4.44)	1.87 (0.40 to 3.34)
Secondary care
Length of stay	33, 3154.36 (2946.53)	400, 2468.29 (2341.09)	912.73 (–164.41 to 1989.87)	686.07 (–288.09 to 1660.24)
Outpatient visit	30, 0 (0)	334, 0.844 (8.88)	1.00 (–2.40 to 0.39)	–0.84 (–4.50 to 2.81)
Visit to other providers	28, 0.10 (0.57)	333, 0.61 (6.76)	–0.26 (–0.99 to 0.48)	0.50 (–3.38 to 2.38)
Inpatient readmissions	30, 54.67 (299.42)	332, 0 (0)	57.19 (–38.88 to 153.25)	54.67 (18.20 to 91.13)
Medications
Antibiotics	31, 3.32 (2.68)	239, 0.14 (0.85)	3.14 (2.07 to 4.21)	3.18 (2.66 to 3.70)
Total	33, 3229.65 (2944.67)	403, 2456.51 (2342.01)	988.45 (–64.47 to 2051.50)	773.14 (–200.85 to 1747.14)
Total without cost of catheter (used in model)	33, 3225.45 (2944.27)	403, 2452.31 (2342.01)	988.75 (–64.01 to 2051.74)	773.15 (–200.86 to 1747.15)
QALYs	25, 0.099 (0.02)	252, 0.102 (0.02)	–0.002 (–0.012 to 0.007)	–0.003 (–0.011 to 0.004)

TABLE 37 - Cost results of exploratory analysis based on whether or not patient had a symptomatic antibiotic-treated UTI at 3 days post catheter removal

All differences adjusted for sex, age, reason for catheterisation, comorbidities, antibiotic use at 7 days, antibiotic use at catheterisation and CI based on bootstrapped data.

Type of sensitivity	UTI (n), mean (£)	No UTI (n), mean(£)	Differencea
Costs using 3-day post-catheter removal UTI outcomea	370, 4748.72	6023, 3331.24	1417.48 (981.60 to 1925.18)
QALYs	184, 0.076	3132, 0.081	–0.004 (–0.009 to 0.000)

Table 38 summarises the data used for each subgroup analysis, along with details of the distribution and the data used to define that distribution.

TABLE 38 - Cost and QALY values used for each subgroup analysisa

Details of the values and calculations used to derive the distributions are described in Appendix 9.

Variable name	Value	Source
Maternity specialty participants
Health-care costs for participants without reported UTI	£1805.41	Based on cost estimate reported in Table 35. Log-normal distribution: derived from adjusted mean = 1805.41 and median costs = 1383 derived from trial data
Cost difference between UTI and no UTI	£127.93	Based on the adjusted analysis results cost difference estimate in Table 35. Normal distribution: mean = 127.93, SD = 116.55
QALYs for participants who experienced UTI over 6-week trial period	0.080	Based on data reported in Table 35. Beta distribution: α and β derived from mean (0.07961) and SD (0.02061) of QALYs for a UTI
QALY difference UTI and no UTI	0.007	Based on the adjusted analysis difference in QALYs reported in Table 35. Normal distribution: SD = 0.0015
All other parameters	Same as base case
EQ-5D score at 3 days = 1
Health-care costs for participants without reported UTI	£2452.31	Based on cost estimate reported in Table 36. Log-normal distribution: derived from adjusted mean = 2452.31 and median costs = 1844 derived from trial data
Cost difference between UTI and no UTI	£988.75	Based on the adjusted analysis results cost difference estimate in Table 36. Normal distribution: mean = 988.75, SD = 471.97
QALYs for participants who experienced UTI over 6-week trial period	0.099	Based on data reported in Table 36. Beta distribution: α and β derived from mean (0.09909) and SD (0.01904) of QALYs for a UTI
QALY difference UTI and no UTI	0.002	Based on the adjusted analysis difference in QALYs reported in Table 36. Normal distribution: SD = 0.0050
All other parameters	Same as base case
At 3 days post catheterisation
Health-care costs for participants without reported UTI	£3331.25	Based on cost estimate reported in Table 37. Log-normal distribution: derived from adjusted mean = 3331.25 and median costs = 2317 derived from trial data
Cost difference between UTI and no UTI	£1417.48	Based on the adjusted analysis results cost difference estimate Table 37. Normal distribution: mean = 1417.48, SD = 209.10
QALYs for participants who experienced UTI over 6-week trial period	0.076	Based on data reported in Table 37. Beta distribution: α and β derived from mean (0.07553) and SD (0.02553) of QALYs for a UTI
QALY difference UTI and no UTI	0.004	Based on the adjusted analysis difference in QALYs reported in Table 36. Normal distribution: SD = 0.0021
All other parameters	Same as base case

Table 39 describes the results of each of these three subgroup analyses. Also shown in this table are the results of the base-case analysis to facilitate comparison. In the first subgroup analysis, considering those admitted to the obstetrics and gynaecology specialty ward, the cost associated with PTFE was on average the lowest and silver alloy catheters were on average the most costly (£5.47 more costly than PTFE). The most effective catheter was nitrofurazone. When society was unwilling to pay any price for additional QALYs, PTFE had almost an 80% chance of being considered cost-effective. This likelihood fell to 30% when the threshold value for society’s willingness to pay was £30,000. When society’s willingness to pay was £30,000, nitrofurazone had a 70% chance of being considered cost-effective; however, this result is driven by the small difference in QALYs gained over PTFE, which may not be of clinical significance.

TABLE 39 - Results of subgroup analysesa

Plots of cost and QALYs and CEACs are shown in Appendix 10.

Intervention	Cost (£)	Incremental cost (£)	QALY	Incremental QALY	ICER	Probability (%) of being cost-effective at different threshold values for society’s willingness to pay for an additional QALY:
						£0	£10,000	£20,000	£30,000	£50,000
Base-case analysis
Nitrofurazone	3438.4		0.08232			73	77	80	84	89
PTFE	3445.5	7.1	0.08218	–0.0001	Dominated	27	23	20	16	11
Silver alloy	3450.55	12.1	0.08219	0	Dominated	0	0	0	0	0
Participants admitted into the obstetric and gynaecology specialty ward
PTFE	1918.27		0.08712			77	60	41	29	15
Nitrofurazone	1920.01	1.74	0.08726	0.00015	£11,497	23	40	59	71	85
Silver alloy	1923.74	3.73	0.08712	–0.00014	Dominated	0	0	0	0	0
EQ-5D score at 3 days = 1 (full health)
Nitrofurazone	2561.8		0.10106			91	91	91	91	92
PTFE	2578.2	16.3	0.10098	–0.00008	Dominated	9	9	9	9	8
Silver alloy	2582.8	20.9	0.10098	–0.00007	Dominated	0	0	0	0	0
Three-day symptomatic antibiotic-treated UTI outcome
Nitrofurazone	3485.9		0.08118			97	97	97	97	97
PTFE	3511.3	25.3	0.08108	–0.00010	Dominated	3	3	3	3	3
Silver alloy	3515.5	29.5	0.08109	–0.00009	Dominated	0	0	0	0	0

When the analysis was restricted to those with an EQ-5D score of ‘1’ at 3 days, or based on UTI status at 3 days post catheter removal, the nitrofurazone-impregnated catheter was least costly and the silver alloy-coated catheter was most costly. Nitrofurazone was also associated with more QALYs gained on average. In both of these analyses, nitrofurazone had a greater than 90% chance of being considered cost-effective when society was unwilling to pay any price for additional QALYs (willingness-to-pay value = 0). The likelihood of nitrofurazone being cost-effective increased as society’s threshold value for a QALY increased, although it should be noted that any QALY gains from the use of nitrofurazone over standard catheters were small and may not be clinically important.

For illustrative clarity, Figures 14 and 15 show the cost difference for nitrofurazone compared with PTFE and silver alloy compared with PTFE for those participants admitted to obstetrics and gynaecology ward. For this subgroup the 97.5% CI for nitrofurazone versus PTFE was –£6.65 to 7.59, and for silver alloy compared with PTFE it was £5.21 to £5.73.

FIGURE 14.

Distribution of health-care cost differences for participants from the obstetrics and gynaecology specialty subgroup randomised to nitrofurazone compared with PTFE. Based on 1000 Monte Carlo simulations.

FIGURE 15.

Distribution of health-care cost differences for participants from the obstetrics and gynaecology specialty subgroup randomised to silver alloy compared with PTFE. Based on 1000 Monte Carlo simulations.

Figures 16 and 17 show the cost differences for the subgroup of participants who had an EQ-5D score of ‘1’ at 3 days. For this subgroup the 97.5% CI for nitrofurazone compared with PTFE was –£61.42 to 7.09, and for silver alloy compared with PTFE it was £3.53 to £5.62.

FIGURE 16.

Distribution of health-care cost differences for participants from the subgroup with perfect health (EQ-5D score = 1) at 3 days post catheter removal randomised to nitrofurazone compared with PTFE. Based on 1000 Monte Carlo simulations.

FIGURE 17.

Distribution of health-care cost differences for participants from the subgroup with perfect health (EQ-5D score = 1) at 3 days post catheter removal randomised to silver alloy compared with PTFE. Based on 1000 Monte Carlo simulations.

Figures 18 and 19 show the cost differences when costs and QALYs are based on those with and without a UTI at 3 days post catheter removal. For this analysis, the 97.5% CI for nitrofurazone compared with PTFE was –£62.62 to £3.78, and for silver alloy compared with PTFE it was £3.71 to £4.66.

FIGURE 18.

Distribution of health-care cost differences for participants with the outcome of UTI at 3 days post catheter removal randomised to nitrofurazone compared with PTFE. Based on 1000 Monte Carlo simulations.

FIGURE 19.

Distribution of health-care cost differences for participants with the outcome of UTI at 3 days post catheter removal randomised to silver alloy compared with PTFE. Based on 1000 Monte Carlo simulations.

Other sensitivity analysis

As already noted, one driver of the cost differences between trial intervention groups and between participants who did or did not report UTI was the differences in length of stay, which may have been influenced by imbalances in care and illness factors between groups. In this sensitivity analysis the base-case analysis is repeated but with inpatient costs excluded in an attempt to neutralise the influence of initial inpatient treatment costs incurred during the period of catheterisation.

Table 40 describes the differences in costs and QALYs between those with and without a UTI once inpatient costs are excluded. Details of parameter values used in the model are described in Appendix 9, Tables 56–58, and the results of the analysis are shown in Table 41, below.

TABLE 40 - Costs (£ sterling) results of exploratory analysis based on whether or not participant had a UTI at 3 days post randomisation and excluding costs associated with the inpatient stay

All differences adjusted for sex, age, reason for catheterisation, comorbidities, antibiotic use at 7 days, antibiotic use at catheterisation and CI based on bootstrapped data.

Type of sensitivity	UTI (n), mean (£)	No UTI (n), mean (£)	Differencea
Costs excluding the length of stay	762, 81.54	5629, 13.30	67.41 (21.88 to 112.87)
QALYs	424, 0.077	2891, 0.083	–0.006 (–0.009 to –0.003)

TABLE 41 - Results of sensitivity analysesa

Plots of cost and QALYs and CEACs are shown in Appendix 10.

Intervention	Cost (£)	Incremental cost (£)	QALY	Incremental QALY	ICER	Probability (%) of being cost-effective at different threshold values for society’s willingness to pay for an additional QALY:
						£0	£10,000	£20,000	£30,000	£50,000
Base-case analysis
Nitrofurazone	3438.4		0.08232			73	77	80	84	89
PTFE	3445.5	7.1	0.08218	–0.0001	Dominated	27	23	20	16	15
Silver alloy	3450.55	12.1	0.08219	0	Dominated	0	0	0	0	0
No inpatient costs
PTFE	23.0		0.08102			100	93	69	47	23
Nitrofurazone	26.0	3.0	0.08113	0.00011	28,602	0	7	31	53	77
Silver alloy	28.6	2.5	0.08103	–0.00010	Dominated	0	0	0	0	0

The results of the analysis described in Table 41 show that when inpatient costs are excluded then the savings in health-care costs incurred after discharge from hospital are not large enough to compensate for the higher unit price of the nitrofurazone or silver alloy catheter. On average the nitrofurazone catheters are marginally more effective than the PTFE or silver alloy catheters but the difference in QALYs may not be of clinical significance. When society is unwilling to pay anything for an additional QALY then PTFE would have a probability of being considered cost-effective that approaches 100%. When society’s willingness to pay for a QALY approaches £30,000 then the likelihood that nitrofurazone would be considered cost-effective increases to 50%. It should be noted that this result arises from the assumption made in the analysis that any difference in QALYs, no matter how small, is important.

Figures 20 and 21 display the differences in costs more clearly. For this analysis the 97.5% CI for nitrofurazone compared with PTFE is £0.82 to £4.53 and for silver alloy compared with PTFE it is £5.49 to £5.57.

FIGURE 20.

Distribution of health-care cost (£ sterling) differences for participants randomised to nitrofurazone compared with PTFE and excluding costs of the initial episode of hospital stay. Based on 1000 Monte Carlo simulations.

FIGURE 21.

Distribution of health-care cost (£ sterling) differences for participants randomised to silver alloy compared with PTFE and excluding costs of the initial episode of hospital stay. Based on 1000 Monte Carlo simulations.

Nitrofurazone-impregnated catheter

The primary end point results of the trial are summarised in Table 13. The adjusted OR when comparing nitrofurazone with PTFE was 0.81, with 97.5% CI of 0.65 to 1.01. The best estimate of effect (0.81) is thus less than the clinical effect that was the basis of the sample size calculation (OR 0.67). The CI just includes the OR sought (lower boundary 0.65) and just crosses the point of ‘no difference’ (1.0). Our conclusion, therefore, is that nitrofurazone catheters do not provide the protective effect that we prespecified as clinically important. Putting this another way, the best estimate is that the use of a nitrofurazone catheter would prevent 1 UTI in every 48 people catheterised but that the ‘true’ figure lies anywhere between 1 in 24 people and no protective effect at all. Reflecting the considerations about costs above, the cost–utility analysis was more encouraging, suggesting that there was a 70% chance that the nitrofurazone catheter would be the least costly option, with an 80% chance of cost-effectiveness at a threshold of £20,000 per QALY.

Silver alloy-coated catheter

The primary end point results are again summarised in Table 13. The adjusted OR for the comparison of the silver alloy catheters compared with PTFE catheters was 0.96, with 97.5% CI of 0.78 to 1.19. The best estimate of effect 0.96 is thus close to no difference (1.0). The lower end of the CI (0.78) is well short of the pre-set difference considered to be clinically important (0.67) and hence can be said to rule out such a difference. Furthermore, the other end of the CI (1.19) is consistent with a substantially increased risk of UTI. To put this in terms of numbers needed to treat, the best estimate is that 1000 people would need to receive a silver alloy catheter to prevent one UTI, but that the true effect may be anywhere between 1 infection prevented in 42 people and 1 infection caused in 45 people. In the economic analysis, when compared with the other catheters, it was very unlikely that the silver alloy catheter could be cost-effective.

Effectiveness

The trial was designed at the outset to provide primary outcome information that could be directly used by NHS policy-makers to help decide whether or not antimicrobial catheters should be implemented as the standard for short-term catheterisation, predominantly following surgery or other interventional procedures. The resultant pragmatic trial design encompassed recruitment of a large sample of the relevant population across a representative spectrum of NHS hospitals and services. The only additional interventions to routine care for trial participants were random allocation of type of catheter, a urine sample at 3 days after catheter removal and trial questionnaires. The primary outcome, symptomatic UTI, defined as the presence of symptoms suggestive of UTI together with physician prescription of an antibiotic to treat UTI within 6 weeks of randomisation, was recorded by completion of case report forms during hospital stay, participant questionnaire after discharge and review of primary care records. These methods were highly effective in that all participant-reported prescriptions of antibiotic for UTI were confirmed from clinician record and that primary outcome data were obtained for all except one non-responder (for whom we assumed no UTI). We therefore believe that the results of this trial are indicative of the clinical effectiveness of the different antimicrobial catheters in preventing UTI. Estimates of QoL changes and, in the context of a large RCT, costs should also be representative. However, given the wide variety of health problems and interventions experienced by participants, some parameters, such as length of hospital stay, show such variability that chance differences in illness trajectory are likely to obscure any impact of urethral catheterisation.

A key criticism of previous studies has been the use of outcomes that are not directly relevant to patient experience or health-care costs, such as a microbiological finding of bacteriuria alone without taking account of patient-reported symptoms or prescription of antibiotic for UTI. 42 Indeed, the lack of clinically meaningful end points in all previous clinical trials of antimicrobial-coated catheters has hindered the assessment of clinical effectiveness and cost-effectiveness of such catheters. 23,42,58 There is some evidence that clinician-defined CAUTI may not align well with the presence of bacteriuria on microbiological culture. 10 To address this, we chose a definition of CAUTI that included both patient-reported symptoms and clinician action in terms of prescription of an appropriate antibiotic. This did conform to criterion 2g of the 2004 CDC definition of CAUTI, and was more relevant to both patient experience and the requirement for additional health-care resource use. 38 We chose a 6-week period of trial observation following randomisation, as this would generally encompass at least a 4-week period following catheter removal and capture events occurring following discharge from hospital; lack of community follow-up being a further criticism of previous trials. 42 Reassuringly, we found agreement in direction of effect between bacteriuria and clinical diagnosis of CAUTI with 13.5% of the nitrofurazone group and 17.4% in the silver alloy group having urinary bacterial counts of 10⁴ CFU/ml within 3 days of catheter removal compared with 17.5% of those receiving the standard catheter (see Table 16). In addition, when a concurrent finding of bacteriuria was used as an additional criterion to the primary outcome of symptomatic UTI, the relative risk reduction was maintained with rates of 3.2%, 5.0% and 4.6% seen in the nitrofurazone, silver alloy, and control groups, respectively (see Table 14). These additional data provide some evidence that microbiological and symptomatic infection were linked in the present trial. In summary, we feel that our primary outcome did successfully address the aim of this pragmatic trial and our use of other definitions of UTI allows comparison with previous studies. The differing overall incidence rates found for these outcomes suggest that they are capturing distinct diagnostic constructs that do, however, have considerable overlap in terms of attribution of each outcome to individual trial participants.

As expected for a RCT of this size, all groups were well balanced at baseline on predicted risk factors of CAUTI (female sex, older age, diabetes and pre-existing lower urinary tract dysfunction). Furthermore, clinician-driven variables (e.g. antibiotic use prior to catheterisation, duration of catheterisation29,35,78) were also balanced between groups. Regarding patient factors thought to confer increased risk, we did find higher CAUTI rates among women, older participants, and those who experienced longer duration of catheterisation, but not among those with comorbidities such as diabetes, urinary tract dysfunction or immune suppression. Our analyses for interaction did not provide any evidence that these risk factors differentially affected the incidence of UTI within the three trial groups. We acknowledge that there were many possibly relevant characteristics of participant care that we did not capture. The wide variety of surgical interventions received by participants for disparate conditions may have resulted in imbalance of uncollected characteristics, which may have particularly affected the small effect sizes recorded and may also have influenced the magnitude and distribution of key economic parameters such as length of stay.

For the purposes of this trial a key eligibility criterion was an intended duration of catheterisation of between 1 and 14 days. Recruitment was therefore concentrated on clinical areas admitting men and women for elective surgical or interventional procedures where temporary indwelling catheterisation was part of the standard care pathway, with few participants recruited from patients primarily admitted to medical or critical care areas. Similarly, we deliberately avoided recruiting patients from the urology wards, as such patients represent a group at high risk of developing CAUTI, and they are not representative of the majority of patients receiving short-term catheters in NHS hospitals. The problems associated with using urology patients in antimicrobial catheter trials have been well documented, the most important of which is the higher background bacteriuria rate, which may lead to an overestimation of the effect size of the catheters. 23,58 The prevalence of short-term catheterisation in different clinical areas is uncertain but a previous study did document that elective surgical areas predominate, accounting for 78% of cases. 10 In addition an increased rate of CAUTI has been documented in patients treated in medical areas, although it was unclear if this was adjusted for duration of catheterisation, which is likely to be longer for medical patients who are generally admitted as emergencies. 39

We were unable to carry out a planned subanalysis concerning participants who were catheterised as part of their admission to a critical care area. Many of our trial participants did have part of their hospital stay on a critical care ward, particularly those undergoing neurosurgery or cardiothoracic surgery. The duration of such stays was difficult to measure within the logistical constraints of such a large trial, as they varied from a few hours to a number of days. We therefore decided against capturing this information. Recruitment of patients undergoing unplanned catheterisation as part of urgent care of critical illness or trauma proved logistically difficult and resource intensive and therefore was not prioritised. In addition, the need for randomisation prior to obtaining participant consent, although approved by our Ethics Committee, was associated with a relatively high rate of subsequent refusal of consent.

Given the recruitment policy for our trial, the results may therefore not be generalisable to patients admitted as emergencies to medical, trauma or critical care areas, or to urology patients, as the baseline bacteriuria rate is likely to be higher owing to underlying disease, the average duration of catheterisation is likely to be longer and the range of infecting organisms may differ. In line with the recruitment policy, the average duration of catheterisation was short, being ≤ 3 days for 75% of participants, which is similar to that found in a large cohort study from the USA. 11 As expected, we found CAUTI rates to be higher in those catheterised for at least 4 days but the three trial groups were well balanced for duration of catheterisation, with all showing a median [interquartile range (IQR)] of 2 (1–3) days.

In theoretical terms, the clinical and bacteriological effectiveness of both technologies involves a balance between the duration of their antimicrobial effect in terms of profile of antimicrobial activity and the underlying risk of infection in each individual. All studies, including ours, document that the risk of infection increases with increasing duration of catheterisation. Previous reviews have suggested that the key potential benefit of such catheters will occur during short periods of catheterisation of up to 2–3 weeks. 30 This recommendation is in line with in vitro findings suggesting that the antimicrobial activity of nitrofurazone-impregnated catheters was limited to 5 days’ exposure and that for silver alloy hydrogel-coated catheters was limited to 1 day of exposure to common bacterial strains causing CAUTI. 79 Given that the antimicrobial activity for both devices would be predicted to be maximal within the first 2 weeks of catheterisation and that for the majority of hospitalised patients any planned urinary catheterisation is of short duration,11 our decision to restrict trial participation to those patients predicted to have a period of catheterisation of between 1 and 14 days appears appropriate and does not restrict valid conclusions to be drawn concerning the clinical effectiveness of either technology. A test subanalysis adjusting for interaction catheterisation confirmed that the incidence of UTI did increase with increasing duration but did not provide any evidence that either antimicrobial catheter had greater or lesser effectiveness with its continued use up to 14 days compared with control.

Another key possible confounding variable was the use of prophylactic antibiotics at the time of catheterisation, given principally to reduce infective complications of the particular surgery being carried out. Again, all groups were well balanced for this variable, with approximately 72% receiving prophylactic antibiotic, and CAUTI rates were similar irrespective of prophylactic antibiotic use. These findings were not altered when only participants receiving antibiotics active against common uropathogens were separately considered (data not shown).

In contrast to the majority of previous randomised trials concerning this technology, allocation concealment was managed using a computerised system that was remote from the users, thereby protecting against selection bias. 42 The slightly higher rates of participants receiving a different catheter to the one allocated seen among those randomised to nitrofurazone catheters (6.7%) and those randomised to silver alloy catheters (4.9%) compared with control (1.1%) are most likely to be related to PTFE being the standard widely available control catheter and hence this was more likely to be chosen as an alternative if there were difficulties inserting one or other of the antimicrobial catheters. All analyses were based on the intention-to-treat principle to guard against any bias that this might have introduced. The lack of difference in duration of catheterisation across the groups also suggests the inability to blind participants and clinicians to the allocated intervention did not result in a systematic bias in terms of one type tending to be removed earlier than another. In addition, the median (IQR) length of hospital stay was similar in all three groups at 6 (3–9) days. Follow-up was high across the groups, especially in terms of the primary outcome where antibiotic use was confirmed by the participant’s GP, who was likely to be unaware of the intervention received. Attrition throughout the trial (due to participants declining further follow-up or not responding to requests or due to intervening death) was low and there was no apparent differential loss to follow-up in the trial arms. In all other respects there was no evidence of bias. The pragmatic nature of our trial did have drawbacks. We did not monitor any other interventions that may have influenced CAUTI rates, such as methods of catheter insertion and catheter care violations,48 although efforts were made at all participating centres to emphasise the need for adherence to established best practice guidelines throughout the duration of the trial. To assess whether or not catheter care differed between centres, we did carry out a subanalysis adjusting for centre and found no difference in rates of infection between trial groups. There was variation in overall infection rates between centres, the reasons for which are likely to be multifactorial. Possible factors include differing specialties from which participants were recruited, variation in duration of catheterisation and variation in use of surgical antibiotic prophylaxis. Our pragmatic approach to trial design was intended to ensure that it was representative of current practice across the NHS, which enhances the generalisability of the trial results. It is also possible that adherence to emerging NHS-wide standards for catheter care may have increased over the relatively long duration of the trial as a consequence of the High Impact Actions initiative in NHS England. 28

We decided on the 6-week period of trial participation after review of the literature, internal discussion and consultation with external clinical and microbiologist experts. Our rationale was that we aimed to have outcome data collected for at least 4 weeks after catheter removal for all participants and therefore allowed for a 14-day duration of catheterisation and ensured that we had a certain starting time and date for each participant. This was longer than previous trials because, in line with the pragmatic design and anticipating the short hospital stay experienced by most participants, we captured all relevant events and included an ongoing patient and community health-care perspective. It is likely that a small number of participants suffered a community-acquired UTI at some time during the 6-week period of trial participation, although their recent catheterisation will have remained a risk factor. Given the large sample size and consequently well-matched baseline participant characteristics we cannot envisage that the rate of occurrence of community-acquired UTI subsequent to catheter removal differed between the trial groups.

Cost-effectiveness

Model-based analysis

As an alternative to a within-trial analysis, a model-based analysis was also planned primarily to focus more closely on the costs and QALY differences between those who suffered an infection and those who did not. The model assumed that differences in costs and QALYs between randomised groups were solely a consequence of potential differences in the incidence of UTI. The base-case analysis did not account for unobserved heterogeneity in uncontrolled characteristics between those who suffered an infection and those who did not. The implication of this is that the analysis may be confounded by the scenario that those participants who suffered an infection were more likely to (1) suffer an infection because of underlying ill health; (2) incur extra costs not only because they have an infection but also because of the increased severity of their underlying condition; and (3) lose QoL again not just because they have an infection but because of the increased severity of their underlying condition. In order to explore these possibilities, a number of sensitivity analyses were performed, considering comparisons between those who had a UTI and those who did not in a number of subgroups that were hypothesised post hoc to be more homogeneous than the whole trial population.

For the model-based analysis, the mean health-care cost differences between the different trial groups were considerably less than the mean cost differences estimated in the within-trial analysis, although the direction of the differences was unaltered. This was primarily because the costs in the model were determined by the modest differences in infection rates and the difference in costs between those participants who did or did not suffer a UTI. Again, differences in QALYs were very small and not statistically significant.

The sensitivity analyses were conducted using more homogeneous patient subsets defined post hoc. These included considering only patients admitted to obstetric and gynaecology specialty wards; using infection rates at 3 days after catheter removal; and including only those patients who had an EQ-5D score of ‘1’ at 3 days. For each of these three sensitivity analyses, differences in QALYs were very small, but for two analyses (infection rate at 3 days and EQ-5D = 1) the magnitude of the estimated cost of a UTI was greater than that observed in the analysis based on all participants and hence the cost saving for the nitrofurazone group was greater. For the third sensitivity analysis, both nitrofurazone and silver alloy catheters were, on average, slightly more costly than PTFE and there was a < 25% probability that nitrofurazone would be less costly than PTFE (the probability for silver alloy was approximately 0%). However, given the distributions associated with catheter infections, cost and QALYs there was an approximately 70% chance that nitrofurazone would be the most likely option to be considered cost-effective when the threshold value for society’s willingness to pay for a QALY was £30,000.

As already noted, one of the main components of cost was the inpatient stay. These data were highly skewed to the right, as they are naturally bounded by zero, but they have no logical upper bound and with such a large sample size there was likely to be a number of outliers with long stay and hence high cost. Hospital stay was truncated in line with the duration of trial participation at maximum of 6 weeks. This has the impact of omitting very long stays. A recent review of difficulties faced in handling skewed data concluded that most of the methods identified had undergone limited testing in different situations and their use in practice was very restricted; therefore, no detailed guidance could be provided. 80 The authors outlined three groups of methods, termed orbits, which had been previously used but all related to studies with small sample size. The simple truncation method we applied seems a credible option for the data produced by our trial.

During planning of the trial we elected to measure QoL changes by participant completion of a relevant questionnaire (EQ-5D) at baseline, 3 days after catheter removal, at 1 and 2 weeks after catheter removal, and at 6 weeks after randomisation. The collection of QoL measures at these time points would reflect the global changes incurred by each individual as a consequence of undergoing and recovering from specific health-care interventions. However, the number of different factors influencing each participant’s score makes determining the impact of a particular catheter on the risk of UTI and on QoL difficult to identify. Our sensitivity analyses, using homogeneous participant subsets defined post hoc, suggest that there would be a trend in favour of nitrofurazone because, based on the primary trial outcome, there was a trend towards reduced infections and lower QALYs among those who suffered a UTI. For the subgroup with a UTI 3 days post catheter removal the mean (97.5% CI) loss of QALYs was –0.004 (–0.009 to 0), for those who suffered a UTI but had a EQ-5D score of ‘1’ at 3 days post catheter removal it was –0.002 (–0.012 to 0.007) and for those admitted to obstetrics and gynaecology it was –0.007 (–0.011 to –0.003). It is likely that the pain and discomfort associated with catheter insertion or removal was not captured by the EQ-5D, as the EQ-5D elicits QoL on the day it is completed, potentially missing the impact of preceding events of short duration.

Alternative approaches to capture short but severe effects on QoL would have been to increase the frequency of QoL measurement or to ask respondents to complete QoL measures when events occurred. The practicalities of this in terms of administrative and respondent burden make it unfeasible. Furthermore, even if it were possible to ask those suffering a UTI or discomfort to complete a QoL measure, we would need some form of control to know what the QoL decrement should be compared with those who are not suffering a UTI or discomfort. An alternative approach would be to use some of the stated preference techniques used in economics, such as a time trade-off, standard gamble or contingent valuation method to elicit valuations for the different states of health that might exist. These approaches might value the profile of outcomes expected following the use of each of the different types of catheter or elicit the valuations of the presence or absence of specific events, such as UTI or discomfort of use. Such data could then be used in either a within-trial or a model-based analysis.

Estimation of NHS costs

Tables 45 and 46 are conversions of Tables 43 and 44 into costs to the NHS. As indicated by the SD, the cost data were highly skewed to the right; while most of the patients had low costs, some of them had very high costs. In terms of NHS costs (see Tables 45 and 46) incurred after the patients received the catheters, the mean total cost per patient in the nitrofurazone group was £3259 (SD £3152), the mean cost in the silver alloy group was £3438 (SD £3270) and the mean cost of the PTFE group was £3390 (SD £3405). There was, however, no evidence of statistically significant differences in the total mean costs or other costs, although the CIs are sufficiently wide to include economically important differences that may favour any of the catheters. The mean difference in costs per patient of catheters for the intervention groups compared with PTFE was higher for both: nitrofurazone (£4.19, 97.5% CI 4.11 to 4.26) and silver alloy (£5.39, 97.5% CI 5.30 to 5.49). Reflecting the findings described above, other costs tended to be lower in the nitrofurazone group, particularly differences in length of stay, which was the main driver of differences in mean costs between groups. The implications of differences in length of stay are explored later as part of the sensitivity analysis.

Incremental cost per infection averted

Taking the results of the primary outcome reported in detail in Table 13 in Chapter 5 and the differences in cost reported in Table 46, a three-way comparison of the different catheter groups was made (Table 49). On average, nitrofurazone was associated with lower costs, with a cost reduction of £108, and was more effective, being associated, on average, with 0.021 UTIs compared with PTFE. An incremental cost per infection avoided is not calculated in this circumstance, as, on average, nitrofurazone is less costly and more effective. On average, the care of participants in the silver alloy group cost £81 more than PTFE but the patients suffered 0.001 fewer UTIs. Therefore, the incremental cost per number of infections was £81,370 for silver alloy compared with PTFE.

The data presented above do not reflect the statistical imprecision surrounding estimates of costs and infections. Therefore, Tables 49 and 50 also show the probability that an intervention would be considered cost-effective at different threshold values for society’s willingness to pay to avoid an infection. These data show that nitrofurazone had an approximately 90% chance of being considered cost-effective at all willingness-to-pay thresholds considered. The main driver of these data is the trend towards lower costs in the nitrofurazone group compared with control (which, in turn, are driven by the trend towards a shorter length of stay). The probability of either silver alloy or PTFE being cost-effective over the range of threshold values considered was low (approximately 9% for PTFE and between 1% and 3% for silver alloy).

Sensitivity around cost per day

A sensitivity analysis around length of stay considered the use of alternative unit cost data. In this sensitivity analysis, elective inpatient excess bed-days based on Health Resource Group (HRG) data from the National Schedule of Reference Costs 2009–10 for NHS Trusts76 were used instead of the data from ISD. 71 As Table 53 illustrates, the direction of cost results was similar to that of the base-case analysis using NHS Scotland ISD data, although the magnitude in the differences was lower.

Use of unadjusted differences in costs and quality-adjusted life-years instead of the adjusted differences used in the base-case analysis

The base-case analysis was based on differences between different groups that were adjusted for age (< 60 years, ≥ 60 years), sex, comorbidity (pre-existing urological disease, diabetes, immunosuppression), duration of catheterisation (< 4 days, ≥ 4 days), indication for catheterisation (incontinence, urinary retention and monitoring purpose) and antibiotic use prior to enrolment. The results of the sensitivity analysis using unadjusted cost and QALY differences are presented in Table 54.

The unadjusted results of nitrofurazone compared with PTFE were similar to those obtained from the adjusted analysis. The mean difference in QALY was similar to that of the base case and again this difference was not statistically significant. Similarly, the cost difference was slightly higher than estimated in the base case; again, this difference was not statistically significant. The net effect of this was a small increase in the probability of nitrofurazone being considered cost-effective over the threshold values for society’s willingness to pay for an additional QALY considered (Figures 27 and 28).

FIGURE 27.

Representation of the uncertainty in differential mean costs and QALYs for nitrofurazone vs PTFE.

FIGURE 28.

Cost-effectiveness acceptability curve for nitrofurazone vs PTFE.

The unadjusted results of the silver alloy group compared with PTFE were similar to those of the adjusted analysis as the differences in QALYs and cost were smaller than reported for the base-case analysis. As a consequence this slightly increased the probability of silver alloy being considered cost-effective at all threshold values for society’s willingness to pay for a QALY considered (Figures 29 and 30).

FIGURE 29.

Representation of the uncertainty in differential mean costs and QALYs for silver alloy vs PTFE.

FIGURE 30.

Cost-effectiveness acceptability curve for silver alloy vs PTFE.

The CATHETER trial

Title of trial: types of urethral catheter for reducing symptomatic urinary tract infections in hospitalised adults requiring short-term catheterisation.

This protocol describes a major multicentre UK trial to establish whether using different types of urethral catheters, coated with antibiotic or antiseptic, in adults requiring short-term catheterisation can reduce symptomatic urinary tract infections. The study is designed to be as simple as possible for those participating and those involved in clinical care.

Recruitment officers in each centre will identify and recruit patients who require short-term catheterisation and collect patient information and urine samples. Patients will be followed up at 3 days and 1 and 2 weeks post catheter removal, and at 6 weeks following randomisation.

Full title of study

Types of urethral catheter for reducing symptomatic urinary tract infections in hospitalised adults requiring short-term catheterisation: multicentre randomised controlled trial of antibiotic and antiseptic impregnated urethral catheters (the CATHETER trial).

(Keywords: short-term urethral catheter, anti-microbial, silver alloy, nitrofurazone, silicone, randomised controlled trial, catheter-associated symptomatic urinary tract infection, bacteriuria, bacteraemia, cost–benefit analysis)

What is the problem to be addressed?

Development of symptomatic urinary tract infection in adults following short-term urethral catheterisation in hospitalised adults.

What are the principal research questions to be addressed?

What is the clinical benefit and cost-effectiveness of using antibiotic- or antiseptic-impregnated urethral catheters over standard urethral catheters in hospitalised adults requiring short-term catheterisation? Two pragmatic comparisons will be made comparing catheters, as they would be used in the NHS:

• antibiotic-impregnated (nitrofurazone) catheter versus ‘standard’ PTFE (PolyTetraFluoro-Ethylene)-coated latex catheter

• antiseptic-impregnated (silver alloy) catheter versus ‘standard’ PTFE (PolyTetraFluoro-Ethylene)-coated latex catheter.

The hypothesis being tested is that use of either of the impregnated catheters will reduce the incidence of catheter-associated symptomatic urinary tract infection by 40% relative to the standard PTFE coated latex catheter (an absolute reduction of around 3%).

Why is a trial required now and what is the available evidence?

25% of patients admitted to hospital will require urethral catheterisation at some stage during their stay1, and the risk of developing bacteriuria, the presence of bacteria in the urine, in catheterised patients is approximately 5% per day increase1. It has been estimated that symptomatic urinary tract infection occurs in approximately 20% of patients with bacteriuria2,3 whilst bacteraemia, the presence of bacteria in the blood, occurs in up to 4% of these patients4,5. Catheter-associated symptomatic urinary tract infections (CASUTI) are the leading cause of hospital acquired infections, accounting for between 23% and 40% of all cases6,7, and such infections result in additional morbidity3,5 and mortality8 and represent a considerable economic burden to the health-care sector, patients and their carers9,10. Consequently, any intervention that reduces the incidence of CASUTI may have wide-ranging repercussions. A wide variety of preventive approaches have been investigated, including mental care measures, pre-connected catheter-collection systems, antiseptic drainage bags, routine bladder irrigation, and prophylactic antibiotics, but the evidence in support of these measures is weak at best11. However, coating of urinary catheters with antibiotic or antiseptic compounds is thought to be a potentially effective preventative measure. Both, however, are more expensive than standard uncoated catheters.

A recent Cochrane review of randomised controlled trials12 concluded that the silver alloy impregnated catheter (an antiseptic impregnated catheter) has the most evidence of benefit out of the antibiotic/antiseptic impregnated urethral catheters available. However, the included trials were small and of poor or moderate quality. A reduction in risk of catheter-associated urinary tract infections by up to 40% was reported in hospitalised adults having a short-term catheter. The evidence for antibiotic-impregnated urethral catheters was even weaker, with only one type of catheter (i.e. minocycline and rifampicin impregnated catheter) from one trial being considered in the review12. However, minocycline and rifampicin impregnated catheters are no longer available (oral communication with Cook Urological, 03/10/05). Another recent small trial investigating a different type of antibiotic-impregnated catheter (nitrofurazone impregnated urethral catheter, marketed by Rochester Medical) revealed no evidence of a difference in the incidence of bacteriuria between the nitrofurazone impregnated and standard silicone catheters in patients catheterised for up to 1 week13, although confidence intervals were wide.

In summary, the majority of clinical trials conducted thus far have been small and of poor to moderate quality in terms of trial methodology and design, outcome measures (such as reporting of bacteriuria rather than symptomatic urinary tract infection) and the lack of a comprehensive evaluation.

How will the results of this trial be used?

The trial results will have implications for the management of patients requiring short-term urethral catheterisation in hospital and rationalise catheter-purchasing policies for large organisations like the NHS. If the use of an antibiotic or antiseptic impregnated urethral catheter leads to a significant reduction in CASUTI compared with a standard catheter and proves to be cost-effective, it will inform future short-term catheter policies in secondary care. In addition, the trial will also explore whether in high-risk sub groups more expensive catheters might be more likely to be cost-effective. The trial will clarify amongst currently available catheters, which should be used in the NHS.

Results of audit of short-term catheter policies

A one week audit was conducted in Aberdeen Royal Infirmary and Freeman Hospital (Newcastle) in October 2005. The purpose of the audit was to determine current short-term catheter policies across specialities and to identify high volume users. Elements of practice reviewed included the number of catheterisations performed, types of catheters used, indication for catheterisation, estimated duration of catheterisation, and use of prophylactic and concurrent antibiotic therapy. Data from hospital purchasing department records for each individual ward was also reviewed to determine general trends in number and type of catheters used over a 12-month period and to provide further supporting evidence of high volume users. There were 148 short-term urethral catheterisations reported during that week. PTFE-coated latex catheters were most commonly used (74%). Antibiotic prophylaxis was used in 39%, and 20% of patients were already on concurrent antibiotics at the time of catheterisation. Catheter-related urinary sepsis was documented in 29 cases (20%), of which 17 (11%) had culture-positive urinary infections.

Results of feasibility study of short-term catheter in secondary care

Following on from the audit we undertook a feasibility study. The aim of the 2 week feasibility study was to devise and test recruitment methods, baseline questionnaires, and to provide estimates of numbers likely to be eligible for the trial. The feasibility study aimed to replicate the main study protocol including randomisation and was undertaken in Aberdeen Royal Infirmary in four different clinical specialities. These specialities were identified as high volume users of short-term catheters from the audit (Results of audit of short-term catheter policies above), namely General Surgery, Acute Medical Assessment Unit, the Stroke Unit and Gynaecology. The first week of the two-week feasibility study was dedicated to fine-tuning the practicalities of the study in the Urology ward because of the experience of the medical and nursing staff and then rolled out to the other specialities towards the end of the first week. A dedicated researcher visited the four wards daily. Of 14 patients identified during their hospital admission requiring a short-term catheter, nine were approached and five were missed. All nine approached agreed and consented to take part in the feasibility study. Thus our estimate of 60% of those approached likely to want to participate (Table 1) is conservative, though based on a small sample.

Design

A ten-centre randomised controlled trial testing three short-term urinary catheter policies in a range of high-volume clinical settings. Participants will be randomised either by using a telephone system or a web-based system. The Centre for Healthcare Randomised Trials (CHaRT), Health Services Research Unit, University of Aberdeen, as the Trial Data Centre, will manage the randomisation process. As at June 2008, ten centres have agreed to participate: Newcastle upon Tyne Hospitals NHS Trust, City Hospitals Sunderland NHS Foundation Trust, Gateshead Healthcare NHS Foundation Trust, Aberdeen Royal Infirmary NHS Grampian, Raigmore Hospital NHS Highland, Southampton University Hospitals NHS Trust, Northumbria Healthcare NHS Foundation Trust, Western General and Royal Infirmary of Edinburgh Hospital NHS Lothian, Bristol Royal Infirmary United Bristol Healthcare NHS Trust and Hillingdon Hospital NHS Trust.

Setting

Secondary care units with a high volume of short-term catheterisation. In October 2005, an audit of catheter use in Newcastle and Aberdeen identified high volume units including: surgical specialities, general medical wards, care of the elderly wards, high dependency and intensive care units. These are the specialities that will be targeted in the main trial.

Inclusion/exclusion criteria

• Inclusion:

  1. – Adult patients (≥16 years of age) requiring urethral catheterisation (expected to be required for a maximum of 14 days), in pre-selected units with a high volume of short-term catheterisation.

• Exclusion:

  1. – Patients for whom urinary catheterisation is expected to be long-term (i.e. > 14 days).

  2. – Urological intervention or instrumentation within preceding 7 days (e.g. catheterisation, cystoscopy, prostatic biopsy and nephrostomy insertion).

  3. – Non-urethral catheterisation (e.g. suprapubic catheterisation).

  4. – Known allergy to any of the following: latex, silver salts, hydrogel, silicone or nitrofurazone

  5. – Any patient who has a microbiologically confirmed symptomatic urinary tract infection, at time of randomisation.

  6. – Unable to give informed consent.

The planned trial interventions

1. Experimental groups: There will be two experimental groups managed with:

   1. silver alloy impregnated hydrogel urethral catheter (S).

   2. nitrofurazone impregnated silicone urethral catheter (N).

2. Control group

There will be one control group managed with a PTFE coated latex urethral catheter – the ‘standard’ control (P).

Participants will be randomised 1:1:1 to the three groups. Size 14 Ch catheters will be used for all three arms. The choice of catheter as ‘standard’ control was based on the results of the audit of short-term catheter use in all secondary care wards in Newcastle and Aberdeen (see Results of audit of short-term catheter policies), which confirmed that the PTFE-coated latex urethral catheter was the most commonly used in both hospitals (over 70%); it is also relatively inexpensive compared with the coated catheters (at approximately £0.86 each for the PTFE-coated catheter vs £5.50 for the antiseptic coated and £4.50 for the antibiotic coated, 2007 prices).

Each catheter will have a detachable sticker attached to the outer packaging. This sticker will display the ‘CATHETER’ logo and either ‘N’, ‘S’ or ‘P’ to denote catheter type (N, Nitrofurazone; S, Silver; P, PTFE). These stickers will then be stuck directly onto Patient Consent Forms so as to determine that the patient was given the catheter they were randomised to.

Proposed outcome measures

Proposed duration of intervention

The mean period of urethral catheterisation is expected to be between 1 and 14 days.

What is the proposed frequency and duration of follow-up and how will outcome measures be measured?

Following enrolment in the trial, a mid-stream specimen of urine will be sent for microbiological analysis (i.e. microscopy, culture and sensitivity) immediately prior to catheterisation, if one has not been sent within the preceding 48 hours (baseline sample). Where this is not possible, a specimen of urine will be obtained during catheterisation (i.e. catheter-specimen of urine) using standard aseptic techniques. Urine will also be collected again for microbiological analysis within, or at, 3 days after catheter removal. If patients are discharged home prior to the third post-catheter removal day a sample will be taken as close to the 3 day after removal time point as possible. However, patients will also be provided with sterile urine collection bottles to be filled and submitted by post, or local NHS courier service, on the third post-catheter removal day if the sample taken in hospital is before the 3 day removal time point. If a clinical diagnosis of symptomatic UTI is made at any stage, including during the period of catheterisation, either a catheter-specimen or mid-stream specimen of urine will be obtained by ward staff or the patient’s GP according to normal clinical practice.

Participants will complete the following questionnaires at the specified time points:

• baseline questionnaire (EQ-5D and a urinary symptom questionnaire) around the time of recruitment

• a 3 days after catheter removal questionnaire (Urinary Symptoms and EQ-5D)

• patient diary to be completed at 1 and 2 weeks after catheter removal (will contain specific UTI questions and EQ-5D)

• final set of questionnaires (i.e. EQ-5D and specific Health Economic questionnaires) 6 weeks after randomisation.

The questionnaires will be targeted at identifying symptomatic UTI as well as other catheter-associated problems (e.g. urethral discomfort), quality of life, and any health economic implications such as costs to the patients and the NHS. Questionnaires completed prior to hospital discharge will be collected by the Recruitment Co-ordinator. Questionnaires completed after hospital discharge will be posted back to the Data Centre (CHaRT, University of Aberdeen); patients will be supplied with a discharge pack containing the relevant questionnaires and stamped addressed envelopes. The Recruitment Co-ordinator will telephone participants when at home to remind and help them to complete Trial questionnaires and submit urine samples. This will also provide another opportunity to answer any questions the participants may have about the Trial and the associated paperwork. This phone call system will not be standard practice and will only be used when participants fail to send back questionnaires or urine samples once they have been discharged from hospital.

Collection of information to describe UTIs

Information will be collected in five phases to contribute data to identify UTIs.

Proposed sample size

Based on the Cochrane review and other data12,15,16, the anticipated incidence of UTI in the standard control group is 7%. Given that this trial is assessing patient reported outcomes rather than microbiology or clinician reported symptoms, the predicted incidence rate of UTIs is 11% and it is reasonable to hypothesise that the impregnated catheters will reduce this to about 8%. The increased cost associated with use of the silver alloy and nitrofurazone catheters is further justification for aiming to identify an effect of this size. With 90% power and alpha set at 2.5% rather than 5% (to adjust for the two comparisons of antibiotic impregnated catheter and control catheter, and antiseptic impregnated catheter and control catheter), and using chi-square tests of association to compare the incidence of CASUTI in an intervention group and the control group, the trial would require an estimated 1970 fully followed up participants per group, inflated to 2362 per group to allow for an observed loss to follow up of 17%, giving a total of 7086.

Statistical analysis

All primary analyses will be according to the intention-to-treat principle, and will be governed by a Statistical Analysis Plan, which will be agreed by the Trial Steering Committee (TSC). The outcomes listed in Proposed outcome measures will be compared with (a) antibiotic impregnated and control and (b) antiseptic impregnated and control using generalised linear models (for example, for the primary outcome of proportion of participants with symptomatic catheter associated infection, a logistic regression model will be used, adjusting for any covariates felt to be of prognostic importance). Nominal 95% confidence intervals will be calculated. Subgroup analyses will examine possible effect modification of age, gender, co-morbidity, duration of catheterisation, indication for catheterisation, and antibiotic use prior to enrolment using tests for interaction (all at stricter levels of significance P< 0.01). We will also explore the comparison of antibiotic and antiseptic groups. A single main analysis will be performed at the end of the trial when all follow up has been completed. An independent Data Monitoring Committee (DMC, see below) will review confidential interim analyses as frequently as requested (see Independent supervision) of accumulating data but at least annually.

Economic evaluation

The economic evaluation will be integrated into the trial. Outcomes and costs will be assessed from the perspective of the NHS and patients for a 6-week time horizon. The alternatives compared are described in The planned trial interventions.

Effectiveness will be measured in terms of number of symptomatic urinary tract infections up to 6 weeks after randomisation and quality-adjusted life-years (QALYs) at six weeks. QALYs will be derived using data from EQ-5D administered at baseline, 3 days, 1 week, 2 weeks and at 6 weeks post randomisation as part of the main study questionnaires. These responses will be converted into health state utilities using UK population tariffs. The estimation of QALYs will take account of the mortality of study participants. Participants who die within the study follow-up will be assigned a zero utility weight from their death until the end of the study follow up. QALYs will be estimated using linear extrapolation between the QALY scores at baseline and all available EQ-5D.

Costs for the six week follow-up will be assessed from the trial. The number of infections will be collected using the baseline questionnaire, the UTI questionnaire completed 3 days after catheter removal, the patient diary at 1 and 2 weeks after catheter removal and from the final questionnaire at 6 weeks. Other cost generating events of the interventions such as the use of primary care services including contacts with primary care practitioners (e.g. GPs and practice nurses) and prescription medications, will be collected using the health-care utilisation questionnaires administered at 1, 2 and 6 weeks follow-up. Use of secondary care services following the period of catheterisation will be collected using the 6 week follow up questionnaire. If patients are identified as having been re-admitted following their discharge, within the 6 weeks post randomisation, they will be further investigated using the NHS Patient Administration System (PAS). This will record information on non-protocol outpatient visits, readmissions relating to the use and consequences of urinary catheters. This approach will ensure that information on the use of secondary care services that could be main determinants of incremental costs can be identified without overburdening the participants.

Estimates of resource utilisation will be combined with unit costs to derive total costs. Unit costs will be based on study-specific estimates and data from standard sources. Participant costs will be based on self-purchased health care (e.g. prescription costs, over the counter medications), will be collected as part of the health-care utilisation questions (see above). The results of these analyses will be presented as point estimates of mean incremental costs, QALYs and cost per QALY. Measures of variance for these outcomes will be derived using bootstrapping and deterministic sensitivity analysis will be used to explore other forms of uncertainty (e.g. the exploration of the implication of extrapolating outcomes beyond the 6 weeks time horizon). Similar sub-group analysis to that described in Statistical analysis will also be undertaken.

A simple decision model will be developed to compare the different catheters in terms of the loss of quality of life (based on the responses to the EQ-5D collected as part of the trial) and change in cost caused by a symptomatic catheter-associated infection. In this analysis it will be assumed that the only loss of quality of life will be caused by a symptomatic catheter-associated infection and that the type of catheter does not affect quality of life except by changing the risk of a symptomatic catheter-associated infection occurring. Regression methods will be used to estimate the loss of quality of life for those with a symptomatic catheter-associated infection compared with those without a symptomatic catheter-associated infection. A similar approach will be used to estimate costs associated with developing a symptomatic catheter-associated infection compared with costs for those without a symptomatic catheter-associated infection. Estimates of cost for each arm of the decision model (i.e. each type of catheter) will be adjusted to reflect the cost of using the catheter. Probabilistic sensitivity analysis will be performed along with deterministic analysis to reflect other forms of uncertainty within the model.

Practical arrangements for identifying and allocating participants to trial groups

(a) Identifying potentially eligible participants: A potentially eligible participant will be identified by either ward staff, operating theatre staff, or by the local recruitment co-ordinator. The recruitment co-ordinator will visit the wards daily to check on possible activity that day and by doing so, also promote the trial. Laminated recruitment sheets will be placed prominently at the nurses station/doctors room on the wards. Participants will be identified according to the pre-stated inclusion and exclusion criteria from pre-determined high volume units. The local recruitment co-ordinator will obtain informed consent (see below) and ensure completion of in-hospital questionnaires and other data collection.

(b) Informed consent: Participants will be provided with written information and will sign a consent form once they have had enough time to understand the implications and requirements of the trial. The only exception to this will be for adults in an emergency situation.

In emergency situations common reasons for short-term catheterisation are acute urinary retention or for monitoring purposes in an unwell patient. These patients will be provisionally included in the trial according to a protocol which will respect both their right to best quality care and the ethical requirement of informed consent. As the decision to catheterise is based solely on clinical need by the caring physician, we propose that a randomised catheter is used in an emergency situation. Once the patient’s condition has settled, they will be provided with an information sheet and an opportunity to opt out of the trial if they wish, with assurances that such a decision will not affect the level of care they receive. As the antiseptic/antibiotic-impregnated catheters are reported to lessen the risk of infection compared with the usual standard catheter, we feel this proposed protocol for including adults in emergency situations is justified. Ethics Committee approval was secured from the all Research Ethics Committees involved in the trial to include such patients in the study using the above arrangements for post-procedure consent or withdrawal.

Although unlikely, we have anticipated a situation whereby a patient may have been randomised and is discharged before the recruitment co-ordinator can obtain informed consent (e.g. if over the weekend). Therefore, we propose to send a letter to any patient that may have been randomised to the trial and ‘missed’ by the recruiter. This letter will describe the trial, a patient information sheet will also be sent, and will invite them to participate in the trial. They will be asked to contact the recruitment co-ordinator if they are interested in taking part in the trial. A copy of the letter that will be sent to the patient is enclosed.

(c) Recruitment: The feasibility study and audit indicated that although the ‘office hours’ are busy periods for the flow of potentially eligible participants, it will be necessary to be able to recruit throughout the 24 hour period. The telephone randomisation system is the easiest way to effectively recruit throughout this period.

(d) Randomisation: Eligible participants will be randomised centrally by simple randomisation to the three groups using either by telephone to the automated IVR telephone randomisation application at CHaRT in Aberdeen.

(e) Data Management. The recruitment co-ordinator, during the daily round, will input the last session’s data via the study web portal to the Study Data Centre in Aberdeen on a daily basis.

(f) Use of Posters. The inclusion of posters to be displayed in the treatment rooms of wards taking part in the trial as reminders of:

• the trial and

• the catheters to be used in the trial.

Posters are also to be included in admission lounges and ward areas to highlight the trial to eligible patients. It is hoped that this will alert patients to the trial so that when they are approached to take part, they have some familiarity with the research. Each site will have posters than contain site specific contact details.

(g) Withdrawal: Any patient who is withdrawn from the study for personal or medical reasons will be recorded appropriately using a specific Withdrawal Form. The form will record what the patient has withdrawn from e.g. questionnaires being sent, their hospital and GP records being accessed (for the results of medically indicated urine samples and antibiotics prescribed for UTIs) and contact by the CATHETER team.

Some patients will be classed as post-randomisation exclusions as opposed to withdrawals. Patients who will be categorised as post-randomisation exclusions are as follows:

• emergency patients who retrospectively refuse consent

• those patients who are randomised due to medical intention to catheterise but are never catheterised

• those patients who are randomised but receive a suprapubic catheter in theatre.

In line with this a notice (a CATHETER compliments slip, see enclosed) documenting that the patient was never catheterised or received a suprapubic catheter will be placed in the patients notes. The recruiter will explain to the patient that they are no longer in the trial.

Methods for protecting against other sources of bias

As indicated, the study will be open, with at least the antibiotic catheter easily identifiable by its unavoidably bright yellow colour. The concern about ‘open’ use of catheters is that knowing which catheter has been used might influence clinical decisions, such as about sending a urine specimen for culture or removing a catheter. To guard against bias introduced in this way, we shall standardise clinical policies in these respects. Further protection comes from the fact that the person inserting the catheter is unlikely to be the same person making later clinical decisions specifically regarding timing of catheter removal. All urine samples will be tested ‘blind’ to the catheter allocation. Outcome assessment in terms of symptoms of UTI will be made by patient self-completed questionnaires. Participants will not be told until the end of the study which catheter has been used to protect against the influence of preconceptions on the participants’ side in respect of self-reported outcomes. While researchers will be blinded to type of catheter, clinical staff performing the catheterisation will not be.

To attempt to standardise catheter insertion technique, catheter care, catheter urine sampling and catheter removal policies, a protocol has been developed and tested in the feasibility study and will be implemented in all centres involved in the trial. This protocol incorporates principles of ‘Best Practice’ based on published guidelines17. In addition, the protocol also advocates that patients with asymptomatic bacteriuria should not routinely be treated with antibiotics18. Catheter care violations, such as inappropriate collection of urine (e.g. from catheter bag rather than from the sampling port or accidental disconnection of the closed-drainage system) will be recorded to attempt to explain false-positive diagnosis of bacteriuria12. Every attempt will also be made to record any catheter changes or reinsertion of a catheter during the trial period.

Ethical arrangements

This project has been approved by the Grampian MREC. We believe this is a very low risk study, with a correspondingly favourable risk: benefit ratio, since both patients and society (through efficient allocation of resources within NHS) will benefit from the results of the research. We do not anticipate that there will be risks to participants, and they may benefit from participation in the trial. Society will benefit from the conclusions of the research, which will be used to inform future short-term catheter policies in secondary care. An information leaflet will be given to each potential participant to inform them of the benefits and known drawbacks that may apply. Informed signed consent will be obtained from the participants in all centres. Participants who cannot give informed consent (e.g. due to their mental state) will be excluded. The trial will be co-ordinated from a centre with considerable experience of multicentre trials (CHaRT), cognisant of the implications of research governance and other legal frameworks for the conduct of trials. This is not classed as a trial of any investigational medicinal products or Medical Devices, and so does not come under the EU Clinical Trials Directive. Nevertheless we will conduct the study to the standards required by ICH GCP, and the NHS and Universities Research Governance as well as all other applicable legal, ethical and regulatory requirements.

The risk of adverse events associated with catheterisation is minimal. Please see appendix 1 for further information on SAE reporting.

Arrangements for independent supervision are described below.

Independent supervision

A Trial Steering Committee that will include an independent Chairperson and other independent members and a consumer representative will oversee the trial. The Steering Committee will meet at least three times over the course of the study, and at least annually. The DMC will meet early in the trial to agree its terms of reference and other procedures. While its deliberations will take into account the full implications of any recommendations it may choose to make (such as whether or not any interim findings are sufficiently convincing to change clinical practice), the statistical guidelines adopted are likely to be those suggested by Peto19 based on a difference in a main outcome measure of at least three standard deviations. The DMC will report any recommendations to the Chairperson of the TSC.

Duration 33 months

1–6 months: set up office, assemble team, secure ethical approval

7–12 months: establish eight centres

7–24 months: identify and recruit 5700 participants requiring short-term catheterisation (an average of 720 is feasible in each of the eight centres based on the Newcastle and Aberdeen audits)

10–25.5 months: follow up at 6 weeks

25.5–33 months: complete data collection, analysis and dissemination.

Figure 1 shows the projected recruitment of centres and participants, and projected number who would be approached. Our plan is to establish two centres relatively early in the project (by seven months) and to roll out to the others over the subsequent five months. We recognise that this will need prior preparation. An application to MREC will be submitted between funding decision and the start date.

FIGURE 1.

Projected recruitment chart.

The participant recruitment graph in Figure 1 has been modelled to take into account: the rollout to the centres over the first 5 months and that there are likely to be slightly fewer short-term catheters inserted around August and over Christmas due to a reduction in elective surgical throughput.

All of the following are defined as serious adverse events

• Resulted in death.

• Life-threatening.

• Required inpatient hospitalisation or prolongation of existing hospitalisation.

• Septicaemia.

• Persistent or significant disability/incapacity.

Prioritisation Group

1. Director, NIHR HTA programme, Professor of Clinical Pharmacology, Department of Pharmacology and Therapeutics, University of Liverpool

2. Professor Imti Choonara, Professor in Child Health, Academic Division of Child Health, University of Nottingham

   Chair – Pharmaceuticals Panel

3. Dr Bob Coates, Consultant Advisor – Disease Prevention Panel

4. Dr Andrew Cook, Consultant Advisor – Intervention Procedures Panel

5. Dr Peter Davidson, Director of NETSCC, Health Technology Assessment

6. Dr Nick Hicks, Consultant Adviser – Diagnostic Technologies and Screening Panel, Consultant Advisor–Psychological and Community Therapies Panel

7. Ms Susan Hird, Consultant Advisor, External Devices and Physical Therapies Panel

8. Professor Sallie Lamb, Director, Warwick Clinical Trials Unit, Warwick Medical School, University of Warwick

   Chair – HTA Clinical Evaluation and Trials Board

9. Professor Jonathan Michaels, Professor of Vascular Surgery, Sheffield Vascular Institute, University of Sheffield

   Chair – Interventional Procedures Panel

10. Professor Ruairidh Milne, Director – External Relations

11. Dr John Pounsford, Consultant Physician, Directorate of Medical Services, North Bristol NHS Trust

    Chair – External Devices and Physical Therapies Panel

12. Dr Vaughan Thomas, Consultant Advisor – Pharmaceuticals Panel, Clinical

    Lead – Clinical Evaluation Trials Prioritisation Group

13. Professor Margaret Thorogood, Professor of Epidemiology, Health Sciences Research Institute, University of Warwick

    Chair – Disease Prevention Panel

14. Professor Lindsay Turnbull, Professor of Radiology, Centre for the MR Investigations, University of Hull

    Chair – Diagnostic Technologies and Screening Panel

15. Professor Scott Weich, Professor of Psychiatry, Health Sciences Research Institute, University of Warwick

    Chair – Psychological and Community Therapies Panel

16. Professor Hywel Williams, Director of Nottingham Clinical Trials Unit, Centre of Evidence-Based Dermatology, University of Nottingham

    Chair – HTA Commissioning Board

    Deputy HTA Programme Director

HTA Commissioning Board

1. Professor of Dermato-Epidemiology, Centre of Evidence-Based Dermatology, University of Nottingham

2. Professor of Bio-Statistics, Department of Public Health and Epidemiology, University of Birmingham

3. Professor of Clinical Pharmacology, Director, NIHR HTA programme, Department of Pharmacology and Therapeutics, University of Liverpool

1. Professor Zarko Alfirevic, Head of Department for Women’s and Children’s Health, Institute of Translational Medicine, University of Liverpool

2. Professor Judith Bliss, Director of ICR-Clinical Trials and Statistics Unit, The Institute of Cancer Research

3. Professor David Fitzmaurice, Professor of Primary Care Research, Department of Primary Care Clinical Sciences, University of Birmingham

4. Professor John W Gregory, Professor in Paediatric Endocrinology, Department of Child Health, Wales School of Medicine, Cardiff University

5. Professor Steve Halligan, Professor of Gastrointestinal Radiology, Department of Specialist Radiology, University College Hospital, London

6. Professor Angela Harden, Professor of Community and Family Health, Institute for Health and Human Development, University of East London

7. Dr Joanne Lord, Reader, Health Economics Research Group, Brunel University

8. Professor Stephen Morris, Professor of Health Economics, University College London, Research Department of Epidemiology and Public Health, University College London

9. Professor Dion Morton, Professor of Surgery, Academic Department of Surgery, University of Birmingham

10. Professor Gail Mountain, Professor of Health Services Research, Rehabilitation and Assistive Technologies Group, University of Sheffield

11. Professor Irwin Nazareth, Professor of Primary Care and Head of Department, Department of Primary Care and Population Sciences, University College London

12. Professor E Andrea Nelson, Professor of Wound Healing and Director of Research, School of Healthcare, University of Leeds

13. Professor John David Norrie, Director, Centre for Healthcare Randomised Trials, Health Services Research Unit, University of Aberdeen

14. Professor Barney Reeves, Professorial Research Fellow in Health Services Research, Department of Clinical Science, University of Bristol

15. Professor Peter Tyrer, Professor of Community Psychiatry, Centre for Mental Health, Imperial College London

16. Professor Martin Underwood, Professor of Primary Care Research, Warwick Medical School, University of Warwick

17. Professor Caroline Watkins, Professor of Stroke and Older People’s Care, Chair of UK Forum for Stroke Training, Stroke Practice Research Unit, University of Central Lancashire

18. Dr Duncan Young, Senior Clinical Lecturer and Consultant, Nuffield Department of Anaesthetics, University of Oxford

1. Dr Tom Foulks, Medical Research Council

2. Dr Kay Pattison, Senior NIHR Programme Manager, Department of Health

HTA Clinical Evaluation and Trials Board

1. Director, Warwick Clinical Trials Unit, Warwick Medical School, University of Warwick and Professor of Rehabilitation, Nuffield Department of Orthopaedic, Rheumatology and Musculoskeletal Sciences, University of Oxford

2. Professor of the Psychology of Health Care, Leeds Institute of Health Sciences, University of Leeds

3. Director, NIHR HTA programme, Professor of Clinical Pharmacology, University of Liverpool

1. Professor Keith Abrams, Professor of Medical Statistics, Department of Health Sciences, University of Leicester

2. Professor Martin Bland, Professor of Health Statistics, Department of Health Sciences, University of York

3. Professor Jane Blazeby, Professor of Surgery and Consultant Upper GI Surgeon, Department of Social Medicine, University of Bristol

4. Professor Julia M Brown, Director, Clinical Trials Research Unit, University of Leeds

5. Professor Alistair Burns, Professor of Old Age Psychiatry, Psychiatry Research Group, School of Community-Based Medicine, The University of Manchester & National Clinical Director for Dementia, Department of Health

6. Dr Jennifer Burr, Director, Centre for Healthcare Randomised trials (CHART), University of Aberdeen

7. Professor Linda Davies, Professor of Health Economics, Health Sciences Research Group, University of Manchester

8. Professor Simon Gilbody, Prof of Psych Medicine and Health Services Research, Department of Health Sciences, University of York

9. Professor Steven Goodacre, Professor and Consultant in Emergency Medicine, School of Health and Related Research, University of Sheffield

10. Professor Dyfrig Hughes, Professor of Pharmacoeconomics, Centre for Economics and Policy in Health, Institute of Medical and Social Care Research, Bangor University

11. Professor Paul Jones, Professor of Respiratory Medicine, Department of Cardiac and Vascular Science, St George‘s Hospital Medical School, University of London

12. Professor Khalid Khan, Professor of Women’s Health and Clinical Epidemiology, Barts and the London School of Medicine, Queen Mary, University of London

13. Professor Richard J McManus, Professor of Primary Care Cardiovascular Research, Primary Care Clinical Sciences Building, University of Birmingham

14. Professor Helen Rodgers, Professor of Stroke Care, Institute for Ageing and Health, Newcastle University

15. Professor Ken Stein, Professor of Public Health, Peninsula Technology Assessment Group, Peninsula College of Medicine and Dentistry, Universities of Exeter and Plymouth

16. Professor Jonathan Sterne, Professor of Medical Statistics and Epidemiology, Department of Social Medicine, University of Bristol

17. Mr Andy Vail, Senior Lecturer, Health Sciences Research Group, University of Manchester

18. Professor Clare Wilkinson, Professor of General Practice and Director of Research North Wales Clinical School, Department of Primary Care and Public Health, Cardiff University

19. Dr Ian B Wilkinson, Senior Lecturer and Honorary Consultant, Clinical Pharmacology Unit, Department of Medicine, University of Cambridge

1. Ms Kate Law, Director of Clinical Trials, Cancer Research UK

2. Dr Morven Roberts, Clinical Trials Manager, Health Services and Public Health Services Board, Medical Research Council

Diagnostic Technologies and Screening Panel

1. Scientific Director of the Centre for Magnetic Resonance Investigations and YCR Professor of Radiology, Hull Royal Infirmary

2. Professor Judith E Adams, Consultant Radiologist, Manchester Royal Infirmary, Central Manchester & Manchester Children’s University Hospitals NHS Trust, and Professor of Diagnostic Radiology, University of Manchester

3. Mr Angus S Arunkalaivanan, Honorary Senior Lecturer, University of Birmingham and Consultant Urogynaecologist and Obstetrician, City Hospital, Birmingham

4. Dr Diana Baralle, Consultant and Senior Lecturer in Clinical Genetics, University of Southampton

5. Dr Stephanie Dancer, Consultant Microbiologist, Hairmyres Hospital, East Kilbride

6. Dr Diane Eccles, Professor of Cancer Genetics, Wessex Clinical Genetics Service, Princess Anne Hospital

7. Dr Trevor Friedman, Consultant Liason Psychiatrist, Brandon Unit, Leicester General Hospital

8. Dr Ron Gray, Consultant, National Perinatal Epidemiology Unit, Institute of Health Sciences, University of Oxford

9. Professor Paul D Griffiths, Professor of Radiology, Academic Unit of Radiology, University of Sheffield

10. Mr Martin Hooper, Public contributor

11. Professor Anthony Robert Kendrick, Associate Dean for Clinical Research and Professor of Primary Medical Care, University of Southampton

12. Dr Nicola Lennard, Senior Medical Officer, MHRA

13. Dr Anne Mackie, Director of Programmes, UK National Screening Committee, London

14. Mr David Mathew, Public contributor

15. Dr Michael Millar, Consultant Senior Lecturer in Microbiology, Department of Pathology & Microbiology, Barts and The London NHS Trust, Royal London Hospital

16. Mrs Una Rennard, Public contributor

17. Dr Stuart Smellie, Consultant in Clinical Pathology, Bishop Auckland General Hospital

18. Ms Jane Smith, Consultant Ultrasound Practitioner, Leeds Teaching Hospital NHS Trust, Leeds

19. Dr Allison Streetly, Programme Director, NHS Sickle Cell and Thalassaemia Screening Programme, King’s College School of Medicine

20. Dr Matthew Thompson, Senior Clinical Scientist and GP, Department of Primary Health Care, University of Oxford

21. Dr Alan J Williams, Consultant Physician, General and Respiratory Medicine, The Royal Bournemouth Hospital

1. Dr Tim Elliott, Team Leader, Cancer Screening, Department of Health

2. Dr Joanna Jenkinson, Board Secretary, Neurosciences and Mental Health Board (NMHB), Medical Research Council

3. Professor Julietta Patrick, Director, NHS Cancer Screening Programme, Sheffield

4. Dr Kay Pattison, Senior NIHR Programme Manager, Department of Health

5. Professor Tom Walley, CBE, Director, NIHR HTA programme, Professor of Clinical Pharmacology, University of Liverpool

6. Dr Ursula Wells, Principal Research Officer, Policy Research Programme, Department of Health

Disease Prevention Panel

1. Professor of Epidemiology, University of Warwick Medical School, Coventry

2. Dr Robert Cook, Clinical Programmes Director, Bazian Ltd, London

3. Dr Colin Greaves, Senior Research Fellow, Peninsula Medical School (Primary Care)

4. Mr Michael Head, Public contributor

5. Professor Cathy Jackson, Professor of Primary Care Medicine, Bute Medical School, University of St Andrews

6. Dr Russell Jago, Senior Lecturer in Exercise, Nutrition and Health, Centre for Sport, Exercise and Health, University of Bristol

7. Dr Julie Mytton, Consultant in Child Public Health, NHS Bristol

8. Professor Irwin Nazareth, Professor of Primary Care and Director, Department of Primary Care and Population Sciences, University College London

9. Dr Richard Richards, Assistant Director of Public Health, Derbyshire County Primary Care Trust

10. Professor Ian Roberts, Professor of Epidemiology and Public Health, London School of Hygiene & Tropical Medicine

11. Dr Kenneth Robertson, Consultant Paediatrician, Royal Hospital for Sick Children, Glasgow

12. Dr Catherine Swann, Associate Director, Centre for Public Health Excellence, NICE

13. Mrs Jean Thurston, Public contributor

14. Professor David Weller, Head, School of Clinical Science and Community Health, University of Edinburgh

1. Ms Christine McGuire, Research & Development, Department of Health

2. Dr Kay Pattison, Senior NIHR Programme Manager, Department of Health

3. Professor Tom Walley, CBE, Director, NIHR HTA programme, Professor of Clinical Pharmacology, University of Liverpool

External Devices and Physical Therapies Panel

1. Consultant Physician North Bristol NHS Trust

2. Reader in Wound Healing and Director of Research, University of Leeds

3. Professor Bipin Bhakta, Charterhouse Professor in Rehabilitation Medicine, University of Leeds

4. Mrs Penny Calder, Public contributor

5. Dr Dawn Carnes, Senior Research Fellow, Barts and the London School of Medicine and Dentistry

6. Dr Emma Clark, Clinician Scientist Fellow & Cons. Rheumatologist, University of Bristol

7. Mrs Anthea De Barton-Watson, Public contributor

8. Professor Nadine Foster, Professor of Musculoskeletal Health in Primary Care Arthritis Research, Keele University

9. Dr Shaheen Hamdy, Clinical Senior Lecturer and Consultant Physician, University of Manchester

10. Professor Christine Norton, Professor of Clinical Nursing Innovation, Bucks New University and Imperial College Healthcare NHS Trust

11. Dr Lorraine Pinnigton, Associate Professor in Rehabilitation, University of Nottingham

12. Dr Kate Radford, Senior Lecturer (Research), University of Central Lancashire

13. Mr Jim Reece, Public contributor

14. Professor Maria Stokes, Professor of Neuromusculoskeletal Rehabilitation, University of Southampton

15. Dr Pippa Tyrrell, Senior Lecturer/Consultant, Salford Royal Foundation Hospitals’ Trust and University of Manchester

16. Dr Nefyn Williams, Clinical Senior Lecturer, Cardiff University

1. Dr Kay Pattison, Senior NIHR Programme Manager, Department of Health

2. Dr Morven Roberts, Clinical Trials Manager, Health Services and Public Health Services Board, Medical Research Council

3. Professor Tom Walley, CBE, Director, NIHR HTA programme, Professor of Clinical Pharmacology, University of Liverpool

4. Dr Ursula Wells, Principal Research Officer, Policy Research Programme, Department of Health

Interventional Procedures Panel

1. Professor of Vascular Surgery, University of Sheffield

2. Consultant Colorectal Surgeon, Bristol Royal Infirmary

3. Mrs Isabel Boyer, Public contributor

4. Mr Sankaran Chandra Sekharan, Consultant Surgeon, Breast Surgery, Colchester Hospital University NHS Foundation Trust

5. Professor Nicholas Clarke, Consultant Orthopaedic Surgeon, Southampton University Hospitals NHS Trust

6. Ms Leonie Cooke, Public contributor

7. Mr Seumas Eckford, Consultant in Obstetrics & Gynaecology, North Devon District Hospital

8. Professor Sam Eljamel, Consultant Neurosurgeon, Ninewells Hospital and Medical School, Dundee

9. Dr Adele Fielding, Senior Lecturer and Honorary Consultant in Haematology, University College London Medical School

10. Dr Matthew Hatton, Consultant in Clinical Oncology, Sheffield Teaching Hospital Foundation Trust

11. Dr John Holden, General Practitioner, Garswood Surgery, Wigan

12. Dr Fiona Lecky, Senior Lecturer/Honorary Consultant in Emergency Medicine, University of Manchester/Salford Royal Hospitals NHS Foundation Trust

13. Dr Nadim Malik, Consultant Cardiologist/Honorary Lecturer, University of Manchester

14. Mr Hisham Mehanna, Consultant & Honorary Associate Professor, University Hospitals Coventry & Warwickshire NHS Trust

15. Dr Jane Montgomery, Consultant in Anaesthetics and Critical Care, South Devon Healthcare NHS Foundation Trust

16. Professor Jon Moss, Consultant Interventional Radiologist, North Glasgow Hospitals University NHS Trust

17. Dr Simon Padley, Consultant Radiologist, Chelsea & Westminster Hospital

18. Dr Ashish Paul, Medical Director, Bedfordshire PCT

19. Dr Sarah Purdy, Consultant Senior Lecturer, University of Bristol

20. Dr Matthew Wilson, Consultant Anaesthetist, Sheffield Teaching Hospitals NHS Foundation Trust

21. Professor Yit Chiun Yang, Consultant Ophthalmologist, Royal Wolverhampton Hospitals NHS Trust

1. Dr Kay Pattison, Senior NIHR Programme Manager, Department of Health

2. Dr Morven Roberts, Clinical Trials Manager, Health Services and Public Health Services Board, Medical Research Council

3. Professor Tom Walley, CBE, Director, NIHR HTA programme, Professor of Clinical Pharmacology, University of Liverpool

4. Dr Ursula Wells, Principal Research Officer, Policy Research Programme, Department of Health

Pharmaceuticals Panel

1. Professor in Child Health, University of Nottingham

2. Senior Lecturer in Clinical Pharmacology, University of East Anglia

3. Dr Martin Ashton-Key, Medical Advisor, National Commissioning Group, NHS London

4. Dr Peter Elton, Director of Public Health, Bury Primary Care Trust

5. Dr Ben Goldacre, Research Fellow, Division of Psychological Medicine and Psychiatry, King’s College London

6. Dr James Gray, Consultant Microbiologist, Department of Microbiology, Birmingham Children’s Hospital NHS Foundation Trust

7. Dr Jurjees Hasan, Consultant in Medical Oncology, The Christie, Manchester

8. Dr Carl Heneghan, Deputy Director Centre for Evidence-Based Medicine and Clinical Lecturer, Department of Primary Health Care, University of Oxford

9. Dr Dyfrig Hughes, Reader in Pharmacoeconomics and Deputy Director, Centre for Economics and Policy in Health, IMSCaR, Bangor University

10. Dr Maria Kouimtzi, Pharmacy and Informatics Director, Global Clinical Solutions, Wiley-Blackwell

11. Professor Femi Oyebode, Consultant Psychiatrist and Head of Department, University of Birmingham

12. Dr Andrew Prentice, Senior Lecturer and Consultant Obstetrician and Gynaecologist, The Rosie Hospital, University of Cambridge

13. Ms Amanda Roberts, Public contributor

14. Dr Gillian Shepherd, Director, Health and Clinical Excellence, Merck Serono Ltd

15. Mrs Katrina Simister, Assistant Director New Medicines, National Prescribing Centre, Liverpool

16. Professor Donald Singer, Professor of Clinical Pharmacology and Therapeutics, Clinical Sciences Research Institute, CSB, University of Warwick Medical School

17. Mr David Symes, Public contributor

18. Dr Arnold Zermansky, General Practitioner, Senior Research Fellow, Pharmacy Practice and Medicines Management Group, Leeds University

1. Dr Kay Pattison, Senior NIHR Programme Manager, Department of Health

2. Mr Simon Reeve, Head of Clinical and Cost-Effectiveness, Medicines, Pharmacy and Industry Group, Department of Health

3. Dr Heike Weber, Programme Manager, Medical Research Council

4. Professor Tom Walley, CBE, Director, NIHR HTA programme, Professor of Clinical Pharmacology, University of Liverpool

5. Dr Ursula Wells, Principal Research Officer, Policy Research Programme, Department of Health

Psychological and Community Therapies Panel

1. Professor of Psychiatry, University of Warwick, Coventry

2. Consultant & University Lecturer in Psychiatry, University of Cambridge

3. Professor Jane Barlow, Professor of Public Health in the Early Years, Health Sciences Research Institute, Warwick Medical School

4. Dr Sabyasachi Bhaumik, Consultant Psychiatrist, Leicestershire Partnership NHS Trust

5. Mrs Val Carlill, Public contributor

6. Dr Steve Cunningham, Consultant Respiratory Paediatrician, Lothian Health Board

7. Dr Anne Hesketh, Senior Clinical Lecturer in Speech and Language Therapy, University of Manchester

8. Dr Peter Langdon, Senior Clinical Lecturer, School of Medicine, Health Policy and Practice, University of East Anglia

9. Dr Yann Lefeuvre, GP Partner, Burrage Road Surgery, London

10. Dr Jeremy J Murphy, Consultant Physician and Cardiologist, County Durham and Darlington Foundation Trust

11. Dr Richard Neal, Clinical Senior Lecturer in General Practice, Cardiff University

12. Mr John Needham, Public contributor

13. Ms Mary Nettle, Mental Health User Consultant

14. Professor John Potter, Professor of Ageing and Stroke Medicine, University of East Anglia

15. Dr Greta Rait, Senior Clinical Lecturer and General Practitioner, University College London

16. Dr Paul Ramchandani, Senior Research Fellow/Cons. Child Psychiatrist, University of Oxford

17. Dr Karen Roberts, Nurse/Consultant, Dunston Hill Hospital, Tyne and Wear

18. Dr Karim Saad, Consultant in Old Age Psychiatry, Coventry and Warwickshire Partnership Trust

19. Dr Lesley Stockton, Lecturer, School of Health Sciences, University of Liverpool

20. Dr Simon Wright, GP Partner, Walkden Medical Centre, Manchester

1. Dr Kay Pattison, Senior NIHR Programme Manager, Department of Health

2. Dr Morven Roberts, Clinical Trials Manager, Health Services and Public Health Services Board, Medical Research Council

3. Professor Tom Walley, CBE, Director, NIHR HTA programme, Professor of Clinical Pharmacology, University of Liverpool

4. Dr Ursula Wells, Principal Research Officer, Policy Research Programme, Department of Health