Interventions for treating supracondylar fractures of the humerus in children
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
To assess the effects (benefits and harms) of interventions for treating supracondylar fractures of the humerus in children. We intend to compare different types of conservative intervention, surgical versus conservative treatment, and different surgical interventions.
Background
Description of the condition
Supracondylar fractures of the humerus (upper arm bone) are the most common elbow fractures in children, accounting for 50% to 70% of all paediatric elbow fractures (Wilson 2006). These fractures have a peak incidence in children of between five and seven years of age (Kasser 2001) and occur more often in boys (2:1 male to female ratio). These injuries usually occur after either a fall onto an outstretched hand or a fall onto a flexed elbow (Farnsworth 1998), and the non‐dominant arm is most commonly affected. Children under three years typically sustain a supracondylar fracture after a fall from a height of less than three feet (one metre). Fractures in older children are associated with greater energy, such as a fall from playground equipment. Extension type injuries (arm is extended) account for 95% of supracondylar fractures; flexion fractures (arm is bent) are much rarer, accounting for 5% (Carson 2006).
There is a relatively high incidence of neurovascular complications (3% to 14%) associated with these fractures due to the close proximity of nerves and vascular structures to the elbow joint; particularly the median nerve and brachial artery (Griffin 2008). Neurovascular compromise occurs with entrapment or laceration of the brachial artery at the fracture site, or damage to the median nerve as it passes close to the elbow joint. In addition, children with supracondylar fractures of the humerus and concurrent forearm fractures are at particularly high risk of developing forearm compartment syndrome (Baratz 2006; Shrader 2008). The risk of neurovascular compromise and significant residual deformity make this a serious injury that requires careful management.
Gartland 1959 described a useful classification for supracondylar fractures of the humerus: undisplaced (type I), displaced with intact posterior cortex (type II), and displaced with no cortical contact (type III). Type III fractures may be displaced in three directions; posteromedially (the most common), posterolaterally or anterolaterally. The direction of displacement is important in determining which neurovascular structures are at greatest risk of injury.
Description of the intervention
Several treatment options exist for supracondylar fractures; all require appropriate assessment of the severity of the injury including neurovascular compromise. Typically non‐surgical (conservative) management of undisplaced fractures involves immobilisation of the arm in a cast or splint for three to five weeks. For displaced fractures, this would be preceded by closed reduction where the fractured parts are manipulated through the skin back into place. Other conservative treatment of displaced fractures involves side‐arm skin traction or overhead skeletal traction; both of which require extended periods in hospital. Surgical treatment may involve closed reduction and percutaneous K‐wire fixation (followed by immobilisation in a cast or splint), open reduction and internal fixation, or external fixation.
How the intervention might work
The choice of intervention used to treat these fractures will depend on the type of fracture and whether there is neurovascular compromise. Cast or splint immobilisation or protracted traction aim to keep the fractured bone parts in position while they heal. Conservative treatment avoids potential surgical complications such as neurovascular damage from incorrect pin placement and infection. Surgical treatment generally enables a more accurate reduction of the fracture and better stabilisation of the reduced fracture to achieve bony union. Surgery is generally indicated when the fracture is unstable. Closed reduction and percutaneous K‐wire fixation is the least invasive method but open reduction and internal fixation with K‐wires may be necessary if anatomic or near anatomic reduction is unobtainable by closed methods. The use of crossed pins or two (or three) lateral pins, the timing of surgery, the role of open reduction and the approach for open reduction are controversial. External fixation may provide more rigid stabilisation but can be cumbersome for the patient. Surgically‐treated patients routinely have subsequent surgery to remove metalwork (pins and wires).
Why it is important to do this review
Despite these fractures being common in children, the optimal management remains controversial, especially for displaced fractures and fractures with neurovascular compromise (Lee 2012; Zamzam 2009). This review will examine and summarise current evidence regarding the management of these fractures in order to help guide their clinical management.
Objectives
To assess the effects (benefits and harms) of interventions for treating supracondylar fractures of the humerus in children. We intend to compare different types of conservative intervention, surgical versus conservative treatment, and different surgical interventions.
Methods
Criteria for considering studies for this review
Types of studies
We will include randomised controlled trials (RCTs) and quasi‐randomised controlled trials (method of allocating participants to a treatment which is not strictly random e.g. by date of birth, hospital record number, alternation) assessing interventions for treating supracondylar fractures of the humerus in children.
Types of participants
We will include studies of children of either gender with fractures involving the supracondylar region of the distal humerus. We will exclude participants over 16 years of age (the age of skeletal maturity).
Types of interventions
We will include all conservative and surgical interventions used in the treatment of these fractures. We will exclude trials that evaluate different methods of treating complications or soft‐tissue injuries, or rehabilitation interventions.
We intend to make the following main comparisons.
-
Different methods of conservative intervention.
-
These include closed reduction, if the fracture is displaced, and cast immobilisation (type and timing) and traction (e.g. side‐arm skin traction versus overhead skeletal traction).
-
-
Surgical versus conservative treatment.
-
Different methods of surgical intervention.
-
K‐wire fixation, including different configurations of K‐wire fixation.
-
Open fixation (minimal versus extensive incision; anterior versus posterior approach; and medial versus lateral approach).
-
External fixation.
-
We will also consider the timing of initial surgery, specifically emergency (immediate) versus urgent (within 24 hours) surgery.
Types of outcome measures
Primary outcomes
-
Elbow function. We have been unable to locate a specific patient‐reported functional measure used for these injuries. Thus where possible, we will report on numbers achieving a good or full recovery (thus normal elbow function without pain, loss of motion or deformity). This appraisal may draw upon the commonly used Flynn's criteria (Flynn 1974).
-
Pain (post‐intervention) at short‐, medium‐ and long‐term follow‐up.
-
Treatment failure (re‐operation, conversion of conservative to surgical management, extensive physiotherapy for an adverse effect).
Secondary outcomes
-
Adverse effects:
-
short‐term: iatrogenic neurological or vascular injury, compartment syndrome (where the complications were not present prior to surgery);
-
medium‐term: infection (including septic arthritis), malunion/non‐union, myositis ossificans; and
-
long‐term: cosmetic deformity, posterolateral rotatory instability.
-
-
Measures, preferably validated, of musculoskeletal function in children such as the Paediatric Outcome Data Collection Instrument (PODCI) (Daltroy 1998).
-
Range of motion (loss of motion, loss of carrying angle) using Baumann's angle (Williamson 1992), the humero‐ulnar angle, and Flynn's criteria for loss of motion.
-
Economic considerations (length of stay in hospital, cost of hospital admission, cost of intervention).
We will consider the timing of follow‐up and outcome measurement.
Search methods for identification of studies
Electronic searches
We will search the Cochrane Bone, Joint and Muscle Trauma Group's Specialised Register (to present), the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library, current issue), MEDLINE (1948 to present) and EMBASE (1980 to present). We will also search Current Controlled Trials and the WHO International Clinical Trials Registry Platform for ongoing and recently completed trials. We will not apply any language restrictions.
In MEDLINE (Ovid Online), we will combine the subject‐specific strategy with the sensitivity‐maximising version of the Cochrane Highly Sensitive Search Strategy for identifying randomised trials (Lefebvre 2011). Search strategies for The Cochrane Library and MEDLINE can be found in Appendix 1.
Searching other resources
We will search the reference lists of all retrieved articles and we will contact expert researchers in the field.
Data collection and analysis
Selection of studies
Two review authors (SBR and PM) will independently screen search results for eligible studies. Where suitable, we will seek full articles for any study judged eligible by either author. The same two authors will perform study selection. We will resolve any disagreements by discussion or adjudication by the third author (AR).
Data extraction and management
Two review authors (SBR and PM) will independently extract data from each included trial using a pre‐derived data extraction form and enter the data into Review Manager 5 (RevMan 2011). We will record qualitative details and data describing the study groups, interventions and outcomes. We will contact trialists for further details as necessary. We will resolve any differences or disagreements by checking trial reports, contacting trial authors, discussion between the two authors (SBR and PM), or adjudication by the third author (AR).
Assessment of risk of bias in included studies
Two review authors (SBR and PM) will independently assess risk of bias using The Cochrane Collaboration's 'Risk of bias' tool, without masking of the source and authorship of the trial reports (Higgins 2011a). Any disagreement will be resolved by consensus of the three authors.
We will judge each study as 'low risk', 'unclear', or 'high risk' of bias for the following domains.
-
Sequence generation.
-
Allocation concealment.
-
Blinding (of participants and personnel).
-
Blinding (of outcome assessors).
-
Incomplete outcome data addressed.
-
Free of selective outcome reporting.
-
Free of other bias.
We will assess the risk of bias associated with patient‐rated outcomes separately from clinician‐rated outcomes for the two blinding domains and incomplete outcome data. We will assess the grade of operating surgeon between the intervention arms as a source of performance bias.
Measures of treatment effect
We will calculate risk ratios (RRs) and 95% confidence intervals (CIs) for dichotomous outcomes, and mean differences and 95% CIs for continuous outcomes. When pooling continuous data for outcomes measured in different ways or scales (e.g. measures of elbow function), we shall use the standardised mean difference and 95% CIs.
Unit of analysis issues
We anticipate that the unit of randomisation in trials on this topic will be the individual patient. We will be alert to potential unit of analysis issues. These include those relating to multiple observations of the same outcome (such as for total complications where a child might have more than one complication) or multiple time points (we will extract data at clinically relevant time points and perform separate analyses for these).
Dealing with missing data
Whenever possible, we will contact study authors to request missing data. We will analyse missing data for dichotomous and continuous data using an available case analysis, and we will perform sensitivity analyses using 'best‐worst case' analysis (Gamble 2005) and 'fixed difference' analysis (Higgins 2011b; see 16.2), respectively. We will make explicit the assumptions of any methods used to cope with missing data (Higgins 2011b; see 16.1).
Assessment of heterogeneity
We will assess heterogeneity by visual inspection of the forest plot (analysis) along with consideration of the Chi² test (which we will consider to be statistically significant at P < 0.10) and the I² statistic (Higgins 2003).
Assessment of reporting biases
If sufficient (at least 10) studies are identified, we will construct funnel plots (trial effect versus standard error) to assess funnel plot asymmetry, which amongst other things could be due to publication bias. We will conduct a test of funnel plot asymmetry for the main outcomes using Eggers test (Egger 1997). We will conduct analyses using Review Manager 5 (RevMan 2011).
Data synthesis
When appropriate and possible, we will pool outcome data from trials testing comparable comparisons. In the first instance, we will use a fixed‐effect meta‐analysis for combining data. However, where there is substantial and unexplained heterogeneity between studies we will use a random‐effects meta‐analysis.
Subgroup analysis and investigation of heterogeneity
Specific subgroup analysis will be according to:
-
fracture classification (Wilkins modification of the Gartland classification (flexion/extension type and I, II, or III (Wilkins 1984));
-
K‐wire fixation configuration (lateral pins versus crossed pins); and
-
timing of surgery (emergency (immediately) versus urgent (within 24 hours)).
Sensitivity analysis
If appropriate, we will perform sensitivity analyses examining various aspects of trial and review methodology, including the inclusion of trials at high or unclear risk of bias (such as from lack of allocation concealment and lack of blinding of outcome assessors), and of trials only reported in abstracts.
'Summary of findings' tables
Where there are sufficient data, we will summarise the results for the main comparisons described in Types of interventions in 'Summary of findings' tables. We shall use the GRADE approach to assess the quality of evidence related to each of the primary outcomes listed in Types of outcome measures (Higgins 2011c; see 12.2).
