There is no accepted figure for the incidence of new cases of acute spinal cord injury (SCI) globally; estimates range from 10 to 83 cases per million population annually (Wyndaele 2006). It is estimated that between 500 and 700 people sustain a traumatic SCI in the UK each year (Harrison 2000), and that there are approximately 10,000 new cases per year in the USA (McDonald 2002).
SCI can occur at any age. The effects are usually permanent and currently there is no cure (Smith 1999). The modal age of SCI is 19 years, most people with SCI then live a relatively normal lifespan, so the lifetime cost of care may be quite high. The average lifetime cost of treating a person with SCI has been estimated at between US$500,000 and US$2 million, depending on the extent and location of injury (McDonald 2002). The most common mechanism of injury is a sudden unexpected impact or deceleration (e.g. road traffic injury, domestic falls). Further neurological deterioration, resulting from lesion extension after the initial injury, can occur naturally in about 5% of cases (Harrison 2000) and complications associated with the systemic effects of SCI can lead to respiratory compromise. Significant delays and complications - sometimes leading to admission to an intensive therapy unit (ITU) can also arise as a result of inappropriate or poorly informed management.
If the spine is unstable following injury, surgical fusion and bracing may be necessary to obtain vertical stability and prevent re-injury of the spinal cord from repeated movement of the unstable bony elements (Geisler 1988). It has been suggested that spinal fixation surgery may enable early rehabilitation and mobilisation. There does not seem to be an accepted protocol with regard to what type of surgery is used: whether surgical stabilisation or surgical decompression are required, and what type of approach, instrumentation and procedure should be chosen. In some cases the procedure involves posterior decompression and fusion with a bone graft and with hardware consisting of wires or rods.
Different techniques are used for cervical spine surgery and for thoracolumbar spine surgery (Donovan 1994). In the cervical region, several stabilising options exist: the soft cervical collar, the Philadelphia collar, the sternal-occipital-mandibular immobiliser (SOMI), Yale types of cervical-post brace, the halo vest, and the thermoplastic Minerva body jacket (Amar 1999). The procedure may vary between surgeons; for example, surgeons in a specialist spinal injuries unit (SIU) may be more likely to use bone grafts.
Surgical reduction and stabilisation of the spine at the immediate/early stage is done to prevent secondary spinal cord injury (McDonald 2002), but can cause further oedema at the lesion site with a resulting extension of ischaemia. Early internal stabilisation surgery is reported to have substantial pragmatic advantages in later rehabilitation phases compared with external (halo) stabilisation devices alone (McDonald 2002). Indications for both surgical stabilisation and surgical decompression are subjective.
Arguments for spinal fixation, or stabilisation, surgery tend to focus on perceived advantages, such as shorter hospital stays, assurance of stability, correction of deformity and enhancement of neurological recovery. The strength of the case for spinal fixation surgery can depend on factors such as the patient's general medical condition, spinal instability, deformity and completeness and level of the lesion (Donovan 1994). Benefits of stabilisation surgery have been agreed in the UK (by the British Association of Spinal Cord Injuries Specialists, the British Association of Spinal Surgeons and the British Cervical Spine Society) to be protection of the neural tissues, reduction of pain, easier patient handling, earlier mobilisation within physiological restrictions, reduction of respiratory complications and reduction in late deformity with better posture and balance (British Orthopaedic).
On the other hand, surgery is considered by some to potentially lead to hypoxia, hypotension, further mechanical damage and post-operative complications such as bleeding, chest or urinary infection and infection at the wound site (El Masri 2006).
A systematic review (Bagnall 2003) was commissioned by the Health Technology Assessment Programme on the effectiveness and cost-effectiveness of acute hospital-based services for spinal cord injuries. The HTA review aimed to answer five research questions. One of those questions, the effectiveness and cost-effectiveness of spinal fixation surgery, will be addressed in this review in greater detail.
The objective of this review is to answer the following research questions:
- Is there a difference in rate or completeness of neurological recovery between those who have had spinal fixation surgery and those who have not?
- Does spinal fixation surgery have an effect on time to mobilisation, acute recovery from trauma, pain, posture, spinal deformity, surgical complications, post traumatic syrinx and other generally measured outcomes?
Criteria for considering studies for this review
Types of studies
Published and unpublished randomised or non-randomised controlled trials (RCTs or CCTs).
Types of participants
People of any age with a complete or partial interruption of spinal cord function resulting from trauma.
Types of interventions
Surgical spinal fixation (with or without surgical decompression) compared to any other treatment. Studies which compared different types of surgical spinal fixation were not included. Studies which included postoperative external bracing were eligible for inclusion in the review.
Types of outcome measures
The following outcome measures were eligible for inclusion in the review:
- Neurological improvement.
- Functional ability.
- Activities of daily living.
- Discharge venue and associated costs.
- Time to mobilisation.
- Acute recovery from trauma.
- Psychological and social outcomes (including employment).
- Infections (especially methicillin-resistant staphylococcus aureus (MRSA)).
- Incidence of secondary complications (such as pressure sores).
- Other adverse events e.g. spinal deformity, post traumatic syrinx, time spent in intensive care, on ventilation etc.
Search methods for identification of studies
The search strategy was devised to find papers about spinal fixation surgery for spinal cord injuries. This strategy combined terms for 'spinal cord injury' with terms for 'fixation' and 'fusion'. The strategy also used specific search terms for 'spinal cord surgery', but not broader search terms for 'spinal surgery' in order to narrow the search. The terms used in the search strategy were identified through discussion with the research team involved in the HTA review, by scanning background literature and by browsing the MEDLINE thesaurus (MeSH).
Full details of the search strategies are available from the authors and from Bagnall 2003 in electronic format.
Searches were conducted for the HTA review in October 2001. The searches were updated for this Cochrane review in May 2003 and the MEDLINE search was updated in May 2007.
The following databases were searched:
- Allied and Complementary Medicine (AMED, to May 2003),
- Cochrane Controlled Trials Register (CCTR, to May 2003),
- Cumulative Index to Nursing and Allied Health Literature (CINAHL, to May 2003),
- Database of Abstracts of Reviews of Effectiveness (DARE, to May 2003),
- EMBASE (to May 2003),
- Health Economic Evaluations Databases (HEED, to May 2003),
- Health Management Information Consortium (HMIC, to May 2003),
- MEDLINE (to May 2007),
- National Research Register (NRR, to May 2003),
- NHS Economic Evaluation Database (NHS EED, to May 2003).
In addition, the following searches were also carried out on the Internet using OMNI (http://omni.ac.uk), Copernic (http://www.copernic.com/), Alta Vista (http://www.altavista.com/) and Google (http://www.google.com/). Specialist spinal cord injury and spinal injury related web sites were searched, specifically: Spinal Injuries Association (http://www.spinal.co.uk/), the British Association of Spinal Cord Injury Specialists (http://www.bascis.pwp.blueyonder.co.uk/) and the National Spinal Cord Injury Association (http://www.spinalcord.org/).
Searching other resources
The reference lists of all retrieved studies were also scanned for additional studies.
Data collection and analysis
Two authors independently screened all study citations for inclusion. Any discrepancies were resolved by discussion with reference to the original papers and, if necessary, by discussion with a third reviewer. As no studies were found, there was no data to review.
It was intended to extract data onto forms developed for different study designs on a Microsoft Access database. One reviewer was to have extracted the data and a second to check the forms for accuracy. Disagreements were to be resolved by discussion or, when necessary, through discussion with a third reviewer.
The quality of studies that were included were to be assessed according to established criteria (NHS CRD 2001). Briefly, these are as follows:
- for RCTs, method of randomisation, method of allocation concealment, blinding, handling of withdrawals, similarity of groups at baseline, specified eligibility criteria, presentation of results;
- for non-randomised controlled studies, as above but without the randomisation and allocation concealment questions.
Quality assessment was to be carried out by one reviewer on to predefined and piloted forms on a Microsoft Access database, and checked by the second reviewer for accuracy. Any disagreements were to be resolved by discussion or, when necessary, through discussion with a third reviewer. Quality scores would not be assigned to studies, but the results of quality assessment were to be discussed in the report.
As no RCTs or controlled studies were identified, meta-analysis was not possible.
Description of studies
In the HTA review (Bagnall 2003), 68 studies were identified that addressed the question of spinal fixation surgery for acute traumatic SCI. All studies included a control group, in that a group receiving spinal fixation surgery was compared to a group not receiving spinal fixation surgery. However, all studies were retrospective observational studies and of poor quality. There was some doubt over the comparability of groups and/or on confounding factors in many of the studies. Often, the decision on whether to treat surgically or not was made based on the severity of the patient's injuries (more severe injuries led to non-operative treatment in some units and to operative treatment in others). In many studies, results of surgery with and without fixation were reported together. In a number of other studies, few details of baseline severity or patient demographics were reported. A full discussion of these studies, including full data extraction and quality assessment tables, is available in the HTA review.
Update searches, conducted for this review in May 2003 and May 2007, located a further 3106 records. Of these, 10 were retrieved for full inspection. No relevant RCTs or controlled trials were identified. The majority of the identified studies were retrospective case series and did not directly investigate whether spinal fixation surgery resulted in better outcomes than no spinal fixation surgery. One study (Brodke 2003) compared anterior and posterior surgical approaches in a RCT. Another study (Kerwin 2005) was a retrospective review of the effects of spinal fixation within 3 days. One literature review (Kishan 2005) and one systematic review (Fehlings 2006) were also found. Both suggested that early decompression surgery may be beneficial, but did not find strong evidence about the effects of spinal fixation surgery.
Risk of bias in included studies
No studies were included.
Effects of interventions
No studies were included.
No RCTs or controlled trials were identified that answered the question whether there is a difference in functional ability or other commonly measured outcomes for those who have had spinal fixation surgery and those who have not. All of the studies identified in the original and update searches were retrospective and of poor quality. The limitations of this type of study design meant that all of the identified studies suffered from a number of methodological flaws. The validity of the studies may be affected by confounding and other biases, but often important variables that could affect the study results are not reported in sufficient detail to allow the reader to make a judgement. In the future, studies of any design should report more details of participants and outcomes (for example, the Injury Severity Score (ISS) or similar standardised measure, and the level of SCI should be reported). Data could then be stratified according to injury level. This is important because conservative treatment for SCI differs according to the level of injury (conservative treatment for cervical fractures is quite different from conservative treatment of thoracic or lumbar fractures). Patient characteristics should be reported as these may also differ according to injury level (for example cervical fractures are predominantly seen in the elderly). MRI findings should also be reported, where applicable, along with details of all treatments given, including whether post operative external bracing was used. It should also be noted that in retrospective studies the severity of injury will impact the decision whether to treat surgically or not: more severe injuries will lead to non-operative treatment in some studies and operative treatment in others. Studies with historical controls have additional biases in that aspects of care other than the decision to treat surgically or not will differ between treatment and control groups.
Well-designed, prospective randomised controlled trials with appropriately matched controls are needed. If this is not possible, at the very least well-designed prospective cohort studies with concurrent and appropriate controls are required. Outcomes should be reported in a standardised way, giving as much information as possible for treatment and control groups, and outcomes that are important to patients, their families and carers, as well as clinicians, should be measured and reported. People with SCI, and their representatives, should be involved in the design of future research studies to ensure the research is relevant and useful.
Implications for practice
The current evidence is insufficient to enable the author to comment on the benefits or harms of spinal fixation surgery in patients with traumatic SCI.
Implications for research
Well-designed, prospecive experimental studies with appropriately matched controls are required to assess the benefits or harms that may be associated with spinal fixation surgery. All future research should be planned in association with people with SCI and their carers to ensure that appropriate and relevant research is carried out.
We wish to thank Steven Duffy for conducting the initial and 2003 update searches for this review.
Data and analyses
This review has no analyses.
Last assessed as up-to-date: 9 October 2007.
Protocol first published: Issue 2, 2004
Review first published: Issue 1, 2008
Contributions of authors
AM Bagnall - wrote protocol
L Jones - wrote protocol
S Duffy - designed and carried out search strategy
R Riemsma - oversaw project, provided comments and input at all stages
Declarations of interest
Sources of support
- Centre for Reviews and Dissemination, University of York, UK.
- Health Technology Assessment (HTA) Programme, NHS, UK.
Medical Subject Headings (MeSH)
MeSH check words