The British antibiotic and silver-impregnated catheters for ventriculoperitoneal shunts multi-centre randomised controlled trial (the BASICS trial): study protocol

The BASICS trial is a national three-arm, multi-centre, phase III RCT comparing antibiotic-impregnated, silver-impregnated, and standard VPS in patients with hydrocephalus undergoing insertion of their first permanent shunt. The design of the trial is presented in Figure 1.

The trial will be conducted across 17 paediatric and adult regional neurosurgery units in the UK and Ireland. A feasibility survey completed by 14 centres provided information on the number of eligible hydrocephalus patients and confirmed that the sample size is achievable.

The BASICS trial is funded by a £2.04 m grant from the National Institute of Health Research Health Technology Assessment (NIHR-HTA) programme. CLM is the chief investigator and MDJ is the co-chief investigator. The study protocol, patient information sheets, and consent forms have received ethical approval from the North West Greater Manchester Research Ethics Committee (reference: 12/NW/0790).

The trial will be open to all patients (children and adults) with hydrocephalus requiring treatment with a first permanent VPS who meet the eligibility criteria.

Patients will be randomised to standard, antibiotic, or silver-impregnated VPS in a ratio of 1:1:1. These are all CE marked medical devices used in accordance with the manufacturer’s instructions for their intended purpose. Written informed consent and assent, where appropriate, will be obtained pre-operatively, but randomisation will not occur until the trial participant is in the theatre suite so that only patients demonstrating continuing eligibility will be randomised at this stage, avoiding the randomisation of ineligible patients as far as is practically possible. Theatre staff cannot be blinded to the type of shunt inserted, but once the shunt is in position there will be no visible indications of the shunt type used. The type of shunt inserted will not be disclosed outside the operating theatre. Training on non-disclosure of shunt type will be provided to all staff members present at surgery. The majority of shunt revisions and removals for infection happen as emergencies and are managed by the on-call team. The likelihood of the same surgeon who inserted the shunt being involved in the decision to remove the shunt is low given the work rotas of surgical staff. However, this will be tracked by logging the staff involved in surgery and in the decision to remove the shunt. This process is the same as that successfully implemented for the catheter infections in children (CATCH) trial (ISRCTN34884569) and will form part of the trial monitoring report presented to the trial management group.

Approximately 30% of new VPS will fail within the first 6 months of insertion due to shunt complications, which will include infection but also malfunction and mechanical failure, typically secondary to catheter or valve obstruction and peritoneal complications. When a VPS has failed for reasons other than infection then this prevents infection being observed. Infection is the event of interest with all other reasons for shunt failure being competing risks. The trial will compare antibiotic-impregnated versus standard VPS, and silver-impregnated versus standard VPS. A Bonferroni adjustment has been made to allow for the multiple comparisons and a statistical significance level of 0.025 will be used accordingly. The sample size calculation is based on Pintilie[24] assuming time to any type of event (of interest or competing risks) follows an exponential distribution. Using a 2-year accrual period with 6-month follow-up, once accrual has been completed and a competing risks event rate of 30% across trial arms with the event rate of interest (infection) being 8% in the standard arm and 4% in the treated arms (hazard ratio of 0.49), then with 5% loss to follow-up we would require 989 patients to provide 80% power. The power varies according to changes in the event rate with a fixed total sample size of approximately 1,143. Allowing for 5% loss to follow-up a sample size of 1,200 participants allows a hazard ratio of 0.49 to be detected over a range of baseline event rates (0.05 to 0.1) with good statistical power. The value of 8% is supported elsewhere[17], in addition, infection rate surveillance over 10 years (1993 to 2003) at Great Ormond Street Hospital, London, UK, involving over 1,500 insertions has demonstrated that while there are fluctuations in monthly infection rates that overall the rate has remained remarkably stable around this level. Observed fluctuations in infection rates within a centre will be influenced by the size of the denominator. Reduction of infection is a priority and therefore it could be argued that much smaller differences than those the study is powered to detect are still important. However the investigators believe that a strong effect is required to inform clinical practice and establish a first-line treatment policy across the National Health Service (NHS). This is in part due to the large differences in cost between the VPS; we need to warrant expenditure on the type of VPS as opposed to other infection control activities.

The primary analysis will be by intention-to-treat principle as far as is practically possible and a full statistical analysis plan will be developed prior to comparative analysis. Results will be presented throughout using 97.5% confidence intervals and a 2.5% level of statistical significance. Analysis of the primary outcome will be by cumulative incidence. The Fine and Gray[25] regression method will be used to directly model the cumulative incidence function. In addition two semi-parametric models described in Scheike and Zhang[26] will be used to consider time-varying effects. Time-to-event outcomes where competing risks is not an issue will be analysed using Kaplan–Meir curves, log-rank tests, and Cox proportional hazard models. Assumptions of proportional hazards will be investigated. Categorical outcomes will be analysed using Chi-squared test. An interim analysis will be conducted halfway through the trial after approximately 50% of the total events have been observed using Haybittle[27].

The health economic analysis will adopt the perspective of the NHS in the UK and personal social services. Costs will include those of the catheters, duration of intensive care stay and hospital admission, antibiotic treatment, and contact with health professionals and social services. Resource use will be based on entries made in designated sections of patients’ case report forms which will include questions on use of personal social services[28], complemented with hospital.

Hospital Episode Statistics (HES) data will be sourced from the NHS Information Centre. Unit cost data will be obtained from routine hospital data (NHS reference costs)[29] and other standard sources[30, 31]. Given that the economic question under consideration is one of technical efficiency (that is, the aim to maximise health outcomes given the resource inputs), we shall approach this as a cost-effectiveness analysis, based on the incremental cost per shunt failure averted. Secondary economic outcomes will include: 1) incremental cost per shunt infection avoided; and 2) incremental cost per QALY gained, estimated by administering the EuroQol (Rotterdam, Netherlands) EQ-5D-3 L, EQ-5D-Y (youth version), or Health Utilities Index Mark 2 (HUI2; Dundas, ON, Canada) to patients (or their parents, according to age) at each follow-up visit. Costs and benefits occurring after the first year will be discounted at 3.5% per annum. Total costs will be combined with the measures of health outcome to calculate the incremental cost-effectiveness (utility) ratios of each technology. Where appropriate, missing resource use or health outcome data will be imputed. The number of QALYs experienced by each patient will be calculated as the area under the curve, using the trapezoidal rule, and corrected for baseline utility score. Non-parametric bootstrapped 95% confidence intervals will be estimated (10,000 replicates). We will also employ parametric approaches for analysing cost and QALY data that assume normal distributions given the large samples where the near-normality of sample means is approximated. Should the data indicate otherwise, we will develop a generalised linear model to deal with problems such as skewness. Stratified cost-effectiveness analyses will be conducted on important, pre-specified patient subgroups (including neonates, patients under 1 year old, and patients with a previous EVD). Cost-utility estimates will be compared with the £20,000 to £30,000 per QALY threshold of cost-effectiveness, specified by National Institute for Health and Care Excellence (NICE), and a range of one-way sensitivity analyses will be conducted to assess the robustness of the analysis. Multivariate sensitivity analyses will be applied where interaction effects are suspected, for instance in the assessment of the combined effect of development of resistant organisms and the costs associated with their management. The joint uncertainty in costs and benefits will be considered through the application of bootstrapping and the estimation of cost-effectiveness acceptability curves[32].

The study has ethical approval and funding to obtain an additional 5 mL of blood and CSF gifted by patients at VPS insertion and revision. These samples will be stored until trial completion and then used for research studies to characterise the blood and CSF patterns of host response during confirmed shunt infection (via molecular methods including proteomics and transcriptomics), with an aim to use these patterns to strengthen the diagnosis of shunt infections and improve pathogen detection.

The communication and dissemination strategy will actively involve participating centres, their staff and service users, professional bodies involved (Society of British Neurological Surgeons, SBNS), and relevant charitable organisations (spina bifida, hydrocephalus, information, networking, equality, Shine). Communication and dissemination of results will be assisted by members of the study team, including using social networking sites. Findings of the trial will also be presented at national and international meetings of relevant professional bodies and research groups. The results of the trial will be published in peer-reviewed journals.

If the trial favours a particular impregnated catheter, we will initiate a forum for discussion with other specialties such as neurologists, microbiologists, and infection control specialists and industry to discuss the implications for practice for research into impregnated catheters for other patient groups requiring implants in the central nervous system, such as baclofen pumps and catheters, spinal cord stimulators, and deep brain stimulation.