Silver-impregnated, antibiotic-impregnated or non-impregnated ventriculoperitoneal shunts to prevent shunt infection: the BASICS three-arm RCT

Article history

The research reported in this issue of the journal was funded by the HTA programme as project number 10/104/30. The contractual start date was in December 2012. The draft report began editorial review in April 2019 and was accepted for publication in November 2019. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HTA editors and publisher have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the draft report document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

Michael J Griffiths has a patent 068347A1 pending for a novel method of detection of bacterial infection. Tom Solomon reports grants from the National Institute for Health Research (NIHR) outside the submitted work and other support from the Data Safety and Monitoring Committee of the GlaxoSmithKline plc (London, UK) study to evaluate the safety and immunogenicity of a candidate ebola vaccine in children (GSK3390107A) (ChAd3 EBO-Z), outside the submitted work. He also chairs the Siemens Healthineers (Munich, Germany) Clinical Advisory Board. Dyfrig Hughes was member of the Health Technology Assessment (HTA) programme Pharmaceuticals Panel (2008–12) and the HTA programme Clinical Evaluation and Trials board (2010–16). Carrol Gamble reports grants from NIHR outside the submitted work and is a member of the NIHR Efficacy and Mechanism Evaluation programme committee (January 2015–present).

Copyright statement

© Queen’s Printer and Controller of HMSO 2020. This work was produced by Mallucci et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2020 Queen’s Printer and Controller of HMSO

Current practice

Hydrocephalus affects one in every 500 births,3 and is thus one of the most common developmental disabilities in children. The condition also affects older children and adults of all ages, and can be secondary to a variety of causes, including intracranial tumours, haemorrhage and infection. In the late 1950s, the development of a treatment with cerebral shunts revolutionised the management of these patients.

Standard treatment for hydrocephalus remains the ventriculoperitoneal shunt (VPS) catheter. A VPS comprises silicone tubing with the addition of an in-line valve that is designed to control the rate of cerebrospinal fluid (CSF) flow. The tubing passes from the brain fluid cavities (ventricles) under the skin to the peritoneum (abdominal cavity). The shunt drains CSF from the ventricles to the peritoneal cavity.

Insertion of a VPS to treat hydrocephalus is now one of the most common procedures performed in neurosurgical units, and between 3000 and 3500 shunt operations are carried out per year in the UK in adults and children. 4 Once inserted, a shunt is generally required for life; it will inevitably be susceptible to failure, in terms of both infection (as it remains an implanted foreign body) and mechanical failure, usually due to blockage of tubing or valve failure. Thus, patients with shunts will need lifelong follow up and usually require multiple surgeries. Therefore, VPS treatment for hydrocephalus is a major health burden to the NHS.

Industry produces a number of different VPS types, and costs associated with these can vary. The market comprises a wide variety of different valves and, more recently, different types of shunt catheter. The treating surgeon and hospital often chooses the type of valve and shunt tubing based on personal preference and/or associated costs.

There are three types of VPS catheter available (standard, antibiotic impregnated and silver impregnated). There is no standard practice or guidance in the UK as to which shunt catheter is the most effective at reducing infection. Practice is variable across the UK and the world. There are no current National Institute for Health and Care Excellence (NICE) guidelines, nor is there a position statement from the Society of British Neurological Surgeons regarding the use of any type of VPS.

As an infection in a newly implanted shunt can have such devastating consequences for the patient, with far-reaching health economic sequelae,5 the industry has led the way in trying to develop types of shunt catheters that will reduce infection. It is incumbent on clinical researchers to assess the effectiveness of these developments; this study attempts to answer this question.

Rationale

Shunt failure due to infection has plagued this neurosurgical advance ever since it was developed. The reported incidence of shunt infection varies markedly in the literature from 3% to 27%6–10 and is higher in certain groups, for example neonates and children aged < 1 year, and patients treated with a previous temporary external ventricular drain (EVD). Episodes of shunt infection have a major impact on both patients and the NHS and require prolonged inpatient hospitalisation, additional surgery to remove the infected hardware, placement of a temporary EVD, intravenous and intrathecal antibiotics and further surgery to place a new shunt once the infection has been treated. Other clinical consequences of infection, including epilepsy, reduced intelligence quotient (IQ) and loculation, have often been reported8 but never formally studied in the context of a prospective clinical trial. The number of shunt infections is an independent predictor of death in patients requiring CSF shunts [hazard ratio (HR) 1.66, 95% confidence interval (CI) 1.02 to 2.72]. 11

The most common pathogens detected were staphylococcus species, but, in a proportion of patients with suspected infection, the organism is never determined, especially if the patient has already received antimicrobial treatment or if there was a delay in culturing the organism, both of which hamper microbiological treatment. 5 However, newer molecular approaches are being developed,12 including the substudy within this trial.

Data from the UK shunt registry (to which most neurosurgical units contribute) report that 15% of shunt revisions are for infection. 13 In the largest randomised controlled shunt trial worldwide, the infection rate was 8.4% for primary VPSs. 14

Impregnated VPS catheters have been introduced as a means to reduce VPS infection, in addition to the usual surgical site infection prevention care bundles that are not standardised across neurosurgery clinical practice.

There are three types of VPS catheters available, and there are cost implications associated with impregnated shunt catheters that, typically, are more than double the cost of the standard non-impregnated VPS catheters:

1. standard VPSs are made of silicone and are available and supplied by a number of different companies

2. antibiotic-impregnated VPSs are made of silicone and are impregnated with antibiotics (0.15% clindamycin and 0.054% rifampicin) [available as Bactiseal^® (Codman^®; Integra LifeSciences Holdings Corporation, Plainsboro, NJ, USA) and Ares™ (Medtronic plc, Dublin, Ireland)]

3. silver-impregnated VPSs are made of silicone and impregnated with silver [available as Silverline® (Spiegelberg GmbH & Co. KG, Hamburg, Germany)].

Despite a large number of publications15–22 prior to our study, there has been limited evidence to date indicating the clinical effectiveness of these impregnated shunt catheters. Prior to our study, a systematic review and meta-analysis23 of the Bactiseal VPS identified one randomised controlled trial (RCT)15 and 11 observational studies. The RCT,15 conducted in a single centre in South Africa, demonstrated a trend favouring impregnated VPSs, but did not show a statistically significant difference between the two trial groups [relative risk (RR) 0.38, 95% CI 0.11 to 1.30; p = 0.12]; however, meta-analysis of the 11 observational studies showed a statistically significant difference favouring the Bactiseal VPS (RR 0.37, 95% CI 0.23 to 0.60; p < 0.01). 23 Research on the Bactiseal VPS conducted in Liverpool, UK, has shown that, over a 2-year period, the infection rate reduced among paediatric patients who were given the Bactiseal VPS compared with historical controls. 16 However, continued data collection over 3.5 years, published as part of a Liverpool-led multicentre observational study in collaboration with two other UK paediatric neurosurgical units, showed no significant reduction in infection. 17 Indeed, the reduction in infection achieved by the Bactiseal VPS in the multicentre observational study17 was seen only in neonates and was heavily weighted by the results from one unit. This study17 was not part of the published systematic review. 23

Silver-impregnated shunts were launched in the UK in March 2011. There is little doubt that silver ions have antimicrobial effects and they elute from Silverline shunt catheters. However, the efficacy of Silverline shunt catheters at preventing VPS infections is not proven. In vitro models have shown varying results and clinical studies are limited. 18,19 There is one observational study of the Silverline VPS,20 in which the Silverline VPS was used to successfully treat seven patients with active CSF infection. There are no observational studies comparing Silverline VPS infection rates with those of either standard or Bactiseal VPSs. However, in a RCT of EVDs (an EVD is a temporary tube placed in the ventricles that is prone to infection) in children and adults, Silverline EVDs have been shown to reduce infection from 21.4% (30/140) with standard shunt catheters to 12.3% (17/138) (p = 0.043) for silver shunt catheters. 22 Two further observational studies comparing standard with Silverline EVDs also show a reduction in infection rates. 21,24

Risks and benefits

The potential beneficial effect on health status of these impregnated shunt catheters is reduced shunt infection and its negative sequelae. Prior to this study, approximately 70%4,13 of shunt operations in the UK were with Bactiseal shunts (verified by feasibility screening logs) and it was felt that, just like Bactiseal, there was likely to be a significant uptake of Silverline shunts by neurosurgeons, despite the lack of evidence of clinical or cost benefit.

The potential adverse effects of impregnated shunt catheters has never been studied prospectively, to our knowledge. One of the potential concerns of antibiotic-impregnated shunt catheters is the potential for selecting out resistant organisms or missing potential infections owing to an inability to culture them.

Thus, before the wide adoption of these impregnated shunt catheters, an adequately powered RCT is needed to assess their effectiveness at reducing infection and to determine their safety (including the type of organisms cultured), antibiotic sensitivities and antibiotic resistances.

Aims and objectives

Trial design

A schematic representation of the trial design is given in Figure 1.

FIGURE 1.

Trial design. CT, computerised tomography; MRI, magnetic resonance imaging. Reproduced from Jenkinson et al. 1 This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The figure includes minor additions and formatting changes to the original.

Reproduced from Jenkinson et al. 1 This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The figure includes minor additions and formatting changes to the original.

Ethics approval and research governance

The protocol was approved by the Greater Manchester South Research Ethics Committee (reference number 12/NW/0790). The trial was funded by the National Institute for Health Research Health Technology Assessment programme (number 10/104/30) and included on the International Standard Randomised Controlled Trial Number registry (ISRCTN49474281). Centre-specific approval was obtained at all of the recruiting centres.

The protocol has been published previously. 1 The trial opened on protocol version 3.0, and the final approved version of the protocol was version 13.0, which contains a complete list of protocol changes [see www.fundingawards.nihr.ac.uk/award/10/104/30 (accessed 22 January 2020)]. A summary of substantial protocol amendments are provided in Table 1.

Selection of trial centres

Participants were recruited from 21 regional adult and paediatric neurosurgery centres in the UK and the Republic of Ireland. To be eligible to participate in the trial, centres had to meet the British Antibiotic and Silver Impregnated Catheters for ventriculoperitoneal Shunts (BASICS) trial centre suitability assessment criteria:

• minimum of three patients per month

• neurosurgical unit treating adults or paediatrics

• evidence of a team to undertake trial activities

• principal investigator (PI) had previous experience of RCTs or a significant role

• no local issues to prevent trial set-up

• completion of prospective screening log.

Participants

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

Recruitment procedure

Trial assessments

Table 2 provides the schedule of trial assessments. Participants were followed up for a minimum of 6 months and a maximum of 2 years, dependent on their randomisation date.

Data collection

Data were collected on paper-based case report forms (CRFs) completed by centre staff who were authorised to do so and returned to the Liverpool Clinical Trials Centre (LCTC) Clinical Trials Unit (CTU). Participants were issued with diaries to record their health-care utilisation every 12 weeks and administered questionnaires to measure quality of life. A planned analysis of data from NHS Digital could not be achieved as the sponsor was unable to meet NHS Digital requirements for obtaining Hospital Episode Statistics data within the project timeline. Electronic health-care data were therefore obtained from Patient-Level Information and Costing Systems (PLICS).

Data were collected at baseline, randomisation, the peri-operative assessment, the early post-operative assessment, the first routine post-operative assessment and the subsequent post-operative assessment, and at the time points described in the remainder of this section.

Measures

Sample size

The sample size for the primary outcome was calculated using the Pintilie26 method with the following assumptions: (1) failure for infection was the event of interest, with all other reasons for failure a competing risk, (2) the rate of infection was 8% in the standard silicone arm14 and 4% in the impregnated shunt catheter arms, (3) the competing risk event rate was 30% and (4) there was a 5% loss to follow-up. A total sample size of 1200 participants with 119 events demonstrated good statistical power (88%), with leverage for a lower event rate if required (Table 4). A feasibility study conducted in trial centres for 1 month indicated an annual eligible participant figure of 1200; a conservative estimate of consent of 50% suggested that the sample size would be achievable within a 2-year recruitment period, with participants followed up for a minimum of 6 months.

An interim analysis was planned after 50% of the total events had been observed, using the Haybittle–Peto method. 27

Monitoring of the infection rate during the trial demonstrated that the majority of events occur within 1 month of VPS insertion (i.e. they are not exponentially distributed), and that the rates of infection, competing risk and loss to follow-up were lower than expected. In January 2016, the Independent Data and Safety Monitoring Committee (IDSMC) reviewed the sample size calculations and recommended increasing recruitment to a target of 1606 participants with 101 events, to provide 80% power; the Trial Steering Committee (TSC) agreed and approved this change. The early occurrence of events and assumption of exponential risk were managed in the Pintilie26 method assumptions by reducing the accrual and follow-up rates to 1 month.

Statistical methods

The main features of the analysis plan were specified in the protocol; the final analyses were undertaken according to a more detailed and prespecified statistical analysis plan (see Report Supplementary Material 1), consistent with the protocol.

Efficacy outcomes were analysed according to the intention-to-treat principle as far as practically possible; AEs and SAEs were reported according to the type of VPS in situ. A Bonferroni adjustment28 was made to allow for multiple comparisons (antibiotic-impregnated vs. standard VPS, and silver-impregnated vs. standard VPS) and a 2.5% level of statistical significance and 97.5% CIs were used throughout.

Outcomes with infection as the event of interest used Fine and Gray29 survival regression models with cause-specific hazard ratios (csHRs) and subdistribution hazard ratios (sHRs) presented. 30,31 Cox regression models were used to analyse time to VPS failure due to any cause. Reason for VPS failure is presented descriptively (see Chapter 3, Secondary outcome 3: reason for shunt failure) and with a chi-squared test. Types of organisms and their resistances and sensitivities are presented descriptively in Chapter 3, Secondary outcome 4: types of bacterial infection. Quality-of-life outcomes were analysed using mixed models. All survival models were adjusted for the age category of the recruiting centre (paediatric or adult), with adult centres further categorised by age > 65 years. A post hoc analysis was conducted that explored revision rates, and reason for revision, by aetiology of the hydrocephalus, type of valve and operative approach. Results of the post hoc analysis are presented descriptively in Chapter 3, Post hoc analyses.

Primary outcome and safety analyses were validated by independent programming from the point of raw data extraction. All analyses were carried out with SAS^® software version 9.4 with SAS/STAT package 14.3 (SAS Institute Inc, Cary, NC, USA).

Patient and public involvement

The trial team collaborated with young people and parent contributors throughout the trial:

• Advice was sought from the Medicines for Children Research Network Young Person’s Group on the content and presentation of patient information leaflets and consent forms. The Medicine for Children’s Research Network is a division of the LCTC, part of the Liverpool Clinical Trials Collaborative.

• Three lay members were invited at the trial outset to join the Trial Management Group (TMG) and TSC; one was recruited to be a member of the TMG.

• Members of the TMG, including the lay member, met via teleconference with the patient and public involvement co-ordinator for the LCTC early in the trial to establish the timings for return of the patient-completed diaries and to identify ways to maximise the return rate of these diaries. They decided that it would be appropriate for the centre team to contact the participant or representative via telephone every 3 months as a reminder to complete and return the diaries.

• The charity Shine (Spina bifida Hydrocephalus Information Networking Equality) was continually supportive of the trial. A Shine representative was a member of the TSC.

Trial oversight and role of funders

The TMG, comprising the chief investigator, other lead investigators (clinical and non-clinical) and members of the LCTC CTU, was responsible for the day-to-day running and management of the trial. The membership of the oversight committees was suggested by members of the TMG to the trial funders and appointed by the funders with their constitution following funder requirements.

The TSC consisted of an independent chairperson, an independent microbiologist, a lay representative from the Shine charity and an independent statistician. The chief investigator was a non-independent member of the TSC. The role of the TSC was as the executive decision-making committee, considering the recommendations of the IDSMC. Monitoring reports viewed by the TSC were not split by treatment group.

The IDSMC consisted of an independent chairperson, plus two independent members: an expert in the field of microbiology and an expert in medical statistics. The IDSMC was responsible for reviewing and assessing recruitment, interim monitoring of safety and effectiveness, trial conduct and external data. The IDSMC provided recommendations to the TSC concerning the continuation of the trial and viewed accumulating data split by treatment group.

All protocol amendments were approved by the funder prior to ethics submission.

Recruitment and screening

The trial opened to recruitment on 26 June 2013 and closed on 9 October 2017.

During this period, 3505 patients were screened for eligibility, of whom 1605 patients were randomised from 21 centres. One patient was randomised twice and their data contributed from the first randomisation only. See Appendix 2, Figure 8 and Tables 32 and 33, for screening and recruitment data.

Screened patients who were not randomised fell into four categories: the patient did not meet eligibility criteria (n = 1020); the patient was eligible but consent was not sought (n = 435); consent was sought but the patient declined (n = 369); and the patient was not randomised for another reason (n = 67). The overall consent rate in patients who were approached for participation was 82%.

A Consolidated Standards of Reporting Trials (CONSORT) flow diagram illustrating the pathway of patients from screening to consent and randomisation is provided in Figure 2.

FIGURE 2.

The CONSORT flow chart. ITT, intention to treat.

Baseline comparability

The three groups were similar in their baseline characteristics (Table 5) and baseline risk assessment (Table 6). Approximately 40% of all participants were paediatric patients, with one-quarter of all participants being aged < 1 year at the time of randomisation. The factors recorded on the baseline risk assessment were those regarded within the literature as being associated with a high risk of infection.

Retention and adherence

Table 7 summarises compliance with the randomly allocated shunt. Of the 1605 participants randomised, four (0.2%) had no VPS inserted and 16 (1%) received a different VPS to the one that was randomly allocated. Reasons for participants having an alternative trial VPS inserted or no VPS are provided in Appendix 2, Tables 35 and 36, respectively. Participants receiving no VPS were excluded from the intention-to-treat population; for the safety analysis, participants were analysed according to the VPS received. The analysis sets are summarised in Table 8.

In total, 53 (3.3%) randomised participants withdrew from the trial. No participants withdrew consent to use collected data. Table 9 summarises the level of and reasons for withdrawal.

Unblinding

A total of 32 participants from 10 centres were unblinded during the course of the trial.

Unblinding could be accidental or intentional. Accidental unblinding was defined as an unplanned occurrence; for example, the allocation was incorrectly recorded in the participant notes. Intentional unblinding occurred when the unblinding envelope was opened; for example, if a patient was transferred to another hospital and staff needed to be aware of their allocation. Twenty-five participants from eight centres were accidentally unblinded and seven participants from four centres were intentionally unblinded. Table 10 summarises unblinding events, both overall and by randomised VPS.

Protocol deviations

Prespecified protocol deviations are summarised in Appendix 2, Table 37. The most common major protocol deviation was that shunt components were not taken for culture at shunt revision/removal (n = 320, 19.9%). This was not routine practice in many units, and the impact of this deviation on the identification of infections for the primary outcome was mitigated by the low numbers of missing CSF at revision (n = 14), as CSF analysis and culture was the primary factor used for defining shunt infection (not culture of the shunt tubing or components).

Antibiotic sensitivity data were not returned on many isolates.

Primary outcome: time to ventriculoperitoneal infection as assessed by the central review panel

The primary outcome was time to VPS infection, as assessed by the central review panel.

Four participants received no shunt and seven participants had an infection at insertion; these participants were excluded from the primary analysis set (see Figure 2).

A summary of the first VPS revisions and infections among participants in the primary intention-to-treat analysis set is provided in Table 11. The overall revision rate of first VPS was 25% (398/1594), and was approximately equal between each of the three VPS groups.

All first revisions were classified as to whether or not the revision was for suspected infection by the central review panel. If there was insufficient information for the central review panel to classify an infection (n = 1/398; see Table 11), the clinical classification, as recorded on the CRFs by the treating surgeon, was used.

Of the total number of first revisions, 75 were classified infections (4.7%). The infection rate was approximately equal in the standard and silver-impregnated VPS arms (6.0% and 5.9%, respectively) and lowest in the antibiotic-impregnated VPS arm (2.2%). The time to infection was similar across all treatment arms {standard VPS arm: median 1 month [lower quartile (LQ)–upper quartile (UQ) 0–1.5 months]; antibiotic-impregnated VPS arm: median 1 month [LQ–UQ 0–2 months]; silver-impregnated VPS arm: median 1 months [LQ–UQ 0–1 months]}.

When compared with the standard VPS, antibiotic-impregnated VPSs decreased the risk of infection (csHR 0.38, 97.5% CI 0.18 to 0.80; p < 0.01) (Table 12). Silver-impregnated VPSs were comparable to standard VPSs (csHR 0.99, 97.5% CI 0.56 to 1.74; p = 0.96) (see Table 12). Figure 3 displays the cumulative incidences of infection and no infection by VPS and age group. Figure 4 displays the cumulative incidence plots of infection or no infection by VPS group, stratified by age group.

FIGURE 3.

Cumulative incidence plots of revisions for infection and competing risk by VPS group and age group. (a) Infection by VPS group; (b) competing risk by VPS group; (c) infection by age group; and (d) competing risk by age group.

FIGURE 4.

Cumulative incidence plots of revisions for infection and competing risk by VPS group, stratified by age group. (a) Infection in the paediatric group; (b) infection in those aged ≤ 65 years; (c) infection in those aged > 65 years; (d) competing risk in the paediatric group; (e) competing risk in those aged ≤ 65 years; and (f) competing risk in those aged > 65 years.

All revisions for infection were further classified by type, with the majority being classified as ‘definite: culture positive’ across all arms (standard VPS arm: 22/32, 68.8%; antibiotic-impregnated VPS arm: 6/12, 50%; silver-impregnated VPS arm: 25/31, 80.6%).

Secondary outcomes

Secondary outcome 2: time to ventriculoperitoneal shunt failure due to any cause

The overall revision rate was 25.0% (398/1594), which was approximately equal between each of the three VPS groups (standard VPS arm: 130/533, 24.4%; antibiotic-impregnated VPS arm: 132/535, 24.7%; silver-impregnated VPS arm: 136/526, 25.9%) (see Table 11).

There was no significant difference for time to failure between the antibiotic-impregnated or silver-impregnated VPS arms when compared with the standard VPS arm (Table 15). Figure 5 shows the Kaplan-Meier curve for time to VPS failure for any cause, split by VPS and age group.

TABLE 15 - Estimates for Cox proportional HRs

Follow-up time from first VPS summary statistics: median 22 months; LQ–UQ 10–24 months; minimum 0 months, maximum 24 months.

Covariate	Revision, HR (97.5% CI); p-value
VPS
Standard	–
Antibiotic impregnated	1.01 (0.77 to 1.33); 0.94
Silver impregnated	1.08 (0.82 to 1.42); 0.54
Age group
Paediatric	–
Adult (≤ 65 years)	0.57 (0.44 to 0.74); < 0.01
Adult (> 65 years)	0.25 (0.18 to 0.35); < 0.01

FIGURE 5.

Kaplan–Meier curves for time to VPS failure for any cause, split by (a) VPS group and (b) age group.

Safety analysis

Efficacy outcomes were analysed by the intention-to-treat analysis population as much as possible. For AEs and SAEs, participants are reported according to the type of VPS in situ at the time of the event. The shunt in situ was known for all patients up to the first revision. However, events occurring following a revision whereby the shunt was not replaced like for like are reported in under ‘other VPS’ group.

The total number of AEs experienced and the number of participants experiencing at least one AE are provided, both overall and split according to the VPS in situ at the time of the event. Of the 1601 participants who received a trial shunt, 413 (25.8%) experienced 654 events. Summarising these events, split by VPS in situ at the time of event, indicates the following:

• Standard VPS – 135 out of 531 participants (25.4%) experienced 201 events.

• Antibiotic-impregnated VPS – 136 out of 545 participants (25.0%) experienced 210 events.

• Silver-impregnated VPS – 140 out of 525 participants (26.7%) experienced 191 events.

• Other VPS – when the initial trial shunt was removed and not replaced like for like, 35 out of 136 participants (25.7%) experienced 52 events.

Common events were ventricular shunt catheter obstruction (96 events in 79 participants), shunt valve obstruction (65 events in 52 participants) and valve change for symptomatic overdrainage (54 events in 50 participants).

Appendix 2, Table 45, provides the summary of related AEs.

No participant experienced a SAE.

Introduction

A number of studies32–36 (although none from the UK) have estimated the costs of managing patients with VPS infections. These often combine cost estimates with observed differences in infection rates, between antibiotic-impregnated and standard VPS catheters, in rudimentary cost-effectiveness analyses, to calculate the incremental cost (saving) per VPS infection avoided. This is relevant to inform the best use of health-care resources, given that impregnated VPS catheters are about twice as expensive as standard shunt catheters, but are limited in not assessing all-cause VPS failure or impacts on patients’ health-related quality of life.

Adopting the perspective of the German health-care setting, Eymann et al. 32 estimated the costs of VPS infections by considering the lengths, and per diem costs, of adult and paediatric intensive care and ward stays. Based on 48 cases of VPS infections among a cohort of 197 hydrocephalus patients with antibiotic-impregnated VPS catheters and 98 patients with standard VPS catheters, the estimated mean cost per infection was US$16,933 in children and US$14,886 in adults. Considering the benefits and additional costs of antibiotic-impregnated shunt catheters, the authors estimated a net saving of US$51,651 for the 197 patients.

Farber et al. 33 conducted a retrospective cohort study of 500 shunt surgeries performed in adult patients with hydrocephalus in a US hospital: 250 received antibiotic-impregnated VPS catheters and 250 received standard VPS catheters. The hospital billing records of all patients who were treated were analysed to estimate the costs associated with hospital stays for infections. The overall mean cost per VPS infection was US$40,371 and, based on an absolute reduction in the risk of shunt infections of 2.8% with antibiotic-impregnated shunt catheters, savings of US$47,193 in infection-related direct costs were estimated per 100 shunt surgeries performed.

Attenello et al. 34 analysed the hospital billing records of 406 paediatric patients who underwent 608 shunt placement procedures (400 antibiotic-impregnated and 208 standard shunt catheters) in a single US hospital. The majority of procedures (93%) were for VPSs. The hospital resources used to treat the 38 patients in whom a shunt infection developed were estimated to cost an average of US$48,454. Based on an observed decreased incidence of shunt infection in the antibiotic-impregnated shunt catheter cohort, the estimated infection-related cost savings per 100 patients was $442,133 per 100 patients with shunts.

Parker et al. 35,36 analysed patient discharge and billing records from 287 US hospitals to estimate the shunt infection rates and costs of antibiotic-impregnated shunt catheters compared with standard shunt catheters. The unadjusted difference in the incidence of shunt infection was 1.4% and 4.5% in adult and paediatric populations, respectively, and the costs of managing infections was US$45,714 and US$56,104, respectively. The authors estimated that the use of antibiotic-impregnated shunt catheters in adults was associated with infection-related cost savings of US$42,125 per 100 de novo shunts placed (US$230,390 per 100 in paediatrics), which corresponds to US$30,854 and US$51,181 per shunt infection avoided, respectively (calculated from data presented in the paper).

Each of the above studies relied on retrospective cohorts of patients, mainly from single centres, but result in estimates for the costs of managing VPS infections in the order of US$50,000 in the US health-care setting. A key limitation, however, is that, in each of these analyses, infection rates were determined from observational data, and were therefore potentially subject to bias, reducing the reliability of the cost-effectiveness estimates. Analyses that considered the RR of VPS infection associated with antibiotic-impregnated shunt catheters, determined from clinical trial data, are limited to three studies. 37–39

Root et al. 37 conducted a meta-analysis of clinical trials and observational studies, and estimated that the number needed to treat with antibiotic-impregnated (compared with standard) EVDs to prevent a shunt catheter-associated infection was 19 (95% CI 15 to 36). Based on assumed costs of US$100 per antibiotic-impregnated shunt catheter and US$30,000 per episode of infection, they estimated that antibiotic-impregnated shunt catheters could result in overall savings, from the perspective of a US neurosurgical unit, of US$28,100 (95% CI US$26,400 to US$28,500) for each shunt catheter-associated infection prevented.

Klimo et al. 38 similarly reviewed the literature for clinical trials and observational studies of antibiotic-impregnated shunts, and conducted a meta-analysis and a simple cost calculation. Assuming a cost of US$50,000 to treat a shunt infection, the cost savings per shunt infection prevented ranged from just under US$90,000 to > US$1.3M per 200 shunts performed. No data were presented on the incremental cost (saving) of avoiding a shunt infection.

Edwards et al. 39 developed a decision-analytic model of the clinical and economic consequences of using antibiotic-impregnated shunts from the perspective of a US hospital. Using trial and observational data, they estimated that, for every 100 patients requiring shunts, antibiotic-impregnated shunt catheters may be associated with 0.5 fewer deaths, 71 fewer hospital days, 11 fewer surgeries and a net saving of US$128,228 owing to decreased infection. The cost of a shunt infection was US$46,394, derived from earlier estimates of Attenello et al. 34 and Farber et al. 33

All analyses have many limitations, not least the assumption that potential avoidance of the cost of managing VPS infections can be equated to a cost saving. Moreover, the analyses lacked any consideration of health outcomes associated with shunt catheter-associated infections or other potential causes of VPS failure.

Aim

The aim of the economic evaluation was to assess the cost-effectiveness of antibiotic-impregnated, silver-impregnated and standard VPS catheters in children and adults with hydrocephalus who were recruited in the BASICS trial.

Methods

The economic analysis adopted the perspective of the NHS and Personal Social Services providers in the UK. The analytical approach for the primary analysis was a cost-effectiveness analysis, based on the incremental cost per first shunt failure (due to any cause) averted for impregnated and standard VPSs. A cost–utility analysis was conducted to estimate the incremental cost per QALY gained in a restricted sample of trial participants.

Results

Economic health outcomes

The proportions of participants who experienced a first VPS failure (due to any cause) within 2 years were 130 out of 533, 132 out of 535 and 136 out of 526 in the standard, antibiotic-impregnated and silver-impregnated VPS groups, respectively. In the base-case analysis, with a 3.5% annual discount rate, the VPS failure rate was 23.3% (97.5% CI 19.1% to 27.3%) in the standard VPS group, 25.9% (97.5% CI 21.8% to 30.3%) in the antibiotic-impregnated VPS group and 25.4% (97.5% CI 20.9% to 29.6%) in the silver-impregnated VPS group.

The distribution of participants (or their parents’ or guardians’) responses to the EQ-5D questionnaires are presented in Figure 6. There was a low return rate of the EQ-5D questionnaire, with combined (EQ-5D-Y, EQ-5D-3L-Proxy and EQ-5D-3L) data available for only about half of the participants. Their responses suggested that there was a general improvement across all dimensions from baseline to the end of the trial, with no clear differences between intervention groups for any given dimension. Similarly, the response rates of participants, their parents or their guardians to the EQ-VAS, which are presented in Appendix 3, Table 50, were also low, but indicate a general trend for improvement from baseline to the end of the trial.

FIGURE 6.

Distribution of participants’ responses to each EQ-5D attribute, by treatment allocated and time. Levels range from 1 to 3, with 3 representing the most severe problem. The numbers of completed responses (n) are reported by intervention group. (a) Mobility; (b) self-care; (c) usual activities; (d) pain or discomfort; (e) anxiety or depression.

Alternative cost-effectiveness and cost–utility analyses

A cost-effectiveness analysis based on the incremental cost per confirmed infection averted indicated that silver-impregnated VPS catheters were dominated by standard VPSs, whereas antibiotic-impregnated VPS catheters were dominant, saving £4059 per 0.030 fewer infection-related VPS failures. Compared with standard VPSs, antibiotic-impregnated VPS catheters save £135,753 per VPS infection avoided (Table 32).

TABLE 32 - Results of alternative cost-effectiveness and cost–utility analyses

Note

Negative ICERs relate to incremental cost and outcome co-ordinates in the south-west quadrant of the cost-effectiveness plane.

VPS	Mean (97.5% CI)				ICER (£ per QALY)
	Total cost (£)	Outcome	Incremental cost (£)	Incremental outcome
Confirmed infections
Antibiotic-impregnated	14,446 (12,660 to 18,054)	0.027 (0.013 to 0.043)	–4059 (–12,567 to 1422)	–0.030 (–0.058 to –0.002)	Dominant
Standard	18,505 (13,872 to 27,274)	0.057 (0.035 to 0.083)	–	–	–
Silver-impregnated	17,331 (14,584 to 22,136)	0.057 (0.038 to 0.080)	–	–	Dominated
Mechanical failures
Standard	14,110 (14,021 to 27,648)	0.092 (0.066 to 0.120)	–	–	–
Silver-impregnated	17,426 (14,682 to 22,445)	0.119 (0.088 to 0.154)	–	–	Dominated
Antibiotic-impregnated	18,749 (12,303 to 17,564)	0.134 (0.103 to 0.167)	–	–	Dominated
Functional failures
Silver-impregnated	17,483 (14,767 to 22,396)	0.069 (0.047 to 0.092)	–1163 (–9349 to 5815)	–0.003 (–0.040 to 0.030)	387,667
Standard	18,646 (13,837 to 27,066)	0.072 (0.048 to 0.101)	–	–	–
Antibiotic-impregnated	14,157 (12,397 to 17,576)	0.084 (0.057 to 0.108)	–4488 (–12,919 to 960)	0.011 (–0.027 to 0.049)	–374,000
Patient factors
Antibiotic-impregnated	14,196 (12,438 to 17,648)	0.008 (0.001 to 0.018)	–4441 (–12,825 to 987)	–0.0006 (–0.015 to 0.012)	7,401,667
Standard	18,638 (13,983 to 27,464)	0.009 (0.001 to 0.018)	–	–	–
Silver-impregnated	17,451 (14,712 to 22,543)	0.009 (0.001 to 0.019)	1186 (–9255 to 5694)	–0.0002 (–0.011 to 0.010)	–3,953,333
Cost–utility analysis based on imputed data
Silver-impregnated	9115 (7596 to 12,682)	1.319 (1.207 to 1.365)	183 (–3035 to 3854)	0.096 (–0.488 to 0.188)	1904
Standard	8932 (7301 to 11,980)	1.223 (1.136 to 1.298)	–	–	–
Antibiotic-impregnated	9643 (7545 to 11,736)	1.250 (1.163 to 1.336)	–	–	Dominated

For the cost-effectiveness measure of incremental cost per mechanical failure averted, both silver-impregnated and antibiotic-impregnated VPS catheters were dominated by standard VPSs, as they were associated with higher rates of mechanical failures and higher costs than standard VPSs. With regards to functional VPS failures, antibiotic-impregnated VPS catheters are both less effective and less expensive than standard VPSs, whereas silver-impregnated shunt catheters cost an additional £387,667 per additional functional failure averted. The opposite was observed when considering the incremental cost per VPS failure due to patient-related factors, although failure rates due to patient influences are much lower, and the reporting of this outcome was less reliable. Antibiotic-impregnated VPS catheters cost an additional £7.4M per failure averted, whereas silver-impregnated VPS catheters save £3.9M per additional failure, each in comparison with standard shunt catheters.

In the cost–utility analysis of trial participants aged ≥ 5 years, and based on multiple imputation to account for missing data, antibiotic-impregnated VPS catheters were dominated by silver-impregnated VPS. Compared with standard VPSs, silver-impregnated VPSs are £183 more costly, and yield 0.096 additional QALYs overall, resulting in an incremental cost of £1904 per QALY gained. The cost-effectiveness acceptability curve showing the probability of each VPS catheter being cost-effective, by a range of cost-per-QALY thresholds, is depicted in Figure 7.

FIGURE 7.

Cost-effectiveness acceptability curves indicating the probability of each VPS catheter being cost-effective (based on the incremental cost per QALY gained) for a range of threshold (willingness-to-pay) values. Vertical lines indicate the usual NICE threshold range53 of £20,000–30,000 per QALY.

Summary of findings

In this trial of patients with hydrocephalus who were undergoing insertion of their first permanent VPS, the infection rates were 6.0% in the standard VPS group, 2.2% in the antibiotic-impregnated VPS group and 5.9% in the silver-impregnated VPS group. Compared with standard VPSs, antibiotic-impregnated VPSs were associated with a significantly lower rate of infection, whereas silver-impregnated VPSs were not. This effect was present across all age categories. There is significant economic benefit for every shunt infection averted. There were notable differences in infection rates for differing aetiologies of hydrocephalus and between different age groups.

Clinical effectiveness

The BASICS trial provides definitive evidence favouring the use of antibiotic-impregnated VPSs to reduce infection. A previously reported underpowered randomised trial15 demonstrated a trend favouring antibiotic-impregnated over standard VPSs, but did not show a statistically significant difference between the two groups in the risk of infection (RR 0.38, 95% CI 0.11 to 1.30). Silver-impregnated shunt catheters have been evaluated only in a randomised trial of EVDs,22 which found a reduction in infection from 21.4% (30/140) with standard shunt catheters to 12.3% (17/138) (p = 0.043) with silver-impregnated shunt catheters, although this is much higher than the UK national reported infection rate (9.3%) for EVDs. 59 The BASICS trial was conceived to evaluate both antibiotic-impregnated and silver-impregnated VPSs before the widespread introduction of these products into routine practice, despite a lack of evidence. The results of this trial will inform neurosurgery practice and choice of VPS for insertion, to the benefit of patients.

In this trial, the classification of VPS infection was determined by the central review committee. The proportion of culture-positive infections was 68.8% in the standard VPS group, 50.0% in antibiotic-impregnated VPS group and 80.6% in silver-impregnated VPS group; there was a lower rate of culture-positive infection with antibiotic-impregnated VPSs than with either the standard-impregnated or the silver-impregnated VPS. The analysis allowed for culture-negative infections to be included when there was sufficient supporting clinical evidence of VPS infection. We postulate that the presence of antibiotic-impregnated shunt catheters reduced the ability to culture organisms in infected VPSs. This trial would have shown an even greater effect in favour of antibiotic-impregnated VPSs if we had included culture-positive infections only.

The reduction of types of infections seen is consistent with the expected microbiological spectrum of the antibiotic-impregnated VPSs, which are especially active against Gram-positive organisms, and were incorporated to prevent staphylococcus species infections. The culture results show a large reduction in staphylococcal infection with antibiotic-impregnated VPSs, compared with the standard and silver-impregnated VPSs, which accounts for the majority of the infections prevented and supports the biological plausibility as clindamycin and rifampicin are preferentially active against Gram-positive organisms. All three VPS types have a similar number of Gram-negative infections, which the antibiotics were not expected to reduce.

It should be noted that the overall VPS failure rate was the same for all groups, although infection was reduced in the antibiotic-impregnated VPS arm. When infection is removed as a cause, the clean non-infected revision rates were slightly higher in the antibiotic-impregnated VPS group. In the post hoc analysis, the only pattern that could be found was a higher rate of valve change in the antibiotic-impregnated VPS group than in the standard and silver-impregnated VPS groups when analysing the components of the shunt that were changed and/or blocked at surgery, as reported by the operating surgeon on the CRF. Interestingly, the valve itself is not impregnated and is common to all shunt types. Unfortunately, in most cases, the valves were not sent for culture as it was not part of most units’ protocol at clean shunt revision. Of note, there were no increased revisions for any group for peritoneal shunt catheters. One of the concerns voiced by some of the surgeons was that silver-impregnated peritoneal shunt catheters were more rigid than the other types, and that this might lead to more peritoneal catheter revisions. This was not borne out in the results.

The reasons for the higher non-infective blockages in the antibiotic-impregnated VPS group are not known. One hypothesis is that the antibiotic-impregnated shunt catheters convert an ‘infected’ shunt revision to an apparently ‘clean’ shunt revision in the following ways:

• Low virulent pathogens are restricted to a biofilm in the valve, not causing a CSF change as there is no ventriculitis and not isolated from the revision CSF as the planktonic bacteria are low in number or not able to grow in the presence of the eluted antibiotics.

• A sufficient change in CSF composition and flow (such as debris or high protein) to promote a blockage in the system such as the valve, which is probably the most vulnerable to content of CSF in terms of blockage in view of the intricate mechanical valve mechanisms. This study has not been powered or designed to answer this question directly, but, in future research, this question will be important.

Of note, from a patient well-being and health economic standpoint, a mechanical clean shunt revision, such as a valve change, will have a minor effect on the patient and a minor cost to the NHS. A clean revision for valve change typically involves an overnight stay only, no brain operation (as the ventricular shunt catheter is not changed) and none of the health implications of suffering from clinical meningitis. In contrast, a shunt infection will involve two brain operations: a first operation to remove the whole shunt system and to insert a temporary ventricular shunt catheter (an average of 2 weeks in a hospital bed for both intravenous and intrathecal antibiotics, and the health implications of clinical meningitis) and a second operation to re-insert a whole new shunt system including a ventricular shunt catheter.

In addition to the above, if a patient has ventricular shunt catheter change at shunt revision [as opposed to valve change or peritoneal shunt catheter change, (not a brain operation)], they will lose their driving licence for 6 months because of the increased risk of epilepsy after ventricular shunt catheter revision (as it is a brain operation).

Cost effectiveness

Complications associated with VPS failures are expensive to manage. 32–38 Consequently, the use of VPS catheters, which result in fewer complications, even if more expensive to purchase, could be cost-effective or yield cost savings to the NHS. The economic analysis within the BASICS trial estimated that, although antibiotic-impregnated shunt catheters are about twice the price of standard silicone shunt catheters, this upfront cost could be justified by the reduced infection rate and associated cost savings of further surgery and prolonged hospital care. Based on the primary economic outcome of incremental cost per VPS failure (due to any cause) averted, there appeared to be large potential savings for additional cases of VPS failures with silver- and antibiotic-impregnated VPSs compared with standard VPSs. This conservative estimate does not assume equivalence, despite no difference between groups in VPS failure rate. Had a cost-minimisation analysis been considered appropriate, the saving per patient having an antibiotic-impregnated rather than a standard VPS would be £4515 (97.5% CI £140 to £9170).

In this context, the secondary outcome based on the incremental cost per confirmed infection averted is more relevant, as well as for comparison with previous such analyses, which use the same outcome measure. Notably, antibiotic-impregnated VPS catheters were dominant, saving £4059 per 0.030 fewer infection-related VPS failures. Compared with standard VPSs, antibiotic-impregnated VPS catheters save £135,753 per VPS infection avoided. In terms of determining technical efficiency, no cost-effectiveness threshold is necessary, given the dominance of antibiotic-impregnated VPS catheters.

For the purposes of informing decisions on allocative efficiency, the cost–utility analysis of participants aged ≥ 5 years indicated that antibiotic-impregnated VPSs were dominant, whereas the incremental cost per QALY gained with silver-impregnated VPSs was £1904, which is within the threshold applied by NICE for the determination of cost-effectiveness. However, the cost–utility analysis was limited, with respect to missing data and the exclusion of participants who were at the highest risk of VPS infections.

The analyses are robust to a range of assumptions and analytic approaches. Important subgroup analyses indicate differences in cost-effectiveness by age. The risk of infection in the BASICS trial was highest in paediatrics; it was lower for adults and was particularly low for the elderly. Some patterns also emerged with the aetiology of hydrocephalus; for example, congenital causes and post-haemorrhagic hydrocephalus (both prevalent in the paediatric age group) resulted in much higher infection rates than those observed for other aetiologies seen in older patients. Although the primary economic analysis was based on all-cause VPS failures, subgroup analyses demonstrated higher cost-effectiveness in paediatrics than in adult populations.

Generalisability and cost impact

In the context of the NHS, there should be a major health benefit and cost-saving impact by routinely adopting antibiotic-impregnated VPSs for all first-time VPS insertions for hydrocephalus.

Clearly, in paediatrics, for whom the effect of reduction in shunt infections is greatest, the economic benefit will also be the most significant. For example, for every 100 paediatric new VPS insertions, using antibiotic-impregnated rather than standard shunt catheters should avert up to six or seven shunt infections, which translates to a potential cost saving of between £814,000 and £950,000. Even in those aged > 65 years, averting one infection per 100 shunts could potentially save £135,753.

Such savings may well translate to equivalent health economies, such as the in the USA, but clearly the potential value and savings will vary in different health models internationally.

Strengths and limitations

The strengths of this trial are that (1) infections were centrally classified by a panel blinded to treatment allocation, thereby removing the risk of treating-surgeon bias; (2) participant retention was very high owing to the nature of the intervention and the primary outcome (patients with infected VPSs are always re-admitted to hospital); (3) patient withdrawal was low (n = 53, 3.3%) so it is unlikely that events were missed; (4) participants were recruited across the whole of the UK and the Republic of Ireland to encompass all ages and socioeconomic classes; and (5) the trial sample size was large.

Some limitations of the trial should be noted. First, it was not possible to blind the treating neurosurgeon to the VPS type, because the physical appearance of each shunt catheter is distinctive. Participants were blinded to the shunt catheter type, and shunt catheter type was not recorded in patient records. The majority of VPS revisions and removals for infection happen as emergencies and are managed by the emergency neurosurgery team. Therefore, the likelihood of the same neurosurgeon who inserted the VPS being involved in the decision to remove the VPS was low, given the work rotas of neurosurgical staff. Furthermore, there was high agreement (95.7%) in the classification of VPS infection between treating surgeons and the central review panel, suggesting that any bias that the treating surgeon may have had did not affect the study conclusion. Second, ventriculoatrial and ventriculopleural shunts were excluded, although we postulate that the results are translatable to patients undergoing these procedures. Finally, the return rate for patient-reported outcomes was low, thereby limiting the analysis of the impact of VPS infection.

Limitations to the economic evaluation include a lack of detailed costing for revisions, and failure to obtain Hospital Episode Statistics data from NHS Digital, as per protocol. Reliance on PLICS data mitigated these limitations, but, as with other data on costs, there was a high degree of missingness due to some centres not being able to share electronic data and patient questionnaires not being returned. This was addressed using robust methods (multiple imputation), but still may have introduced bias to the analysis.

Safety

The data suggest that there is no increased health risk in using impregnated shunt catheters. There were no serious AEs. All of the AEs that were seen were expected events predominantly related to expected shunt revisions and/or due to re-admission for expected known complications associated with VPS surgery; no differences were seen between catheter types.

Implications for practice and health care

From a patient perspective and that of the treating clinician, the hospital and the health service, every effort to reduce shunt infection should be made, and health technologies such as antibiotic-impregnated shunt catheters, with their potential to reduce VPS infections, deserve proper evaluation through appropriately planned and powered trials.

Having demonstrated a marked reduction in such infections, with all of the potentially catastrophic and life-changing health sequalae that result from each infection, the BASICS trial has provided evidence to support that the adoption of antibiotic-impregnated shunt catheters is justified in all patients receiving their first VPS in the UK. The increased upfront cost of the impregnated shunt catheters is offset by the added health economic benefit demonstrated in the health economic analysis and in a previous publication. 39 The benefits and implications, both from an efficacy and health economic point of view, are most pronounced in younger patients.

A wider discussion and analysis, particularly from a health economics point of view, is required when considering recommendations and implications globally.

Conclusions

In conclusion, antibiotic-impregnated shunt catheters significantly reduce the VPS infection rate and the probability of infection, compared with standard silicone shunt catheters in all age groups, whereas silver-impregnated shunt catheters do not. The routine use of these shunt catheters would result in a significant cost saving to the NHS.

These results support the use of antibiotic-impregnated shunt catheters in the treatment of patients undergoing insertion of a first VPS for hydrocephalus.

Implications for future research

To our knowledge, the BASICS trial is the largest-ever prospective randomised trial for VPSs in hydrocephalus worldwide.

A critical value-added aspect of the BASICS trial was the development of systematic clinical sample collection (of CSF and blood) from participants undergoing VPS insertion. We have established a unique neurosurgical patient sample collection, which will enable us to identify surrogate markers of VPS infection in CSF or blood using molecular techniques (transcriptomics/proteomics); to assess whether or not infection may be associated with shunt failure cases (by detection of pathogen nucleic acid via next-generation sequencing); and to begin to explore whether or not host differences contribute to the different rates of infection in children, compared with adults. Proteomic analysis of CSF has already been undertaken to identify a series of candidate surrogate markers of neurosurgical infection. These candidate markers are now being re-tested among the BASICS trial samples to assess their accuracy in identifying confirmed infection. The BASICS trial sample collection, stored at the University of Liverpool, is a valuable resource to help answer the ongoing questions around VPS infection identified in the BASICS trial (as outlined in this paragraph) and to support future additional studies.

A large number of data was collected regarding aetiology, surgical techniques, types of valves and the technology used, which can guide future management and trials. We plan to undertake further analysis exploring patterns related to mechanical shunt failure. This is led by the results and will, therefore, be post hoc. We also plan to create a detailed shunt failure model and predictive score. Failure and infection rates across different aetiologies of hydrocephalus will be analysed further.

Surgical techniques and factors that potentially affect failure rates will be further analysed, such as the use of neuronavigation and the placement of the ventricular shunt catheter, surgical site infection information collected at the time of surgery and seniority of the surgeon. This may feed further questions and prospective randomised studies.

From a methodological perspective, the economic evaluation was significantly affected by missing data. However, organisational familiarity with the procedures and requirements for accessing routine data, post General Data Protection Regulation,60 should facilitate Hospital Episode Statistics data access from NHS Digital in future trials. Acute trusts are now mandated to produce PLICS data, and early engagement with hospital finance offices should resolve many of the difficulties that were faced in the BASICS trial. Poor response rates to questionnaires is a common problem in self-reported resource use. Further research into methods to mitigate this ought to focus not just on technical solutions, but also on operational solutions, including a broader appreciation that achieving complete data on economic outcomes should be regarded as having equal importance to achieving complete data on primary clinical end points. This may be especially relevant for time-to-event analyses, in which costs and disutilities are most likely to accrue after such events.

Contributions of authors

Conor L Mallucci (https://orcid.org/0000-0002-5509-0547) (Chief Investigator and Consultant Neurosurgeon) developed the trial protocol in collaboration with co-investigators. He oversaw the delivery of the trial, prepared trial update reports and oversaw clinical aspects of the statistical analysis plan and clinical interpretation of the trial data. He co-led the preparation of the final report (drafting, reviewing and editing). He was chairperson of the TMG and a member of the central review panel.

Michael D Jenkinson (https://orcid.org/0000-0003-4587-2139) (Co-Chief Investigator and Consultant Neurosurgeon) developed the funding application and trial protocol in collaboration with co-investigators. He contributed to clinical interpretation of the trial data and to the preparation of the final report (drafting, reviewing and editing).

Elizabeth J Conroy (https://orcid.org/0000-0003-4858-727X) (Trial Statistician) contributed to protocol development and data capture methods, wrote the statistical analysis plan, undertook the final statistical analysis under the supervision of Carrol Gamble, prepared data for reports throughout the trial, prepared data tables and figures for final report and co-led the preparation of the final report.

John C Hartley (https://orcid.org/0000-0003-0503-5985) (Trial Microbiologist) contributed to protocol development, data capture methods, central classification of infections, preparation and review of progress and final reports.

Michaela Brown (https://orcid.org/0000-0002-7772-271X) (Senior Statistician) contributed to protocol development and data capture methods, proposed statistical analysis methods, approved the statistical analysis plan and oversaw trial monitoring activities.

Tracy Moitt (https://orcid.org/0000-0002-5579-996X) (Senior Trials Manager) contributed to protocol development, gave guidance and support on all aspects of governance and trial delivery and supported the preparation of progress reports. She also contributed to the final report (drafting and reviewing).

Joanne Dalton (https://orcid.org/0000-0003-1199-3047) (Data Manager) supported the central review panel in its assessment.

Tom Kearns (https://orcid.org/0000-0001-5416-8376) (Trial Co-ordinator) contributed to protocol development, all aspects of governance and study delivery and preparation of progress reports. He also contributed to the final report (reviewing).

Michael J Griffiths (https://orcid.org/0000-0003-1851-6155) (Senior Lecturer in Paediatic Neurology) co-ordinated and developed the sample collection substudy; contributed to protocol development; oversaw sample recruitment, collection and storage; and contributed to the preparation of the final report (reviewing, editing).

Giovanna Culeddu (https://orcid.org/0000-0001-5032-4255) (Study Health Economist) contributed to protocol development and data capture methods, undertook the economic analysis under the supervision of Dyfrig Hughes and contributed to the final report (drafting).

Tom Solomon (https://orcid.org/0000-0001-7266-6547) (Professor of Neurology) helped devise the trial, assisted with obtaining grant funding and reviewed the final report.

Dyfrig Hughes (https://orcid.org/0000-0001-8247-7459) (Senior Health Economist) led the economic evaluation, contributed to protocol development and data capture methods and contributed to the writing of the report (drafting, reviewing and editing).

Carrol Gamble (https://orcid.org/0000-0002-3021-1955) (Statistical Lead) led trial design and developed the funding application, trial protocol and data capture methods in collaboration with co-investigators. She led blind review of the data and supervised the final analysis. She also contributed to the preparation of the final report (drafting, reviewing and editing).

All authors were members of the TMG.

Publications

Jenkinson MD, Gamble C, Hartley JC, Hickey H, Hughes D, Blundell M, et al. The British antibiotic and silver-impregnated catheters for ventriculoperitoneal shunts multi-centre randomised controlled trial (the BASICS trial): study protocol. Trials 2014;15:4.

Mallucci CL, Jenkinson MD, Conroy EJ, Hartley JC, Brown M, Dalton J, et al. Antibiotic or silver versus standard ventriculoperitoneal shunts (BASICS): a multicentre, single-blinded, randomised trial and economic evaluation. Lancet 2019;394:1530–9.

Data-sharing statement

All requests for data should be sent to the corresponding author. Access to available anonymised data may be granted following review.

Patient data

This work uses data provided by patients and collected by the NHS as part of their care and support. Using patient data is vital to improve health and care for everyone. There is huge potential to make better use of information from people’s patient records, to understand more about disease, develop new treatments, monitor safety, and plan NHS services. Patient data should be kept safe and secure, to protect everyone’s privacy, and it’s important that there are safeguards to make sure that it is stored and used responsibly. Everyone should be able to find out about how patient data are used. #datasaveslives You can find out more about the background to this citation here: https://understandingpatientdata.org.uk/data-citation.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health and Social Care.