Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Protocol - Intervention

Ventriculo‐peritoneal shunting devices for hydrocephalus

This is not the most recent version

Collapse all Expand all

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

To compare the effectiveness and safety of conventional and complex shunt devices for CSF diversion in people with hydrocephalus.

Background

Description of the condition

A specialised fluid system circulates around the central nervous system, providing structural and nutritional support for the brain and spinal cord. The fluid is a modified filtrate of the blood called cerebrospinal fluid (CSF), and is produced by groups of ependymal cells, found in the brain and spinal cord, and concentrated in the four choroid plexuses situated in the cerebral ventricles (four normal cavities) of the brain (Champney 2016; Wolburg 2010). The total CSF volume at any time is approximately 150 mL, of which 125 mL is produced inside the brain. The rate of production is 0.35 mL to 0.40 mL per minute. CSF flows from the lateral ventricles to the third ventricle and then through the aqueduct to the fourth ventricle. After that, it leaves the ventricular system through three foramina: two laterally placed (foramina of Luschka) and one medially placed (foramen of Magendie) to circulate in the subarachnoid space (which is a layer of connective tissue between dura mater and the brain surface) and finally it is absorbed into the venous system, through the arachnoid villi (Mancall 2011).

Hydrocephalus, known colloquially as 'water on the brain' or 'dropsy on the brain', is a common neurological disorder, caused by a progressive accumulation of CSF within the intracranial space that can lead to increased intracranial pressure, enlargement of the ventricles (ventriculomegaly) and, consequently, to brain damage. It results from problems with CSF production, circulation or reabsorption. The incidence of hydrocephalus in the USA lies between one and 32 cases per 10,000 births, depending on the definition used and the population studied (Jeng 2011; Kahle 2015). The most recent estimate of incidence comes from a large, population‐based study of idiopathic (unknown cause) infantile hydrocephalus in Denmark over a 30‐year period, which documented 1.1 cases per 1000 births (Munch 2012; Tully 2014).

Hydrocephalus can be classified into communicating and non‐communicating types on the basis of its pathophysiology. The non‐communicating type, also called obstructive hydrocephalus, can appear in the early fetal stage as part of certain congenital malformations (such as aqueductal stenosis, Chiari malformation), or associated with a broad spectrum of brain malformations, which prevent the normal flow of CSF around the brain, and its subsequent reabsorption. Hydrocephalus may also occur secondary to an obstruction of the CSF circulation caused by a ventricular tumour. By comparison, communicating hydrocephalus results from a deficit in CSF reabsorption, for example, in post‐meningitis hydrocephalus. Haemorrhage in preterm infants can cause hydrocephalus by two mechanisms: acute obstruction by a blood clot or a delayed reabsorption caused by thickening of arachnoids, the inner layer of the meninges (Petre 2010). Although infrequent, hydrocephalus can also be associated with an overproduction of CSF, as in the case of a choroid plexus papilloma, which is a type of benign intraventricular tumour.

Normal Pressure Hydrocephalus (NPH) is a condition first reported in adults that refers to a chronic symptomatic hydrocephalus with a normal CSF pressure (i.e. less than 18 mmHg) (Adams 1965). The condition is defined by a clinical triad of symptoms: dementia, gait difficulties, and urinary urge incontinence in the presence of ventriculomegaly documented by computed tomography (CT) or magnetic resonance imaging (MRI) (Ziebell 2013; Halperin 2015). There are signs in the magnetic resonance images that differentiate NPH from other types of hydrocephalus, specifically, the amount of interstitial oedema surrounding the lateral ventricles (which should be minimal or absent), and the presence of deep white matter ischaemia also known as small vessel ischaemia or leukoaraiosis (Bradley 2015). In order to avoid overlap with other Cochrane Reviews, this review will not include people with NPH.

Signs and symptoms of hydrocephalus depend on the cause of the condition and time of presentation. In congenital hydrocephalus, patients present at birth with macrocephaly (overly large head), scalp vein distension, a bulging fontanelle and diastasis of sutures (stretching of fibrous joints between bones in the skull). These can be correctly evaluated using reference tables developed by measuring the circumference of the head of many children at different ages that are considered to establish the normal parameters of head growth. Using this simple tool, rapid head growth as a result of hydrocephalus can be detected clinically (O'Neill 1961). Patients with hydrocephalus may also present with Parinaud´s phenomenon, which is an inability to raise the eyes upwards, also known as 'Sunset Sign', caused by an increased pressure on the tectal plate, which is the area of the midbrain where the oculomotor nerve originates. Symptoms in patients with acquired hydrocephalus (obstructive, post‐infectious, or post‐haemorrhagic) constitute the intracranial hypertension syndrome which includes headache, vomiting and drowsiness (Petre 2010). The appearance of these signs and symptoms of intracranial hypertension is what differentiates developing hydrocephalus from benign ventriculomegaly (dilation of the lateral ventricles in the brain without causing an increase in the CSF pressure and therefore, without risk of brain damage).

In addition to physical examination and assessment of symptoms, confirmation of diagnosis occurs by means of imaging studies that show enlargement of ventricles. For newborns and infants, ultrasonography is frequently used because it can be done at the bedside and is radiation‐free. Nevertheless, it is an operator‐dependent technology and does not assess the posterior fossa (intracranial space occupied by the cerebellum and brain stem), which sometimes makes it difficult to establish the cause of the hydrocephalus. For older children and adults, CT or MRI of the brain is usually used (Carey 1994).

Since Hippocrates in the fifth century BC, a poor understanding of the pathophysiology of hydrocephalus contributed to the failure of the early, and occasional attempts at therapy (Whytt 1768). In 1908 Payr introduced the first ventriculo‐venous shunt for drainage, which consisted of a vein graft that led from the ventricle directly into the sagittal sinus and jugular veins (Mccullough 1990). In the same year, Kaush used a rubber conduit to drain the lateral ventricle into the peritoneal cavity (Kausch 1908), but the medical community received this innovation with little enthusiasm. Research for more effective treatments followed, and eventually the placement of intracranial shunts was tested. Efforts to drain the excess CSF into other body cavities have also been considered; for example, Matson and colleagues at Boston Children´s Hospital first reported a ureteral diversionary procedure (Matson 1949).

The early twentieth century was a period in which knowledge of hydrocephalus, its diagnosis and strategies for treatment evolved. Between January 1938 and December 1957, an observational, case‐series study, documented the progress of 182 patients with congenital or acquired (early after birth, due to trauma or infection) hydrocephalus, who were not operated on. By the end of the study, only 81 patients were alive, and these had spontaneously arrested hydrocephalus, that is, a gradual slowing of the rate of head growth with a reduction in fontanelle tension and improvement in the patient's general condition. Patients with comorbidities, such as myelomeningocele or Chiari malformation, had the worst prognosis (Laurence 1962). Since the introduction of an effective therapy for the symptoms of hydrocephalus in 1956, when the first shunt system became available, mortality rates have fallen from around 45% to 53% ‐ depending on the case series concerned ‐ to 15% (Hagberg 1962; Jansen 1985; Laurence 1967; Yashon 1963). Morbidity rates have also significantly improved, with up to 42% of treated hydrocephalus patients enjoying a normal lifestyle after receiving shunts (Hirsch 1994).

The breakthrough that launched the modern era of surgical treatment for hydrocephalus was the introduction of valve‐regulated shunts and biocompatible synthetic materials in 1952 (Lifshutz 2001). That same year, Nulsen, Spitz and Holter, reported the successful use of a ventriculo‐jugular shunt regulated by a spring and ball valve (Drake 1995), and almost simultaneously, Pudenz and colleagues, created a silicone one‐way slit valve (Pudenz 1957). The development of the valve system and the availability of new biocompatible materials made it possible to divert CSF safely and reliably, and to avoid the many complications associated with unregulated CSF drainage.

Despite this degree of innovation, these devices were not exempt from shunt failure, which represents a serious complication. The most common causes of shunt malfunction include over‐ and under‐drainage, mechanical mismatch (Inadequate selection of the pressure in a programmable valve or inadequate selection of the valve in relation to the type of hydrocephalus), occlusion (blocking) of the shunt and valve failure. These problems are largely resolved by replacing the valve or parts of the shunt, and sometimes by finding alternative drainage locations (for example, using the circulatory system or the pleural cavity in patients in whom the peritoneal cavity can not be used because of peritoneal inflammatory diseases or surgical sequelae) as well as with innovations in valve designs (Stein 2008).

Although implant of ventriculo‐peritoneal shunts is the standard treatment for patients with hydrocephalus, in the 1990s, an additional method to treat hydrocephalus without the use of a prosthetic device became available. Endoscopic third ventriculostomy (ETV) is a minimally invasive procedure originally introduced by Dandy in 1922. During the procedure a neuroendoscope is introduced, through a burr hole, into the lateral cerebral ventricles and a hole is made within the third ventricle to enable intracerebral ventricles to communicate with the subarachnoid spaces. This procedure leads to a more physiological circulation of CSF and is a well established treatment for noncommunicating hydrocephalus in some neurosurgical units (Schroeder 1999). Risk of harms related to treatment is always present, depending on the procedure. Shunts may have complications related to placement, such as brain haemorrhages, brain damage, infections, coma, and even exceptionally, death (Smith 2004). In addition, there is an important and permanent risk of shunt malfunction. On the other hand ETV may also carry serious complications including vascular injuries, hypothalamic injury (brain damage), and occasional death, as has been reported (Hader 2002; Drake 2006). Due to the lack of clinical trials comparing both strategies, the decision about whether to use ETV or ventriculo‐peritoneal shunts remains a topic of considerable debate (Cheng 2015; Limbrick 2014).

Description of the intervention

This review will focus on the placement of various types of ventriculo‐peritoneal shunts for the treatment of hydrocephalus. We define a ventriculo‐peritoneal shunt as a system composed of a ventricular catheter, which is usually inserted into one of the cerebral lateral ventricles and is attached to both a valve and a distal catheter, which is implanted within the peritoneal cavity (abdomen) where CSF is finally reabsorbed (ICD‐9‐CM 2004; Patwardhan 2005).

Ventriculo‐peritoneal shunting is a complex surgical intervention. This procedure is usually performed by an experienced neurosurgeon, in the sterile conditions of an operating theatre, and with the use of general anaesthesia. In the most commonly used surgical technique, the patient lies on the operating table in the supine position (lying on his back) with his head turned to one side. Using sterile instruments, the surgeon makes two incisions. One is made on the scalp in order to expose the posterior parietal part of the skull (above and behind the ear), followed by a burr hole through the skull and a small opening in the dura (membrane that covers the brain), to access the dilated lateral ventricle with a catheter. The second incision is made in the skin of the abdomen; from there a tunnel is formed under the skin towards the cephalic incision, and through this tunnel the distal catheter is passed. This catheter is subsequently connected to the valve part of the shunt. The ventricular catheter is also connected to the valve, which, in turn, is fixed to the periosteum (connective tissue around the skull). Finally, the distal end of the abdominal catheter is placed into the peritoneal cavity, into which the CSF is drained (Warf 2005).

Valves are the most important part of the system, and there are a variety of valve systems available that offer fixed pressures, anti‐siphon devices, and programmable and flow regulation capabilities (Ames 1967; Kaiser 1992; Lumenta 1990; Ojemann 1968).

Fixed differential pressure (DP) valves are considered to be the first generation of valves. When pressure builds up inside the tubing, a slit is forced to open and CSF is allowed to flow outwards. Only unidirectional flow is permitted since an increase in external pressure closes the slit. The opening pressure is determined by the thickness of tubing walls. They are commonly classified according to the opening and closure pressures: low pressure (20 mm H₂O to 40 mm H₂O), medium pressure (40 mm H₂O to 70 mm H₂O) and high pressure valves (80 mm H₂O to 100 mm H₂O) (Post 1985)

Second generation valves include DP valves with flow‐regulating devices, valves with anti‐siphon mechanisms and programmable DP valves. Flow‐regulating devices limit CSF flow through the valve by progressively narrowing its orifice in response to increasing intracranial pressure (ICP), as a pressure sensitive ring moves along a variable‐diameter rod. Its aim is to prevent both postural and vasogenic over‐drainage occurring during rapid eye movement (REM) sleep, physical exertion, coughing, and other physiological conditions (Hanlo 2003).

Through an increase in flow resistance, anti‐siphon devices counteract the effect of hydrostatic negative pressure when the patient stands in a vertical position. The aim of this system is to provide performance characteristics that simulate those of normal CSF absorption, while allowing the regulation of ICP in a fashion similar to that seen in healthy individuals when they change their head position from horizontal to vertical (Watson 1994; Baird 2014).

Programmable variable pressure valves allow selection of different opening pressures (between 30 mm H₂O and 200 mm H₂O in intervals of 10 mm H₂O). The pressure can be selected through a percutaneous noninvasive magnetic programmer that manages valve pressure according to the patient's clinical and radiological evolution in terms of ventricular size. This permits avoidance of over‐drainage and its consequences, which include slit (shaped) ventricles (accompanied by headaches) and retardation of cranial vault growth (secondary craniosynostosis) (Xu 2013; Baird 2014; Miyake 2016).

Finally, third generation valves include DP programmable and anti‐siphon or gravitational systems combined in the same device. The DP unit allows selection of an opening pressure between 0 to 20 cm H₂O, with the possibility of transcutaneous adjustment. Unlike anti‐siphon systems, which use a fixed pressure, the gravitational unit can be set to opening pressures of 15 cm H₂O, 20 cm H₂O, 25 cm H₂O, or 30 cm H₂O, which cannot be changed after placement. This kind of device increases the opening pressure of the shunt by blocking the inlet flow using a gravity‐assisted ball bearing. These devices must be placed vertically in order to counteract the siphoning effect of negative hydrostatic pressures when the patient is standing (Rohde 2009).

How the intervention might work

The balance in production, circulation, and reabsorption of the CSF has a key role in the Monro‐Kelly hypothesis (Champney 2016; Lee 2009). This states that a stable ICP is the result of a rigid sphere (cranium) occupied by a non‐compressible volume of brain tissue, blood and cerebrospinal fluid, which must remain constant. Although physiological fluctuations can occur, these are related to reciprocal changes in the amount of blood and CSF at a given time (Han 2005). ICP varies during the day, normal values are between 7 mmHg to 15 mmHg in a supine adult. Fluctuations are regulated primarily by cranial venous system blood volumes (CVSBV). As a result of gravity, CVSBV diminish when a person is upright. A drop in blood volume causes a retraction in brain tissues away from the normal CSF filled ventricles, causing a drop in ICP. Production of CSF also fluctuates during the day, but as there is a balance between production and reabsorption, the ventricular size remains stable. There are also variations in CSF flow rate at night, with periods of highest flow rate occurring during REM sleep (Watson 1994).

Normally, CSF is produced, circulates, and is reabsorbed within the craniospinal space. Changes in body position between the horizontal and vertical entail transfer of the CSF and blood from the cranial compartment to the spinal compartment in a balanced fashion. When a ventriculo‐peritoneal shunt is in place in a person with hydrocephalus, this balance is broken and hydrostatic pressure plays an important role (Kurtom 2007). Valves provide a resistance mechanism within the shunt system, which serves to control ICP more than CSF flow in patients with hydrocephalus (Miyake 2016).

In the developmental stage, shunt systems were synonymous with CSF diversion to provide a rapid decrease in ICP, with very little attention given to the cause of hydrocephalus. This approach produced good early results in terms of resolution of ventricle enlargement and reduction of ICP, but complications related to over‐drainage (such as subdural haematomas), shunt failures due to catheter obstruction, multiples shunt revisions and infections did appear (Symss 2015).

Valve technology has evolved to demonstrate a more adequate understanding of the physiological balance that needs to be achieved in the production, circulation and reabsorption of CSF. As a result, modern valves decrease ICP, as well as reducing brain damage and shunt‐related complications. Since the first fixed‐pressure shunts, like those of Holter and Pudenz (Drake 1995; Pudenz 1957), shunt systems, and especially valves, have evolved to incorporate flow‐regulating, anti‐siphon and gravitational components that are designed to avoid the over‐drainage that can be caused by changes in patient position (Czosnyka 1998).

Why it is important to do this review

Hydrocephalus is a common chronic neurological disease that places a significant burden both on individuals and society as a whole.The cost of its medical treatment is not completely understood, but appears to be increasing (Pikus 1997). The National Inpatient Sample (NIS) database, a nation‐wide survey in USA healthcare facilities, identified that in 2000 the three most common causes of admission for hydrocephalus were: shunt malfunction (40.7%), non communicating hydrocephalus (16.6%), and communicating hydrocephalus (13.2%). The most common procedures that resulted from these hospitalisations were placement of a new shunt (43.4%) and shunt replacement (42.8%). The total costs of shunt‐related procedures were estimated to be approximately one billion USD (Patwardhan 2005).

Additionaly, a longitudinal study between 1997 and 2003 found an increasing proportion of older children were admitted with shunt malfunctions, and when hydrocephalus treatment and its complications were compared with other chronic illness, such as cystic fibrosis, the inpatient utilisation (admissions, length of stay and hospital charges) were higher for hydrocephalus (Shannon 2011).

In low‐ and middle‐income countries, the treatment of hydrocephalus may encounter additional challenges due to the economic constraints that patients and families face to afford the costs of acquisition, transportation and access to proper care (Warf 2005).

Although ventriculo‐peritoneal shunting has been the most widely used treatment for hydrocephalus in the twentieth century, it is unclear whether the outcomes of shunting have improved significantly over time, despite improvements in the understanding of CSF physiology and the technological advancements in valve design. The Hydrocephalus Clinical Research Network concluded that in comparison with patients treated in the 1990s, there has been a reduction in the risk of complications ‐ assessed as the time to first shunt failure ‐ by about 18% (Kulkarni 2013). However, the reasons for this improvement are not obvious, and thus, open to interpretation.

Despite this, long‐term health and quality of life (QoL) outcomes after shunt implant have been highly variable, and range from patients leading a near normal life to those with physical, cognitive, social and emotional impairments associated with disability. These outcomes rely not only on the cause of hydrocephalus, but also on treatment‐related complications such as infections, haemorrhages leading to brain haematomas, brain injury due to Inadequate positioning of the catheter, coma, and occasionally, death (Smith 2004). They may appear shortly after surgery, and generally within six months of shunt implantation or revision (Kulkarni 2007; Peters 2014; Sciubba 2007).

The main goal of physicians, researchers and engineers when developing a new valve technology is to improve clinical outcomes ‐ both physical and mental ‐ as well as reducing the likelihood of shunt failure. A Cochrane systematic review published in 2013, evaluated the effectiveness of flow‐regulated versus differential pressure‐regulated shunt valves for adults with normal pressure hydrocephalus (Ziebell 2013), but it was limited to two types of valves in a specific group of participants. These reasons, and the fast pace of technological progress, prompt the need for a high‐quality synthesis of the evidence for the effectiveness and safety of common, fixed‐pressure and other more complex ventriculo‐peritoneal shunt devices for people with hydrocephalus with different causes.

Objectives

To compare the effectiveness and safety of conventional and complex shunt devices for CSF diversion in people with hydrocephalus.

Methods

Criteria for considering studies for this review

Types of studies

We will include randomized controlled trials and quasi‐randomised controlled trials. We will include studies regardless of their publication status or language of publication.

Types of participants

People with clinical and imaging diagnosis of communicating or non communicating, congenital or acquired hydrocephalus, irrespective of cause or subtype.

No restrictions will be made with respect to age, except for exclusion of adults considered to have normal pressure hydrocephalus due to clinical symptoms (dementia, gait dyspraxia, and incontinence) and radiological signs (ventricular enlargement disproportionate to the degree of cerebral atrophy), to prevent overlapping with other Cochrane Reviews (Esmonde 2002; Ziebell 2013).

Types of interventions

We will include studies with head‐to‐head comparisons between the different type of devices for ventriculo‐peritoneal shunting listed below:

  • programmable valves: these allow selection of an opening pressure value and are regulated by an external device, with or without an anti‐siphon element;

  • gravitational (gravity‐assisted) valves: these combine a differential pressure unit that allows the selection of an opening pressure between 0 and 20 cm H₂O, with a gravitational unit to counteract siphoning, while the opening pressure can be regulated using telemetry;

  • valves with an anti‐siphon device: these counteract the effect of hydrostatic negative pressure when the patient's body is in the vertical position, through flow resistance;

  • flow‐regulating valves: this type of valve adaptively limits CSF flow through the valve by narrowing its orifice progressively in response to increasing ICP, to avoid the occurrence of both postural and vasogenic over‐drainage during normal physiological conditions;

  • conventional valve: CSF is drained only when ICP rises above the fixed valve pressure.

Types of outcome measures

We will not use measurement of the outcomes assessed in this review as eligibility criteria.

Primary outcomes

  • Treatment failure, defined as progression of hydrocephalus detected by signs or symptoms of intracranial hypertension or imaging studies that led to shunt revision, or both.

  • Adverse events: mortality and morbidity related to the shunt placement (subdural haematomas, slit ventricles (i.e. over‐drainage), obstruction, secondary craniosynostosis and microcephaly, infections).

  • Mortality for all causes.

Secondary outcomes

  • Quality of life and health‐related quality of life measured according to a validated questionnaire, such as: the Medical Outcomes Study (MOS) 36‐Item Short‐Form Health Survey (SF‐36) (Ware 2012), and EuroQOL‐5D (EQ‐5D) for adults (Williams 1990), and the Child Health Questionnaire (CHQ) (Landgraf 1999), or the Hydrocephalus Outcome Questionnaire (HOQ) for children (Kulkarni 2004).

  • Ventricular size reduction measured by cranial sonography or CT scan.

  • Head circumference measured before and after placement of the ventriculo‐peritoneal shunt (occipitofrontal head circumference).

Search methods for identification of studies

We will perform a comprehensive search with no restrictions on the language of publication or publication status.

Electronic searches

The Information Specialist will search the Trials Register of the Cochrane Multiple Sclerosis and Rare Diseases of the CNS Group, which, among other sources, contains trials from: the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, Embase, CINAHL, clinical trials registries including: clinicaltrials.gov and the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) search portal (apps.who.int/trialsearch/). The keywords for this search are listed Appendix 1.

Information on the Group's Trials Register and details of search strategies used to identify trials can be found in the 'Specialised Register' section within the Cochrane Multiple Sclerosis and Rare Diseases of the Central Nervous System Group's module.

In addition, the authors will search the following sources for eligible reports in any language (Appendix 2):

  • Cochrane Central Register of Controlled Trials (CENTRAL; most recent issue);

  • Ovid MEDLINE (1946 to present);

  • Ovid Embase (1974 to present);

  • Latin American and Caribbean Health Sciences Literature database (LILACS; 1980 to present);

  • ClinicalTrials.gov (clinicaltrials.gov); and

  • WHO International Clinical Trials Registry Platform (apps.who.int/trialsearch/).

We will combine the Ovid MEDLINE search with the Cochrane Highly Sensitive Search Strategy for identifying randomized trials in MEDLINE, Ovid format (Lefebvre 2011).

This search strategy will be adapted to search CENTRAL, Ovid Embase and LILACS. Also, we will combine the Ovid Embase search with the trial filter for therapy, best balance of sensitivity and specificity, developed by the Health Information Research Unit (HIRU).

If we detect additional relevant keywords during any of the electronic or other searches, we will modify the electronic search strategies to incorporate these terms and document the changes.

Searching other resources

We will try to identify other potentially eligible trials or ancillary publications by searching the reference lists of retrieved included trials, reviews, meta‐analyses and health technology assessment reports. We will also contact study authors of included trials to identify any further studies that we may have missed. We will contact drug and device manufacturers for ongoing or unpublished trials.

To identify additional studies we will review proceedings of the annual meetings of the International Society of Pediatric Neurosurgery (ISPN) and the European Society of Pediatric Neurosurgery (ESPN) for the last three years.

Handsearches will be considered following the electronic searches, particularly when the abstracts from relevant meetings may not be available electronically.

We will also search other sources of grey literature, such as:

Data collection and analysis

Selection of studies

Independently, two review authors (SP, VV, or JVAF) will scan the abstract, title, or both, of the records retrieved, to determine which studies should be assessed further. Two review authors (SP, VV, or JVAF) will investigate all potentially relevant records as full text, map records to studies, and classify studies as included studies, excluded studies, studies awaiting classification, or ongoing studies in accordance with the criteria for each provided in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a). We will resolve any discrepancies through consensus or recourse to a third review author (AC). If resolution of a disagreement is not possible, we will designate the study as 'awaiting classification' and we will contact study authors for clarification. We will document reasons for exclusion of studies that may have reasonably been expected to be included in the review in a 'Characteristics of excluded studies' table. We will present an adapted Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) flow diagram showing the process of study selection (Liberati 2009).

Data extraction and management

We will develop a dedicated data abstraction form that we will test ahead of time.

For studies that fulfil the inclusion criteria, two review authors (SP, VV, or JVAF) independently will abstract the following information, which we will provide in the 'Characteristics of included studies' table.

  • Study design.

  • Study dates (if dates are not available then this will be reported).

  • Study settings and country.

  • Participant inclusion and exclusion criteria.

  • Participant details, baseline demographics.

  • The number of participants by study and by study arm.

  • Details of relevant experimental and comparator interventions such as: manufacturer, device model and characteristics.

  • Definitions of relevant outcomes, and method and timing of outcome measurement, as well as any relevant subgroups.

  • Sources of study funding.

  • Declarations of interest by primary investigators.

We will extract outcome data relevant to this Cochrane Review as needed for calculation of summary statistics and measures of variance. For dichotomous outcomes, we will attempt to obtain numbers of events and totals in order to populate a 2 x 2 table, as well as summary statistics with corresponding measures of variance. For continuous outcomes, we will attempt to obtain means and standard deviations or data necessary to calculate this information. For time‐to‐event outcomes, we will attempt to obtain hazard ratios (HR) with corresponding measures of variance or data necessary to calculate this information.

We will resolve any disagreements by consensus, or, if required, by consultation with a third review author (AC).

We will provide information, including trial identifiers, about potentially relevant ongoing studies in the table 'Characteristics of ongoing studies'.

We will attempt to contact authors of included studies to obtain key missing data as needed.

Assessment of risk of bias in included studies

Two review authors (SP, VV, or JVAF) will assess the risk of bias of each included study independently. We will resolve disagreements by consensus, or by consultation with a third review author (AC).

We will assess risk of bias using the Cochrane 'Risk of bias' assessment tool (Higgins 2011b). We will assess the following domains:

  • sequence generation (selection bias);

  • allocation concealment (selection bias);

  • blinding of participants and personnel (performance bias);

  • blinding of outcome assessment (detection bias);

  • incomplete outcome data (attrition bias);

  • selective outcome reporting (reporting bias); and

  • other sources of bias.

We will judge risk of bias domains as 'low risk', 'high risk' or 'unclear risk' and will evaluate individual bias items as described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011b). We will present a 'Risk of bias' summary figure to illustrate these findings.

For performance bias (blinding of participants and personnel) and detection bias (blinding of outcome assessment), we will evaluate the risk of bias separately for each outcome, and, when we report our findings in the 'Risk of bias' tables, we will group outcomes according to whether they were measured subjectively or objectively.

We will also assess attrition bias (incomplete outcome data) on an outcome‐specific basis, and will group outcomes with like judgements when reporting our findings in the 'Risk of bias' tables.

In addition, we will summarize the risk of bias across domains for each outcome in each included study, as well as across studies and domains for each outcome.

Measures of treatment effect

We will express dichotomous data as risk ratios (RRs) with 95% confidence intervals (CI). We will express continuous data as mean differences (MDs) with 95% CI, unless different studies use different measures to assess the same outcome, in which case we will express data as standardised mean differences with 95% CI. We will express time‐to‐event data as hazard ratios (HRs) with 95% CI.

Unit of analysis issues

The unit of analysis will be the individual participant. If a patient receives more than one shunt during the course of a trial, we will consider the outcomes related to the first implanted device only.

Dealing with missing data

We will obtain missing data from study authors, if feasible, and will perform intention‐to‐treat (ITT) analyses whenever possible; otherwise we will perform available case analyses. We will investigate attrition rates, e.g. dropouts, losses to follow‐up and withdrawals, and will appraise any issues of missing data critically. If we suspect that missing data are not missing randomly, we will perform sensitivity analyses by imputing outcomes to recreate the most extreme possible data sets, one reflecting the best‐case scenario for the experimental treatment and the other the worst‐case scenario as described in Higgins 2008.

Assessment of heterogeneity

We will evaluate clinical diversity (clinical heterogeneity) across studies by examining the description of the characteristics of participants, interventions and outcome assessment from data extracted (see Data extraction and management). In the event of significant heterogeneity that cannot be explained by subgroup analyses, we will not report outcome results as a pooled effect estimate in a meta‐analysis, but will provide a narrative description of the results of each study.

We will identify heterogeneity (inconsistency) through visual inspection of the forest plots to assess the amount of overlap of CIs, and the I2 statistic: this statistic quantifies inconsistency across studies to assess the impact of heterogeneity on the meta‐analysis (Higgins 2002; Higgins 2003). We will interpret the I2 statistic as follows:

  • 0% to 40%: might not be important;

  • 30% to 60%: may represent moderate heterogeneity;

  • 50% to 90%: may represent substantial heterogeneity;

  • 75% to 100%: represents considerable heterogeneity.

When we find heterogeneity, we will attempt to determine possible reasons for it by examining individual study and subgroup characteristics, as well by sensitivity analyses by risk of bias of domains.

Assessment of reporting biases

We will attempt to obtain study protocols to assess for selective outcome reporting.

If we include ten studies or more investigating a particular outcome, we will use funnel plots to assess small study effects. Several explanations can be offered for the asymmetry of a funnel plot, including true heterogeneity of effect with respect to trial size, poor methodological design (and hence bias of small trials) and publication bias. We will therefore interpret results carefully.

Data synthesis

Unless there is good evidence for homogeneous effects across studies, we will summarize data using a random‐effects model. We will interpret random‐effects meta‐analyses with due consideration of the whole distribution of effects. In addition, we will perform statistical analyses according to the statistical guidelines contained in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a). For dichotomous outcomes, we will use the Mantel‐Haenszel method; for continuous outcomes, we will use the inverse variance method; and for time‐to‐event outcomes, we will use the generic inverse variance method to combine final scores or change score of continuous outcomes. We will use Review Manager 5 software to perform analyses (Review Manager 2014).

Subgroup analysis and investigation of heterogeneity

If sufficient studies are available, we will conduct subgroup analyses according to the type of valve employed and also by a number of factors prespecified below, as it is probable that, given the heterogeneity of the causes of hydrocephalus, clinical outcomes may depend on these subgroups:

  • type of valve employed: programmable valve systems, gravitational, anti‐siphon, flow‐regulating and conventional valves;

  • age of participant: paediatric participants (i.e. from birth to 18th birthday) and adults;

  • congenital or acquired hydrocephalus.

If there are sufficient studies, we will use the test for subgroup differences in Review Manager 2014 to compare subgroup analyses.

Sensitivity analysis

We plan to perform sensitivity analyses in order to explore the influence of the following factors (when applicable) on effect sizes:

  • restricting the analysis by taking into account risk of bias, by excluding studies at high risk or unclear risk, in the following domains: sequence generation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment (detection bias), incomplete outcome data (attrition bias), selective outcome reporting (reporting bias);

  • best‐case and worst‐case scenarios for the experimental intervention when suspecting outcome‐related missing data as described in Dealing with missing data;

  • by using the fixed‐effect model instead of random‐effects model when moderate heterogeneity is present.

'Summary of findings' table

We will present the overall quality of the evidence for each outcome after applying the methods of the GRADE Working Group (GRADE Working Group 2004), which takes into account five criteria that relate to internal validity (risk of bias, inconsistency, imprecision, publication bias), and also to external validity, such as directness of results (Guyatt 2008). For each comparison, independently, two review authors (SP, VV, or JVAF) will rate the quality of evidence for each outcome as 'high', 'moderate', 'low', or 'very low' using GRADEpro 2008. We will resolve any discrepancies by consensus, or, if needed, by arbitration by a third review author (AC). For each comparison, we will present a summary of the evidence for the main outcomes in a 'Summary of findings' table, which provides key information about the best estimate of the magnitude of the effect in relative terms and absolute differences for each relevant comparison of alternative management strategies; numbers of participants and studies addressing each important outcome; and the rating of the overall confidence in effect estimates for each outcome (Guyatt 2011; Schünemann 2011). If meta‐analysis is not possible, we will present results in a narrative 'Summary of findings' table.

We will create a 'Summary of findings' table using the following outcomes:

  • treatment failure;

  • mortality related to the shunt placement;

  • morbidity related to the shunt placement (subdural haematomas, slit ventricles, infections related to shunt placement);

  • mortality for all causes; and

  • quality of life.