Large, well-controlled RCTs comparing different surgical treatments are rare. There are many small studies, often with significant limitations, making it difficult to formulate clear recommendations. Many aspects of surgical management are currently being reviewed by the Cochrane Collaboration. Additional information will also be available from a multicentre prospective randomised controlled trial (the STARS project - Scottish Trial of Arthroplasty or Reduction and fixation in Subcapital hip fractures) which is due to report in 2002 on completion of two years follow up.

7.1 Surgical experience

Evidence suggests that the best results are obtained when hip fracture operations are undertaken by an experienced surgeon.^53,92 The Scottish Hip Fracture Audit has shown considerable variation in the grade of surgeon performing hip fracture surgery.⁷⁴ Although there is no association between the grade of surgeon and mortality, the duration of surgery and incidence of postoperative complications are reduced and outcomes improved with an experienced surgeon.^66,70

7.2 Types of fracture

Hip fractures are classified as intracapsular or extracapsular depending on the site of the fracture in relation to the insertion of the capsule of the hip joint (indicated with an arrow in Figure 1) onto the proximal femur.

Figure 1: Classification of fractures of the proximal femur (hip fractures)

Intracapsular fractures include subcapital and transcervical fractures, and are best subdivided into undisplaced or displaced. Older classifications, such as Garden grades I-IV, offer no further diagnostic, therapeutic or prognostic information.

Extracapsular fractures include per-, inter- and sub-trochanteric, and are best subdivided by their degree of comminution. Basal cervical fracture lines tend to be approximately at the level of the insertion of the joint capsule, and they behave as extracapsular fractures (and should be regarded as such for prognostic and therapeutic considerations).

7.3 Treatment of intracapsular fractures

The treatment of intracapsular hip fractures has stimulated vigorous debate for decades, but with remarkably little good evidence to support clearly one option over another.

7.3.1 UNDISPLACED INTRACAPSULAR FRACTURES

The limited evidence available suggests that there is little difference in outcome between operation and conservative treatment of undisplaced fractures.^109,110 However, surgical treatment allows early mobilisation of the patient and reduces the risk of untreated undisplaced fractures becoming displaced at a later date. Undisplaced intracapsular fractures that are treated surgically should be treated by internal fixation.^{58, 111} Evidence level 3

A meta-analysis of 25 RCTs including 4,925 patients did not demonstrate evidence of the superiority of one device over another, or any benefit from the presence of a side-plate in the treatment of displaced or undisplaced intracapsular fractures.¹¹¹ The meta-analysis did suggest that multiple screws (up to three) were more reliable than pins. Evidence level 1+

There is some suggestion from a small prospective audit that the very elderly (aged over 80 years) may be better served by prosthetic replacement,¹¹² because of the higher risk of failure of fixation. Evidence level 3

Most undisplaced intracapsular hip fractures that are treated surgically should have internal fixation, except in the very elderly, when hemiarthroplasty may be considered.

7.3.2 DISPLACED INTRACAPSULAR FRACTURES

There is no single surgical procedure which has been shown to give the best outcome in all groups of patients with this injury.¹¹³ Two randomised trials indicate that both internal fixation and arthroplasty produce similar final outcomes, but internal fixation has a marginally lower mortality at the expense of an increased re-operation rate:^114,115 some studies have reported that following open reduction and internal fixation the reoperation rate is up to 30%.^116,117

The results of hemiarthroplasty are initially better, but if the patient survives more than three to five years, then function deteriorates. The results from total hip replacement may be better than those for hemiarthroplasty after three years, but a higher incidence of early dislocation is reported.^{118,119,120,121} Results of secondary THR following failure of fixation are better than the results of hemiarthroplasty after a number of years from the initial injury.¹²² Therefore many factors other than the type of fracture must to be considered when deciding surgical approach and choice of implant. These include age, previous physical mobility, previous mental agility, condition of the bone and joint (e.g. presence of arthritis). ¹¹¹ Evidence level 2++

Assessment prior to surgery must consider the patient's: age mobility mental state pre-existing bone and joint pathology.

The Scottish Hip Fracture Audit demonstrated the widespread nature of current clinical practice, with primary reduction and internal fixation of displaced intracapsular hip fractures in younger patients ("biologically" aged less than 65-70 years), and arthroplasty in older patients to reduce healing complications.¹

The complications from internal fixation are dependent upon the quality of the reduction.^{122,123,124,125,126} A meta-analysis of 106 papers showed a re-operation rate of 20-36% after internal fixation compared with 6-18% after hemiarthroplasty.¹¹³ Other studies have suggested reoperation is more common in older patients.^119,127 A rigorous analysis of the Scottish Hip Fracture Audit unitary database of over 12,000 hip fractures has shown a reoperation rate of 17% after internal fixation, compared to 5% after hemiarthroplasty in over 3,300 displaced intracapsular fractures (all age groups).¹ There is a marked difference in management of this type of fracture between Scandinavia, where internal fixation is the preferred treatment, and the UK. It is therefore difficult to generalise from the results of Scandinavian studies to the target population of this guideline.

Surgical techniques for internal fixation

A recent Cochrane review considered surgical techniques for the internal fixation of intracapsular fractures.¹²⁸ Techniques included the impaction of the fracture during surgery, compressing the fracture, and performing an open or closed reduction of a displaced fracture. The review concluded that there was insufficient evidence to determine the relative effectiveness of any of these techniques. As outlined in the surgical treatment of undisplaced intracapsular fractures, a meta-analysis did not demonstrate evidence of the superiority of one device over another, or any benefit from the presence of a side-plate.¹¹¹

Younger, active, fit patients should be considered for internal fixation. Active patients with an anticipated survival of more than a few years should be considered for internal fixation, total hip replacement or hemiarthroplasty, depending on the patient factors outlined above. Patients with an anticipated survival of less than three years and patients whose activity level is low should be considered for hemiarthroplasty. Bed or chair bound patients may be treated conservatively.

7.3.3 TYPES OF HEMIARTHROPLASTY

Hemiarthroplasties may be either unipolar (e.g. Thompson and Austin Moore) or bipolar (e.g. Hastings). Either type may be uncemented or cemented into the femur.

Cemented vs. uncemented stems

The use of bone cement has been associated with intra-operative morbidity. This can be reduced by intramedullary lavage and modern cementing techniques.^129,130 Uncemented stems are associated with more thigh pain and poorer overall function.^131,132,133 Evidence level 2+

Cement should be used when undertaking hemiarthroplasty, unless there are cardiorespiratory complications.

Unipolar vs. bipolar hemiarthroplasty

Radiological studies have suggested that, in many patients, bipolar prostheses move almost entirely at the outer articulation,¹³⁴ and therefore simply act as expensive unipolar prostheses. The main theoretical benefit of a bipolar prosthesis is a reduction in the amount of acetabular wear, minimising pain, joint destruction and mobility problems. Such problems appear to be directly related to the patient's activity levels (degree of mobility and independence of living) and the time since operation.¹³⁵ There does not appear to be any good evidence to show any significant advantage from using bipolar hemiarthroplasty in favour of unipolar hemiarthroplasty.^133,136,137 Evidence level 2++

Bipolar hemiarthroplasty should not be performed in preference to unipolar hemiarthroplasty, as there is limited evidence of any clinical benefit.

Surgical approach for hemiarthroplasty

The common surgical approaches for hemiarthroplasty for intracapsular hip fractures are anterolateral or posterior. Dislocation^138,139 and thrombosis are more common with the posterior approach, but increased operative time, blood loss and infection are more common with the anterior approach.^140,141 Evidence level 2+

The anterolateral approach is recommended for hemiarthroplasty surgery.

7.3.4 THE ROLE OF TOTAL HIP REPLACEMENT

Cohort studies provide conflicting results on the outcome of total hip replacement (THR) as the primary treatment for a hip fracture.^118,121 Dislocation rates of between 10-20% can be expected,¹⁴² but generally, prognosis is good. After three years THRs appear to be doing better than hemiarthroplasties^118,133,143 Evidence level 3

Further information on the outcome of THR as a primary treatment for displaced intracapsular hip fractures in "fit" patients will be available from the STARS trial.

THR as a secondary procedure after failed internal fixation performs better than hemiarthroplasty.¹²² The results of THR after failed hemiarthroplasty are similar to the results after revision for primary THR, although there is a higher complication rate.¹⁴⁴

In patients with pre-existing joint disease, medium/high activity levels and a reasonable life expectancy, THR may be appropriate as the primary treatment.

7.4 Treatment of extracapsular fractures

The standard treatment of extracapsular fractures is operative. The alternative, conservative treatment with prolonged bed rest, is not practised in this country. In elderly patients conservative treatment has been associated with a high incidence of morbidity and mortality, prolonged length of stay and high costs per quality adjusted life year (QALY).¹¹⁰ A systematic review⁷⁵ has not identified any major differences in outcome between these two approaches, but operative treatment appeared to be associated with less deformity, a reduced length of hospital stay and improved rehabilitation.

Extracapsular hip fractures should all be treated surgically unless there are medical contraindications.

The operative treatment of extracapsular fractures is almost always by reduction and internal fixation. This may be accomplished by using implants that are either extramedullary (e.g. sliding screw and plate) or intramedullary (e.g. Gamma nail).

7.4.1 EXTRAMEDULLARY VS. INTRAMEDULLARY FIXATION/ IMPLANTS

Extramedullary fixation by a sliding hip screw appears to give a lower complication rate than a fixed nail plate or intramedullary devices such as the Gamma nail, the IMHS, or condylocephalic implants such as the Ender nail.¹⁴⁵ Further studies are required to determine if the Gamma nail, or its modifications, have advantages for fractures such as subtrochanteric fractures and trochanteric fractures with a reversed obliquity fracture line.¹⁴⁶ These implants, and the indications for their use, continue to evolve.

7.4.2 OSTEOTOMY

It has been proposed that the fixation of unstable extracapsular hip fractures can be improved by an osteotomy to change the displacement and angle of the proximal femur. However, a recent systematic review¹²⁸ found inadequate evidence of any benefits from the routine use of osteotomy in conjunction with fixation by a sliding hip screw for an unstable trochanteric hip fracture.^147,148

Osteotomy is rarely indicated, but may be relevant if used in conjunction with a fixed nail plate.

7.4.3 COMPRESSION

There is only limited amount and poor quality evidence to support the application of compression across the fracture site of a trochanteric fracture during sliding hip screw fixation.^128,149

7.5 Blood transfusion

A retrospective study of 8,787 hip fracture patients, aged >=60 years, found that perioperative transfusion had no effect on mortality in patients with haemoglobin levels >=80 g/l.¹⁵⁰ However, several smaller studies have suggested that patients with known cardiac disease may benefit from transfusion at higher haemoglobin levels.^151,152,153 For further information see the SIGN guideline on perioperative blood transfusion.¹⁵⁴