Description of the condition
The majority of hip fractures occur in older people with an average age of around 80 years. The ratio of females to males with this injury is around four to one and the injury is usually the result of a simple fall. People suffering a hip fracture frequently have other medical and physical problems associated with ageing.
Hip fractures, also termed proximal femoral fractures, can be subdivided into intracapsular (also termed subcapital and transcervical) fractures and extracapsular (also termed trochanteric, intertrochanteric, pertrochanteric and subtrochanteric) fractures. Intracapsular fractures are those occurring proximal to the attachment of the hip joint capsule to the femur at the intertrochanteric line. Extracapsular fractures are those which traverse the femur within the area of bone bounded by the intertrochanteric line proximally up to a distance of five centimetres from the distal part of the lesser trochanter. Numerous subdivisions and classification methods exist for these fractures. 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 (Parker 1999). Extracapsular fractures may also be subdivided into undisplaced and displaced fractures and the displaced fractures may be further subdivided into different groups dependent on the degree of fracture comminution (Parker 2002). The distinction between intracapsular and extracapsular fractures is made in this review.
Description of the intervention
The treatment of a hip fracture may be either conservative or operative. Conservative treatment is also termed non-operative treatment and may entail a period of bedrest, restricted mobilisation, traction or even reduction of the fracture and application of plaster. Some of these methods of treatment, such as the use of plaster, would now be considered of historical interest only. Operative treatment may entail either preserving the femoral head by stabilising the fracture using various implants (internal fixation) or replacing the femoral head with an artificial hip prosthesis (arthroplasty).
How the intervention might work
Undisplaced (impacted) intracapsular fractures
Conservative treatment for undisplaced intracapsular fractures has been described in the literature and its use, as primary treatment, advocated by some (Raaymakers 1991). Opinion differs as to whether conservative treatment is appropriate only for those fractures which are impacted or if it may be used for those fractures which are truly non-displaced (not impacted or displaced). Conservative treatment of an undisplaced intracapsular fracture may involve a period of bed rest followed by mobilisation which may be either unrestricted or with limitations on weight bearing or activities, depending on the physician's preference.
Operative treatment for undisplaced intracapsular fractures generally entails preservation of the femoral head by internal fixation. This involves passing single or multiple screws or pins across the fracture site under X-ray guidance. An alternative is to replace the femoral head by an arthroplasty; this method of treatment is generally considered by most surgeons to be inappropriate for this type of fracture.
The main fracture healing complications that occur after an undisplaced intracapsular fracture are failure of the fracture to heal (non-union) and collapse of the femoral head (avascular necrosis or segmental collapse). The latter complication is due to damage to the blood supply to the femoral head. The term non-union customarily refers both to those fractures which fail to heal, and to those fractures which displace.
Displaced intracapsular fractures
Historically these fractures were treated by closed reduction and the application of plaster or traction, with the hope that the fracture would heal. With the advent of surgical techniques, surgery is now almost invariably used to treat this fracture. This is because if left untreated, non-union of the fracture, which may result in a painful and functionless hip, is very common. Surgical treatment may either entail reduction and internal fixation, or replacement arthroplasty. Conservative treatment generally involves nursing care only, with the use of appropriate analgesia; there is commonly an underlying acceptance that the fracture will fail to heal.
Conservative treatment for extracapsular hip fractures entails applying traction to the injured limb either via a pin inserted into the tibia (skeletal traction), or using adhesive tape or bandages applied to the injured leg (skin traction). Between four to nine kilograms of weight is then applied to reduce the fracture. Traction is then maintained whilst the fracture heals, a period of two to four months.
Operative treatment was introduced around the 1950s using a variety of different implants. The advantage was that it allowed earlier mobilisation of the patient and thereby avoided some of the complications of recumbency in the elderly. The most commonly used implant is the sliding hip screw (SHS): this is synonymous with the term compression hip screw and equivalent models such as the Dynamic, Richards or Ambi hip screws. It consists of a screw passed up the femoral neck to the head and then attached to a plate on the side of the femur. These are considered as 'dynamic' implants as they have the capacity for sliding at the plate/screw junction to allow for collapse at the fracture site. Fixed nail plates such as the Jewett, Thornton and McLaughlin nail plates are similar but have no capacity for sliding and hence are termed 'static' or 'fixed' implants. An alternative form of internal fixation uses an intramedullary nail with a side screw passed up into the femoral head. Examples of such a nail include the Gamma nail and Intramedullary Hip Screw. Cochrane reviews comparing different types of implants used for extracapsular hip fractures are available (for example, Parker 2006; Parker 2008).
Why it is important to do this review
The importance of this review has diminished since it was first published (the review evaluating this comparison for extracapsular hip fractures was first published in Issue 1, 1997; the scope was subsequently extended to cover all hip fractures in 2000). This is because the context has changed considerably. Advances in surgery and anaesthesia have lowered the risks associated with surgery. The benefits of early mobilisation and rehabilitation have become clearer and these, together with economic and health imperatives of avoiding prolonged hospital stay, mitigate against conservative treatment. Far more emphasis is now placed on the possible deleterious effects of delayed surgery, rather than questioning whether surgery is required. Nonetheless, it remains important to present what evidence from randomised trials there is to inform practice. Moreover, in some areas of the world, particularly where low incomes or different attitudes to surgery apply, this question remains relevant to current clinical decision making.
To compare the relative effects (fracture fixation complications, post-operative complications, anatomical restoration, and final outcome measures) of conservative treatment versus operative treatment for all types of hip fracture in adults. Results for the treatment of extracapsular and intracapsular hip fractures were analysed separately. Furthermore, an a priori distinction was made between undisplaced and displaced intracapsular hip fractures.
Criteria for considering studies for this review
Types of studies
All randomised controlled trials comparing conservative treatment with operative treatment. Quasi-randomised trials (for example, alternation) and trials in which the treatment allocation was inadequately concealed were considered for inclusion.
Types of participants
Skeletally mature patients with a proximal femoral fracture.
Types of interventions
All forms of conservative treatment of the fracture including those involving traction and bed rest. Operative treatment (here the control group) involving either internal fixation of the fracture or arthroplasty.
Types of outcome measures
Data for the following outcomes were sought:
(1) Fracture fixation complications
- non-union of the fracture (the definition of non-union was that used within each individual study and will include early displacement or loss of position of the fracture)
- avascular necrosis of the femoral head
- subsequent operation or reoperation (within the follow-up period of the study)
For operative treatment only
- fracture of the femur (around or below the implant)
- 'cut-out' of the implant proximally (penetration of the implant from the proximal femur either into the hip joint or external to the femur).
- breakage of the implant
- separation of the two components of the implant
- other major complications of fixation as detailed in each individual study
- fracture fixation failure rate (a summation of the above five items)
- wound infection: overall and by severity:
- 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)
For conservative treatment only (omitted in original protocol)
- pin tract infection
- pin loosening
- traction pressure sores
(2) Postoperative, clinical course and complications
- pressure sores (sacral and heel)
- thromboembolic complications (deep vein thrombosis, pulmonary embolism)
- any medical complication including mental confusion (as detailed in each individual study)
- transfusion requirements
- length of hospital stay (in days)
- number of patients not discharged from hospital at the end of follow-up period
- cost of treatment
(3) Anatomical restoration
- shortening of the leg (of more than 2 cm)
- varus deformity
- external rotation deformity (of more than 20 degrees)
(4) Final outcome measures (preferably at least six months from injury)
- mortality (within the follow-up period of the study)
- pain (persistent pain at the final follow-up assessment)
- failure to return to place of residence; loss of independence
- failure to regain mobility
- failure to regain activities of daily living
- failure to regain in health related quality of life/functional assessment outcomes
Search methods for identification of studies
We searched the Cochrane Bone, Joint and Muscle Trauma Group Specialised Register (March 2008), the Cochrane Central Register of Controlled Trials (The Cochrane Library 2008, Issue 1), MEDLINE (1966 to March week 4 2008) and EMBASE (1988 to 2008 Week 14). We searched Current Controlled Trials at www.controlled-trials.com (accessed April 2008) for ongoing and recently completed trials. No language restriction was applied.
The generic search strategies for hip fracture trials in The Cochrane Library (Wiley Interscience) and MEDLINE (2004 onwards) are shown in Appendix 1. This MEDLINE search was combined with all three stages of the optimal trial search strategy (Higgins 2006). The general search strategy for hip fracture trials in EMBASE (2004 onwards) is shown in Appendix 2.
Searching other resources
We searched reference lists of articles and our own reference databases. We checked the findings from handsearches of the British Volume of the Journal of Bone and Joint Surgery supplements (1996 to 2006) and abstracts of the American Orthopaedic Trauma Association annual meetings (1996 to 2007: http://www.hwbf.org/ota/am/) and American Academy of Orthopaedic Surgeons annual meeting (2004 to 2007: www.aaos.org/wordhtml/libscip.htm). We also included handsearch results from the final programmes of SICOT (1996 & 1999) and SICOT/SIROT (2003), EFORT (2007) and the British Orthopaedic Association Congress (2000, 2001, 2002, 2003, 2005 and 2006). We scrutinised weekly downloads of "Fracture" articles in new issues of 15 journals (Acta Orthop Scand; Am J Orthop; Arch Orthop Trauma Surg; Clin J Sport Med; Clin Orthop; Foot Ankle Int; Injury; J Am Acad Orthop Surg; J Arthroplasty; J Bone Joint Surg Am; J Bone Joint Surg Br; J Foot Ankle Surg; J Orthop Trauma; J Trauma; Orthopedics) from AMEDEO (www.amedeo.com).
Searches conducted up to June 2004 were mainly those of the Cochrane Bone, Joint and Muscle Trauma Group Specialised Register, locally accessible conference proceedings and orthopaedic journals, reference lists of articles and contacting trialists.
Data collection and analysis
Selection of studies
Two review authors (MP and HH) independently assessed potentially eligible trials for inclusion. Any disagreement was resolved by discussion.
Data extraction and management
Data for the outcomes listed above were independently extracted by at least two review authors and any differences resolved by discussion. Where necessary and practical, we contacted trialists for additional data and clarification.
Assessment of risk of bias in included studies
In this review, risk of bias is implicitly assessed in terms of methodological quality. Each trial was assessed independently, without masking of authors or source, for its quality of methodology by two review authors. All differences were resolved by discussion. In total, 12 aspects of methodology were rated (see Table 1). The scores of the individual items were no longer summed from 2008 onwards.
Measures of treatment effect
For each study, relative risks (RR) and 95% confidence intervals (CI) were calculated for dichotomous outcomes. We planned also to present mean differences and 95% confidence intervals for continuous outcomes.
Assessment of heterogeneity
Heterogeneity between comparable trials was tested using a standard chi² test, with additional consideration of the I² statistic (Higgins 2003).
Trials evaluating treatment of intracapsular fractures were kept separate from those evaluating treatment of extracapsular fractures. Comparable groups of trials on extracapsular fractures were subgrouped by implant type (here, fixed nail plate or sliding hip screw). Where appropriate, data were pooled using the fixed-effect model using 95% confidence intervals (CI). Where there was significant heterogeneity (P < 0.10; I² > 50%) in the pooled results of individual trials, we planned to present these using the random-effects model.
Description of studies
No new trials were found on extending the search from June 2004 to March 2008.
Six studies were identified and all met the pre-defined inclusion criteria. However one (Harper 1994), which was originally available only as a conference abstract, was subsequently excluded as it was methodologically compromised (see 'Characteristics of excluded studies'). The five included studies involved a total of 428 elderly patients.
Two of the five studies were available as published reports (Bong 1981; Hornby 1989), two in abstract form (Hansen 1994; Sher 1985) and one was unpublished (Hornby 1975), with summary data provided by the principal trialist. A conference abstract for Hornby 1975 showed that 70 rather than 67 patients were originally included in the trial (Stevens 1980). Three studies were sited in Newcastle-upon-Tyne in UK, one (Hansen 1994) in Denmark and one (Bong 1981) in Hong Kong. The four studies on extracapsular fractures were conducted from the mid 1970s (Hornby 1975) up to the late 1980s (Hornby 1989), and the one study on intracapsular fractures (Hansen 1994) was conducted in 1994. Details of the individual studies are given in 'Characteristics of included studies'.
The one study on intracapsular fractures (Hansen 1994), compared conservative treatment (including early mobilisation on crutches) with operative treatment of internal fixation with a sliding hip screw in 23 people with impacted fractures.
All four studies on extracapsular fractures (Bong 1981; Hornby 1975; Hornby 1989; Sher 1985) used traction for conservative treatment. Surgical treatment varied. Fixed nail plates were applied in Hornby 1975 and Bong 1981. This was preceded by bone sectioning (osteotomy) in Bong 1981 where there were two surgical groups, one which had medial displacement osteotomy and the other which had valgus osteotomy. Sliding hip screw fixation using the AO dynamic hip screw was used in Hornby 1989. Sher 1985 applied either a sliding hip screw or a Kuntscher Y nail (an intramedullary device). Two studies selected unstable trochanteric fractures (Bong 1981; Sher 1985). Hornby 1989 included both stable and unstable trochanteric fractures. The type of fracture was only stated to be trochanteric in Hornby 1975.
Risk of bias in included studies
The method of randomisation was not stated for Bong 1981 or Hansen 1994. Hornby 1989 used sealed envelopes. Hornby 1975 and Sher 1985 were quasi-randomised trials using odd and even hospital record numbers and alternation respectively. The imbalance in the numbers allocated to the two groups (7 versus 16) was not explained in Hansen 1994.
The methodology scores for the five studies as assessed by the scoring system described earlier are given below:
1 2 3 4 5 6 7 8 9 10 11 12
1 1 0 0 0 0 0 0 1 1 1 0 Bong 1981
1 1 0 0 0 0 0 0 0 0 0 0 Hansen 1994
0 0 0 0 0 0 0 0 1 0 0 1 Hornby 1975
2 1 1 1 1 0 1 0 1 1 1 0 Hornby 1989
0 0 0 0 1 0 0 0 1 0 0 0 Sher 1985
The assessment of methodological quality is likely to reflect more the extent and quality of the trial reports than study design. For instance, Hornby 1989 was a comprehensive report whereas those of Hansen 1994, Hornby 1975 and Sher 1985 were very limited. An enquiry regarding Sher 1985 produced additional details about trial methodology but no information on patient numbers.
Effects of interventions
The outcome measures reported by each study are listed in 'Characteristics of included studies'. These are presented grouped within the four categories listed earlier. The results for intracapsular fractures are presented separately from those for extracapsular fractures. Where pertinent and when data were available, the results subgrouped by implant type (fixed or dynamic) of various outcomes are shown graphically; here only for extracapsular fracture trials. No quantitative results were available for Sher 1985. Data from the two operative groups were combined for Bong 1981.
Treatment of undisplaced intracapsular fractures
(1) Fracture fixation complications
The one study identified (Hansen 1994) involved only 23 participants. Non-union (termed secondary dislocation in this study) only occurred in conservatively-treated participants, at between 2 to 35 days after entry into the trial, (10/16 versus 0/7; relative risk (RR) 9.88, 95% confidence interval (CI) 0.66 to 148.48). The only case of avascular necrosis occurred in a participant of the operatively-treated group.
Nine of the 10 participants with non-union in the conservative group were operated on for secondary displacement; eight had hemiarthroplasty and one had a secondary internal fixation. The other participant with non-union did not have surgery due to poor health. Only one reoperation was undertaken in the internal fixation group. This was for avascular necrosis, which was treated by insertion of a hemiarthroplasty.
(2) Postoperative/clinical complications
These outcomes were not reported in Hansen 1994. Neither were hospital stay or cost of treatment.
(3) Anatomical restoration
These outcomes were not reported in Hansen 1994.
(4) Final outcome measures
These outcomes were not reported in Hansen 1994.
Treatment of extracapsular fractures
(1) Fracture fixation complications
There was no significant difference in non-union between the two treatments for two studies (Bong 1981; Hornby 1989) (3/86 versus 1/133; RR 3.04, 95% CI 0.48 to 19.03).
No reoperations, or secondary traction, were required for the operative group in Hornby 1989. Data for reoperations following internal fixation or operations following conservative treatment failure were not available for the other three trials. Nonetheless it is likely that some of 25 cases of fixation failure, including five cases of fracture collapse, in Bong 1981 resulted in reoperations. Hornby 1975 reported two cases of foot drop, both in the conservative treatment group.
Complications exclusive to each treatment group were reported by three studies (Bong 1981; Hornby 1975; Hornby 1989). Major complications associated with internal fixation were reported in Bong 1981. These comprised 20 breakages at the nail-plate junction and five cases of 'cut-out'. Hornby 1989 reported only five minor complications consisting of three cases of migration of the lag screw in the femoral head without penetration into the hip joint and two instances of backing out of the screw, both in one person. Wound infection in the fracture fixation group was reported by Bong 1981 (4 cases), Hornby 1975 (2 cases) and Hornby 1989 (no cases). Bong 1981 also reported avascular necrosis in one participant. Only Hornby 1989 mentioned complications specific to traction. These were loosening of the traction pin (20 cases), minor pin tract infection (8 cases), and skin sores from traction (5 cases).
(2) Postoperative/clinical complications
These were reported in all four studies but to varying degrees and only comprehensively in Hornby 1989. Bong 1981 only referred indirectly to these in the listing of causes of death.
There was no significant difference in "chest infection" or pneumonia (2/33 versus 5/34; RR 0.41, 95% CI 0.09 to 1.98) between conservative and operative treatment in Hornby 1975. Sher 1985 implied, without qualification, an increased incidence of fatal pneumonia in the conservative group.
Data presented for "deep" pressure sores (heel and sacral) from the two Hornby studies showed no significant difference between the two treatment groups. Sher 1985 reported that pressure sores were more prevalent in the conservative group. At six months follow up there were equal numbers of "unhealed sores" (5/40 versus 5/41) reported by Hornby 1989.
Other complications reported by Hornby 1989 were urinary incontinence (27 versus 34), faecal incontinence (22 versus 25), and mental confusion (19 versus 20), none of which were significantly different between the two groups. Sher 1985 reported an increased incidence of urinary infection in the conservatively-treated group.
The length of hospital stay was increased significantly for the conservatively-treated participants in Hornby 1989, who reported a difference in mean length of stay of 27 days at six months follow up. Hornby 1989 noted that this was an underestimate given that more in the conservative group remained in hospital at six-months follow up (10/51 versus 6/55; RR 1.80, 95% CI 0.70 to 4.59). As the distribution of the values for hospital stay, shown for the first 12 weeks in the trial report, did not approximate to a normal distribution, the derivation of an exact value for the mean difference is misleading. However, from 14 days onwards there were more people in the conservative group remaining in hospital at any time up to and beyond the six-months follow up (see Analysis 2.9 with the four, eight and 12-week figures). Sher 1985 stated that those in the conservative group remained in hospital on average for seven weeks longer than those treated operatively.
(3) Anatomical restoration
No study reported external rotational deformity. Data for varus deformity presented for Bong 1981 and Hornby 1989 show contrasting directions of effect and were not pooled. In Bong 1981, varus deformity was significantly greater in the operated group, whereas the converse was true in Hornby 1989. Sher 1985 stated that "significant shortening and varus deformity had occurred in over half the fractures treated by traction", with no apparent affect on function. Limb shortening of more than one centimetre was more prevalent in the traction group in Hornby 1989 (29/39 versus 11/37; RR 2.50, 95% CI 1.47 to 4.24). No conclusion could be drawn from Bong 1981 who reported the average degree of shortening in each group. Hornby 1989 also reported more rotation deformity (13/36 versus 24/36; RR 0.54, 95% CI 0.33 to 0.89, analysis not shown), less fixed flexion deformity (14/39 versus 0/35), and a greater incidence of persistent leg swelling (9/31 versus 5/35). Overall, in Hornby 1989, the surgically-treated participants achieved better results in terms of anatomical healing. Sher 1985 also reported anatomical restoration was greater in surgically-treated participants.
(4) Final outcome measures
There was no significant difference in mortality at six months for the two studies (Hornby 1975; Hornby 1989) providing mortality data (19/84 versus 21/89; RR 0.96, 95% CI 0.56 to 1.65). Sher 1985 stated there was a higher mortality with conservative treatment. Bong 1981 reported the death of 10 trial participants at two-year follow up without indicating their treatment groups.
The only study to report on residual pain (Hornby 1989) assessed this both by its frequency (none/ occasional/ daily) and intensity (mild/ moderate/ severe). There was no significant difference between the two treatment methods.
The recovery of functional status was depicted by a loss of independence based on residential status in Hornby 1989, and maintenance of mobility in Hornby 1975. At six-months follow up, 16 of the traction group for Hornby 1989 had lost their independence compared to nine of those treated operatively (RR 1.87, 95% CI 0.93 to 3.73). The combined outcome of loss of independence or mortality, still showed a worse outcome for the conservative group (27/51 versus 22/55; RR 1.32, 95% CI 0.87 to 2.00). Hornby 1975 found that fewer conservatively-treated participants (4/20 versus 8/21) had failed to maintain their pre-fracture mobility.
Well conducted and reported randomised trials comparing conservative and operative treatment for hip fractures are disappointingly few. This means that firm and definite conclusions cannot be made from the information available. The relevance of the review to modern day practice in which a high proportion of patients are treated surgically, is diminishing. Some comments on the findings for the two fracture types are given below.
Trials of treatment of undisplaced intracapsular fractures
The one identified study (Hansen 1994) involved only 23 participants, was inadequately reported and may have been subject to serious bias. The imbalance in the numbers allocated to the two groups was unexplained. There were no details of the interventions used for conservative management and no information on the patient characteristics of the surgically-treated group and only the age range for those in the conservatively-treated group. The limited results presented in the abstract suggest that internal fixation of this fracture could markedly reduce the risk of non-union and the requirement for a subsequent operation, generally the replacement of the femoral head with an arthroplasty. These findings are similar to that of a summary of case series of internal fixation (Conn 2004). However, the small numbers of patients involved, the potential for serious bias and the failure to report other outcomes (such as operative complications) means that no definite conclusions can be safely drawn from the results of the one trial on intracapsular fractures included in this review.
Trials of treatment of extracapsular fractures
Whilst conservative treatment has remained relatively unchanged, there have been major improvements in the results of operative treatment over the years connected to the refinement of surgical technique and implant design. This accentuates the unsatisfactory nature of the evidence as some of the implants tested in the included trials have been largely superseded. Thus the results of two studies using fixed nail plates for the operative group of extracapsular fractures (Bong 1981; Hornby 1975), are of restricted value as these implants are less commonly used, have a high failure rate and are almost certainly inferior to the sliding hip screw (Parker 2006).
Since there are no quantitative data for Sher 1985, the only study which provides evidence for current practice is Hornby 1989. This study included 106 out of 134 people aged 60 years or more with an extracapsular fracture admitted to hospital over a 12-month period. The 28 patients who were not entered into the study either refused, were unable to give consent or failed to meet the criteria determining fitness to undergo an operation. The results of Hornby 1989 showed no difference in general medical complications nor long-term mortality, pain and unhealed sores between the two methods of treatment. Surgical complications were few and minor perhaps reflecting the strict protocol followed by the surgeon involved. Rates of non-union of the fracture appeared to be similar in either method of treatment, but operative treatment is less likely to result in limb shortening. The prolonged period of bed rest incurred with conservative treatment resulted in a longer hospital stay with more participants of the conservatively-treated group remaining in hospital at six-months follow up. Although not statistically significant, more participants of the conservative group had lost their independence as assessed by their post-fracture residential status at six months.
No attempt was made to quantify cost in any of the trials, although Hornby 1989 gave a useful reminder of the importance of a general approach taking account of opportunity costs and effects of dependence and limited mobility. Indeed, while surgical treatment will involve some additional costs for the operation and implant, these costs will be greatly offset by the prolonged hospital stay incurred by conservative treatment. An economic analysis using the figures from Hornby 1989 (Parker 1992), reported that conservative treatment would cost about 6000 pounds sterling per patient against 3446 pounds sterling for operative treatment in UK at that time.
Implications for practice
The introduction of operative treatment preceded the first evaluation by randomised trial, and the subsequent randomised trials have been few. The limited evidence from one small randomised trial on undisplaced intracapsular fractures suggests conservative treatment is associated with an increased risk of fracture displacement and later replacement of the femoral head with an arthroplasty. For extracapsular fractures, conservative treatment appears to be associated with a prolonged hospital stay and a greater proportion of patients losing their independence.
The review need to be viewed in the context of advances in surgery and anaesthesia, and clearer understanding of the benefits of early mobilisation and rehabilitation. Clinicians should use evidence from other non-randomised studies in conjunction with their personal experience, patient preference and availability of surgical facilities, to influence their choice of treatment.
Implications for research
Although there is little robust data from randomised trials comparing non-operative with operative management, it is difficult in contemporary practice to conceive circumstances in which future trials would be possible or considered ethical. However, potential exceptions should not be ruled out. For example, despite access to good surgical facilities centres remain in which patients are treated conservatively. Whatever the underlying rationale, any future studies should employ sound methodology including allocation concealment, complete and long-term follow up, and blinded assessment. They should compare available implants of proven effectiveness with good quality conservative treatment. A full evaluation of clinical and functional outcomes, cost effectiveness and social consequences would also be important. Where the provision of surgical services is limited or where there are strong preferences for non-surgical management, trials evaluating conservative interventions are likely to be more appropriate.
We are grateful to Mr R Hornby for providing information on his unpublished study. We thank Dr Amit Bhargava for his earlier contribution as a co-author to the review.
We thank Prof Bill Gillespie for his helpful suggestions throughout the history of the review. We would also like to thank the following for useful comments at editorial review: Prof Bill Gillespie, Prof Rajan Madhok, Prof Gordon Murray, Ms Bev Shea, Prof Marc Swiontkowski, Prof George Wells, Dr Janet Wale, and Prof Rob de Bie.
We thank Dr Joanne Elliott and Mrs Lesley Gillespie for continuing the search for trials for this review.
Dr Helen Handoll's work on the first two versions of the review was supported by the Chief Scientist Office, Department of Health, The Scottish Office, UK.
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
Appendix 1. Search strategies for The Cochrane Library and MEDLINE
Appendix 2. Search strategy for EMBASE
Last assessed as up-to-date: 3 April 2008.
Protocol first published: Issue 1, 1997
Review first published: Issue 2, 1998
Contributions of authors
Martyn Parker initiated and designed the review, and compiled the first draft. Helen Handoll located the review studies and critically rewrote the first draft and subsequent drafts. Helen Handoll initiated and prepared the first drafts of the review updates. All other tasks were shared. On expansion of the review in 2000 to include all hip fractures, Amit Bhargava joined as a co-author. Helen Handoll and Martyn Parker are guarantors of the review.
Declarations of interest
Sources of support
- University of Teesside, Middlesbrough, UK.
- Peterborough and Stamford Hospitals NHS Foundation Trust, Peterborough, UK.
- No sources of support supplied
Details of updates prior to Issue 3, 2008.
(A) Non substantive changes made (January 2000) for Issue 2, 2000.
(1) Literature search extended to August 1999.
(B) Minor update Issue 3, 2000.
(1) Synopsis added
(C) Substantive update for Issue 4, 2000.
(1) Scope of the review extended to all hip fractures. Previous review title was 'Conservative versus operative treatment for extracapsular hip fractures'.
(2) Inclusion of one new study on intracapsular hip fractures.
(3) Relative risks instead of Peto odds ratios presented for dichotomous outcomes.
(D) Minor update for Issue 1, 2002.
(1) Literature search extended to July 2001.
(2) Extra references added for one excluded and one included study.
(3) Implications for practice section rephrased.
(E) Minor update for Issue 4, 2004.
(1) Literature search extended to June 2004.
(2) Text and tables modified to conform to style guidelines and revised guidelines in Cochrane Reviewer's Handbook.
Medical Subject Headings (MeSH)
MeSH check words
* Indicates the major publication for the study