Proximal femoral fractures, more generally termed 'hip fractures', can be subdivided into intracapsular fractures (those occurring proximal to the attachment of the hip joint capsule to the femur) and extracapsular (those occurring distal to the hip joint capsule). Intracapsular fractures can be subdivided into those which are displaced and those which are essentially undisplaced. Undisplaced fractures include those which are termed impacted or adduction fractures. Many other subdivisions and classification methods exist for intracapsular fractures but have not been shown to be of reliable clinical usefulness (Thorngren 2002). These fractures are invariably treated surgically, for which the main treatment options are either replacement of the femoral head or internal fixation of the fracture. This review pertains only to aspects of internal fixation of the fracture.
The rationale for operative treatment by internal fixation for intracapsular fractures is to reduce the risk of the fracture displacing for undisplaced fractures, and to maintain fracture reduction for displaced fractures that have been reduced at the time of surgery. Surgery entails passing single or multiple screws or pins across the fracture site. For displaced fractures the fracture must first be reduced, which may be achieved either by an open (surgical exposure and direct inspection of the fracture) or closed reduction (reducing the fracture under X-ray control). Numerous implants are available for the internal fixation of the fracture. These implants are inserted under X-ray guidance using either an open or percutaneous procedure.
Complications that may occur after internal fixation of an intracapsular fracture mainly relate to failure of the fixation to hold (this may be termed early re-displacement) and failure of the fracture to heal (termed non-union). As there is no clear consensus as to the difference between fracture re-displacement and non-union these two are generally considered together under the term non-union. Damage to the vascular supply to the femoral head may result in the complication of avascular necrosis of the femoral head (also termed segmental collapse and aseptic necrosis). This normally occurs within the first two years from injury. Other complications that may occur are later fracture around the implant and breakage of the implant. Bone scintimetry may be used to assess the risk of the fracture healing complications of non-union and avascular necrosis.
Many different technical aspects of surgical fixation exist. Surgeons have hypothesised that both the quality of the reduction, and the overall stability of the fixation may influence the frequency of these complications. As well as new types of fixation devices, different ancillary surgical techniques have been introduced. These include reduction of the fracture (open or closed), positioning of the implant, forcible impaction of the fracture and compression of the fracture. This review considers only those aspects of surgical fixation of an intracapsular fracture that have been studied within a randomised trial. Trials comparing different implants for internal fixation are examined in a separate Cochrane review (Parker 2004).
To evaluate, based on evidence from randomised trials, the effects of different surgical approaches and ancillary techniques used in the internal fixation of intracapsular proximal femoral fractures.
We tested the following null hypothesis.
- There is no difference between the use or non use of any surgical technique or approach, or between the use of one compared with another surgical technique or approach during surgical fixation of an intracapsular hip fracture in adults.
Criteria for considering studies for this review
Types of studies
All randomised controlled trials comparing surgical techniques used in internal fixation of intracapsular hip fractures. Quasi-randomised trials (for example, allocation by alternation or date of birth) and trials in which the treatment allocation was inadequately concealed were considered for inclusion.
Types of participants
Skeletally mature people with an intracapsular proximal femoral fracture.
Types of interventions
Surgical techniques such as open and closed reduction, compression or impaction of the fracture, variations in the positioning of the implant, and any other surgical techniques that have been evaluated in randomised trials were considered.
Types of outcome measures
The following outcomes were sought.
(1) Operative details
- length of surgery (in minutes)
- operative blood loss (in millilitres)
- post-operative blood transfusion (in units)
- accuracy of fracture reduction (as defined in each study)
- accuracy of screw positioning (as defined in each study)
- radiographic screening time (in seconds)
(2) Fracture fixation complications
- non-union of the fracture within the follow-up period (the definition of non-union will be that used within each individual study and this outcome will include early re-displacement of the fracture).
- avascular necrosis/segmental collapse
- fracture below the implant
- other surgical complications of fixation (as detailed in each study).
- all fracture complications (sum of above four items)
- re-operation (within the follow-up period of the study).
- superficial wound infection (infection of the wound in which there is no evidence that the infection extends to the site of the implant)
- deep wound infection (infection around the implant)
(3) Post-operative complications
- thromboembolic complications (deep vein thrombosis or pulmonary embolism)
- any medical complication (as detailed in each individual study)
- length of hospital stay (in days)
(4) Anatomical restoration
- shortening of the limb (more than 2 cm)
- varus deformity of the femoral neck
- external rotation deformity ( > 20 degrees)
(5) Final outcome measures
- mortality (within the follow-up period of the study)
- pain (persistent pain at the final follow-up assessment)
- return to living at home
- return of mobility
- other functional outcomes as listed in each study
- health related quality of life measures
Search methods for identification of studies
Relevant randomised controlled trials were selected from those identified by application of the general search strategy for the Cochrane Bone, Joint and Muscle Trauma Group (for details see the group strategy contained in The Cochrane Library). This includes a) computer aided searching of various computer databases, routinely MEDLINE, EMBASE, CINAHL, and The Cochrane Library (CENTRAL) b) hand-searching of various core journals and abstract books of orthopaedic conferences, and c) searching reference lists of articles. The authors also checked reference lists of articles, perused conference proceedings and contacted trialists. No language restriction was applied. The date of our most recent search was November 2004.
Data collection and analysis
Both authors independently extracted data for the outcomes listed above and each trial was assessed independently without masking of the study names for its quality of methodology. Differences were resolved by discussion. The main assessment of methodology was the quality of allocation concealment. A further nine aspects of methodology were used giving a maximum score for each study of 12.
(1) Was there clear concealment of allocation? Score 3 (and code A) if allocation clearly concealed (e.g. numbered sealed opaque envelopes drawn consecutively). Score 2 (and code B) if there was a possible chance of disclosure before allocation. Score 1 (and code B) if the method of allocation concealment or randomisation was not stated or was unclear. Score 0 (and code C) if allocation concealment was clearly not concealed such as those using quasi-randomisation (e.g. even or odd date of birth).
(2) Were the inclusion and exclusion criteria clearly defined? Score 1 if text states type of fracture and which patients included and those excluded. Otherwise score 0.
(3) Were the outcomes of patients who withdrew or excluded after allocation described and included in an intention-to-treat analysis? Score 1 if yes or text states that no withdrawals occurred or data are presented clearly showing 'participant flow' which allows this to be inferred. Otherwise score 0.
(4) Were the treatment and control groups adequately described at entry and if so were the groups well matched, or appropriate co-variate adjustment made? Score 1 if at least four admission details given (e.g. age, sex, mobility, function score, mental test score, fracture type) with either no important difference between groups or appropriate adjustment made. Otherwise score 0.
(5) Were the surgeons experienced at both operations prior to commencement of the trial? Score 1 if text states there was an introductory period or all surgeons were experienced in both operations. Otherwise score 0.
(6) Were the care programmes other than the trial options identical? Score 1 if text states they were or this can be inferred. Otherwise score 0.
(7) Were all the outcome measures clearly defined in the text with a definition of any ambiguous terms encountered? Score 1 if yes. Otherwise score 0.
(8) Were the outcome assessors of both X-rays and clinical outcomes blind to assignment status? Score 1 if assessors of anatomical restoration, pain and function at follow up were blinded to treatment outcome. Otherwise score 0.
(9) Was the timing of outcome measures appropriate? A minimum of 12 months follow up for all surviving patients. Score 1 if yes. Otherwise score 0.
(10) Was loss to follow up reported and if so were less than 5% of patients lost to follow up? Score 1 if yes. Otherwise score 0.
For each study, relative risks and 95% confidence intervals were calculated for dichotomous outcomes, and mean differences and 95% confidence intervals calculated for continuous outcomes. Where appropriate, we planned to pool results of comparable groups of trials using both fixed-effect and random-effects models with 95% confidence intervals. However, the clear heterogeneity in the trial characteristics and statistical heterogeneity, assessed using a standard chi squared test and the I squared test, meant that currently no data pooling has been performed.
Description of studies
Four randomised studies were identified and included. One of the authors of Gray 1988 provided additional information. The authors of this review would be pleased to receive any further data from the authors of the individual trials. No ongoing randomised studies were identified.
Jacobsson 1985 evaluated the effect of impaction of the fracture on the blood supply of the femoral head in 103 people with undisplaced or displaced intracapsular hip fractures. Trial participants were randomly allocated to either impaction of the fracture with a one kilogram mallet after release of the traction at the end of the operation or to no impaction of the fracture. All fractures were internally fixed with either three Scand hip pins or a Thornton nail. The only outcome measure reported for this study was isotope bone scintimetry, measured pre-operatively and at around 10 days after surgery.
Frandsen 1984 compared intra-operative fracture compression, using the compression hip screw, during fixation with a sliding screw plate versus no compression in 220 people with displaced intracapsular fractures. Since, operative impaction with a heavy hammer of the fracture was undertaken at the time of surgery for the "control" group, the true comparison for this trial is 'compression with no impaction versus impaction with no compression'. Fracture healing complications were reported for survivors only.
Two trials compared open versus closed reduction of the fracture at the time of internal fixation. Gray 1988, which was only published as a conference abstract, included people over 65 years of age, all of whom were treated by internal fixation with a sliding hip screw. Trial results and further information were obtained from the trial authors but only for 49 participants, whereas extrapolation of the data from the conference abstract indicates a study population of 134 people. The trial was apparently stopped early due to the higher mortality in the open reduction group. Upadhyay 2004 randomised 102 people with displaced intracapsular fractures, predominantly male and aged between 15 and 50 years, who were treated by internal fixation with three cannulated screws.
A summary of the details of these trials is given in the 'Characteristics of included studies' table.
Risk of bias in included studies
Four studies were identified. Except for Upadhyay 2004, the methodology of the included trials was poor.
Jacobsson 1985 did not state their method of randomisation. There were no truly clinically relevant outcomes; only the results of bone scintimetry were reported. Outcome assessment was not blinded.
Frandsen 1984 was a quasi-randomised trial: treatment allocation was determined by the admission of patients on alternate days to one of two orthopaedic units within the same hospital. Participants were followed up for two years from injury but the results of those who died within one year from surgery were not presented. Outcome assessment was not blinded.
Both Gray 1988 and Upadhyay 2004 used sealed envelopes for randomisation but neither provided confirmation of allocation concealment. A blinded observer read the post-operative X-rays in both studies.
Assessment of methodology (maximum score = 12)
1 2 3 4 5 6 7 8 9 10 total
0 1 0 0 0 1 1 0 1 0 4 Frandsen 1984
2 1 0 0 0 0 0 0 1 0 4 Gray 1988
1 1 0 0 0 0 1 0 0 0 3 Jacobsson 1985
2 1 0 1 0 1 1 0 1 1 8 Upadhyay 2004
Effects of interventions
The four trials identified examined the effectiveness of three different manoeuvres carried out during internal fixation of intracapsular hip fractures. Where data are available, these are presented in the Graphs. No data pooling was done.
Impaction of the fracture (Jacobsson 1985)
For trial participants with undisplaced fractures there was no difference in femoral head blood supply as measured by bone scintimetry between the 16 fractures that were impacted (femoral head isotope uptake mean ratio of fracture/intact side: 1.49 ± 0.58) against the 13 fractures that had no impaction (ratio: 1.67 ± 0.48). For displaced fractures it was reported that impaction resulted in a significant (P < 0.05) decrease in femoral head "vitality" as assessed by scintimetry. The results were a ratio of 1.08 ± 0.36 for 29 fractures treated with impaction versus 1.34 ± 0.48 for the 36 fractures with no impaction. There were no data for any of the outcomes listed in 'Types of outcome measures'.
Compression versus no compression of the fracture (Frandsen 1984)
Fewer, though not statistically significantly fewer, people had died in the compression group at one year (27/112 (24.1%) versus 37/108 (34.3%); relative risk (RR) 0.70, 95% confidence interval (CI) 0.46 to 1.07). Fracture healing outcomes were not presented for these 64 people. For the 156 survivors, the incidence of non-union was significantly greater in the compression group (28/85 (32.9%) versus 13/71 (18.3%); RR 1.8, 95% CI 1.01 to 3.20). The combined outcome of 'fracture non-union or dead at one year' showed no difference between the two groups (55/112 (49.1%) versus 50/108 (46.2%); RR 1.06, 95% CI 0.80 to 1.40). There were similar results in the two groups for avascular necrosis in fractures that were united at two years or more (9/37 (24.3%) versus 7/36 (19.4%); RR 1.25, 95% CI 0.52 to 3.00). No other outcome measures were reported.
Open or closed reduction of the fracture (Gray 1988; Upadhyay 2004)
The two trials testing this comparison were very different in terms of study populations (for example, all participants of Gray 1988 were over 65 years old and probably the majority were female, whereas those in Upadhyay 2004 were aged between 15 and 50 years and were mainly male), the method of internal fixation (sliding hip screw versus three cannulated screws) and general circumstances. Despite this, we initially pooled the results for the two outcomes (length of surgery and fracture healing failure rate) that the two trials had in common. The highly significant heterogeneity results were consistent with the fundamental differences in these two trials and we thus have not pooled the results here. When viewing the results of Gray 1988, readers should note that these are very likely to be of a subgroup. In both trials, the mean length of surgery in minutes was significantly increased in the open reduction group (Gray 1988: mean difference (MD) 17.00 minutes, 95% CI 9.32 to 24.68 minutes; Upadhyay 2004: MD 48.42 minutes; 95% CI 42.60 to 54.24 minutes). None of the differences between the two groups in Gray 1988 in the numbers receiving post-operative blood transfusion or in various radiological parameters were statistically significant. Similarly, none of the differences between the two groups in Upadhyay 2004 were statistically significant for the following outcomes: mean time to fracture union; poor quality fracture reduction, poor screw placement; non-union of fracture; and avascular necrosis. The overall failure rate (which included both non-union and avascular necrosis) was in favour, though still not reaching statistically significance, for the open reduction group in Gray 1988 (6/27 (22.2%) versus 11/22 (50.0%); RR 0.44, 95% CI 0.20 to 1.01) but similar in the two groups of Upadhyay 2004 (15/44 (34.1%) versus 16/48 (33.3%); RR 1.02, 95% CI 0.58 to 1.81). Though more participants in the open reduction group had wound infection in Upadhyay 2004, the difference was not statistically significant (4/44 (9.1%) versus 1/48 (2.1%); RR 4.36, 95% CI 0.51 to 37.57). Mortality at six months was higher in those allocated open reduction in Gray 1988, but the difference between the two groups was again not statistically significant (7/27 (25.9%) versus 2/22 (9.1%) at six months, RR 2.85, 95% CI 0.66 to 12.37).
Gray 1988 reported on the Garden alignment index for fracture reduction. The anteroposterior alignment was 164 degrees for open reduction and 160 degrees for closed reduction. For the lateral angles the figures were 169 degrees and 181 degrees respectively. The mean distance in millimeters of the screw tips to the joint was 5.9 mm for closed reduction and 5.8 mm for open reduction. None of these differences between groups was statistically significant.
Where trialists have used the term 'avascular necrosis' it has been retained. It is possible that most had segmental collapse of the femoral head associated with avascular necrosis.
Impaction of the fracture
Only one poor quality study tested the effects of fracture impaction versus control (Jacobsson 1985). Though also randomly allocated, the use of two different types of implant for internal fixation may have affected the outcome. The only outcome measure was bone scintimetry, assessed at around 10 days after surgery, with no reporting of outcomes of direct clinical relevance. The scintimetry results indicated that for displaced fractures there was a significant reduction in the femoral head viability for fractures that had been impacted at the time of surgery. Though these results were not statistically significant for undisplaced fractures, this may in part reflect the fewer numbers of participants with these fractures. It is not possible to determine if the difference in bone scintimetry results produced any clinical improvement although it may be expected that there would be fewer fracture healing complications in those participants with good femoral head viability.
Compression versus no compression of the fracture
Frandsen 1984 was presented as being a comparison of compression versus no compression of the fracture, but the article indicates that those treated without compression also had intra-operative "heavy" impaction of their fractures. It is not clear if no impaction was used for those treated with compression. This importantly changed the actual comparison tested by this trial. This quasi-randomised trial has a high risk of serious bias, especially resulting from the method of treatment allocation where participants were in fact allocated to two different orthopaedic departments on alternating days. The significant finding for non-union in surviving participants at one year should be viewed in the context of the absence of healing complications data for participants who had died. In addition, the excess of deaths in the impaction group is noteworthy. An analysis of unfavourable outcome of either non-union or death at one year produced similar results for the two treatment groups. These findings and the flawed methodology of this study indicate that the healing complication results have to be viewed with caution.
Open versus closed reduction
The unexplained disparity in numbers between the conference abstract and the data provided by the trial authors for Gray 1988 is a cause for concern and we are not sure of the reliability of the data presented here. However, the direction of effect for failure rate (in favour of open reduction) and mortality (in favour of closed reduction) were similar in the abstract and subsequent data. Gray 1988 reported that all the deaths occurred in those aged over 80 years (7/14 versus 2/13); whether there was another reason for this difference in the mortality in this age group is not known. Nonetheless, in the results for Gray 1988, only length of surgery was statistically significantly different between the two groups, in favour of closed reduction. An increased length of surgery for open reduction is to be expected and a similar finding was reported in Upadhyay 2004. The important differences between the two trials, especially in terms of study populations, have already been described and it is more appropriate to view the results of these trials separately. Even so, Upadhyay 2004 also found no statistically significant differences between the open and closed reduction for the outcomes measured aside from length of surgery.
Implications for practice
There is insufficient evidence from randomised trials to confirm the relative effects of open versus closed reduction of intracapsular fractures, or the effects of intra-operative impaction or compression of an intracapsular fracture treated by internal fixation.
Implications for research
Further clinical studies would be required to determine if impaction or compression of an intracapsular fracture at the time of internal fixation are of any benefit.
Dr Yvonne Dynan was co-author for the original version of this review in 2000. We are indebted to Dr Helen Handoll for her help. We also thank the following for useful comments at editorial review: Professor William Gillespie, A/Professor Peter Herbison, Professor James Hutchison, Professor Rajan Madhok, Professor Gordon Murray, Mr John Stothard, Professor Marc Swiontkowski and Dr Janet Wale
Data and analyses
- Top of page
- Authors' conclusions
- Data and analyses
- What's new
- Contributions of authors
- Declarations of interest
- Sources of support
- Index terms
Last assessed as up-to-date: 17 February 2005.
Protocol first published: Issue 3, 1999
Review first published: Issue 3, 2000
Contributions of authors
Martyn Parker initiated and designed the review, contacted trialists for further information and compiled the review. Abhijit Banajee independently extracted data, checked data entry and the review text and conclusions. Martyn Parker is the guarantor of the review.
Declarations of interest
Sources of support
- Peterborough and Stamford Hospitals NHS Foundation Trust, Peterborough, UK.
- No sources of support supplied
The title of the review differs from that in the protocol (Surgical techniques for internal fixation of intracapsular proximal femoral fractures) in order to more clearly distinguish this review from others which also investigate aspects of intracapsular hip surgery.
Medical Subject Headings (MeSH)
MeSH check words
* Indicates the major publication for the study