• Admitted to hospital
• Outpatients post-discharge

• Anti-embolism stockings (AES) (above or below knee)
• Intermittent pneumatic compression (IPCD) devices (full leg or below knee)
• Foot pumps or foot impulse devices (FID)
• Electrical stimulation (including Geko devices)
• Continuous passive motion

• Unfractionated heparin (UFH) (low dose, administered subcutaneously)
• Low molecular weight heparin (LMWH), licensed in UK:

  • enoxaparin (standard prophylactic dose 40mg daily; minimum 20mg daily* to maximum 60mg twice daily*)

    • dalteparin (standard prophylactic dose 5000 units once daily; minimum 1250 units once daily* to maximum 5000 units twice daily*; obese patients – maximum 7500 twice units daily*)
    • tinzaparin (standard prophylactic dose 4500 units once daily; minimum 2500 units once daily* to maximum 4500 units twice daily*; obese patients – maximum 6750 twice daily*)
• LMWH, licensed in countries other than UK:

  • Bemiparin (standard 2500 units daily; minimum 2500 units daily to maximum 3500 units daily)
  • Certoparin (3000 units daily)
  • Nadroparin (standard 2850 units once daily; minimum 2850 units once daily to maximum up to 57 units/kg once daily)
  • Parnaparin (standard 3200 units once daily; minimum 3200 units once daily to maximum 4250 units once daily)
  • Reviparin (minimum 1750 units once daily to maximum 4200 units once daily)
• Vitamin K Antagonists:

  • warfarin (variable dose only)
  • acenocoumarol (all doses)
  • phenindione (all doses)
• Fondaparinux (all doses)*
• Apixaban (all doses)*
• Dabigatran (all doses)*
• Rivaroxaban (all doses)*
• Aspirin (up to 300mg)*

• Other VTE prophylaxis treatment, including monotherapy and combination treatments (between class comparisons for pharmacological treatments only)
• No VTE prophylaxis treatment (no treatment, usual care, placebo)

• Above versus below knee stockings
• Full leg versus below knee IPC devices
• Standard versus extended duration prophylaxis
• Low versus high dose for LMWH
• Preoperative versus post-operative initiation of LMWH

• All-cause mortality (up to 90 days from hospital discharge)
• Deep vein thrombosis (symptomatic and asymptomatic) (7–90 days from hospital discharge). Confirmed by: radioiodine fibrinogen uptake test; venography; Duplex (Doppler) ultrasound; MRI; Impedance Plethysmography (used as rule out tool)
• Pulmonary embolism (7–90 days from hospital discharge). Confirmed by: CT scan with spiral or contrast; pulmonary angiogram; ventilation/ perfusion scan including VQSpect; autopsy; echocardiography; clinical diagnosis with the presence of proven VTE
• Major bleeding (up to 45 days from hospital discharge). A major bleeding event meets one or more of the following criteria: results in death; occurs at a critical site (intracranial, intraspinal, pericardial, intraocular, retroperitoneal); results in the need for a transfusion of at least 2 units of blood; leads to a drop in haemoglobin of ≥2g/dl; a serious or life threatening clinical event. Includes unplanned visit to theatre for control of bleeding
• Fatal PE (7- 90 days from hospital discharge). Confirmed by: CT scan with spiral or contrast; pulmonary angiogram; ventilation/ perfusion scan including VQSpect; autopsy; echocardiography; clinical diagnosis with the presence of proven VTE

• Clinically relevant non-major bleeding (up to 45 days from hospital discharge): bleeding that does not meet the criteria for major bleed but requires medical attention and/or a change in antithrombotic therapy.
• Health-related quality of life (validated scores only)(up to 90 days from hospital discharge)
• Heparin-induced thrombocytopaenia (HIT) (duration of study)
• Technical complications of mechanical interventions (duration of study)
• Infection (duration of study)

1.5.5.

Offer VTE prophylaxis for a month to people with fragility fractures of the pelvis, hip or proximal femur if the risk of VTE outweighs the risk of bleeding. Choose either:

• LMWH^d, starting 6–12 hours after surgery or
• fondaparinux sodium^e, starting 6 hours after surgery, providing there is low risk of bleeding. [2018]

1.5.6.

Consider pre-operative VTE prophylaxis for people with fragility fractures of the pelvis, hip or proximal femur if surgery is delayed beyond the day after admission. Give the last dose no less than 12 hours before surgery for LMWH^f or 24 hours before surgery for fondaparinux sodium^g. [2018]

1.5.7.

Consider intermittent pneumatic compression for people with fragility fractures of the pelvis, hip or proximal femur at the time of admission if pharmacological prophylaxis is contraindicated. Continue until the person no longer has significantly reduced mobility relative to their normal or anticipated mobility. [2018]

7.

What is the clinical and cost effectiveness of aspirin alone versus other pharmacological and/or mechanical prophylaxis strategies (alone or in combination) for people with fragility fractures of the pelvis, hip or proximal femur?

8.

What is the clinical and cost effectiveness of IPCD in combination with pharmacological prophylaxis strategies for people with fragility fractures of the pelvis, hip or proximal femur?

The committee considered all-cause mortality (up to 90 days from hospital discharge), deep vein thrombosis (symptomatic and asymptomatic) (7–90 days from hospital discharge), pulmonary embolism (7–90 days from hospital discharge), fatal PE (7–90 days from hospital discharge), and major bleeding (up to 45 days from hospital discharge) as critical outcomes.

The committee considered clinically relevant non-major bleeding (up to 45 days from hospital discharge), health-related quality of life (up to 90 days from hospital discharge), heparin-induced thrombocytopaenia (duration of study), and technical complications of mechanical interventions (duration of study) and infection (duration of study) as important outcomes.

Please see section 4.4.3 in the methods chapter for further detail on prioritisation of the critical outcomes.

Fifteen studies were included in this review; thirteen of the relevant studies were randomised controlled trials identified from the previous guideline (CG92). One new study was identified and one study published before CG92 is now included in this review. One of the previously included studies in this evidence review was excluded and moved to the major trauma review due to more appropriate applicability of the study population.

Nine comparisons were included; they evaluated both pharmacological and mechanical interventions. Pharmacological interventions included LMWH at standard dose and for a standard duration, UFH, fondaparinux (standard duration and extended duration), VKA and aspirin. Mechanical interventions included AES (length unspecified) and IPCD (thigh-length).

Discussion around the quality of the evidence centred largely on the inclusion of the PEP trial which was excluded from the previous guideline. The PEP trial is one of the larger trials conducted in this population that was published in 2000, evaluating the use of aspirin. The committee noted that the PEP trial allowed centres to include other prophylaxis. The data reported include just over 50% of patients with either LMWH or UFH, and around 30% using AES. It is not reported how many of these patients received both heparin and AES, or who had aspirin alone or no prophylaxis at all. The study also reported a post-hoc analysis for the combined outcome of pulmonary embolism and symptomatic DVT. This showed a reduction in symptomatic VTE events using aspirin (plus or minus AES) without the use of heparin and a reduction of symptomatic VTE events with AES (plus or minus the use of heparin). The outcomes of major bleeding or clinically relevant non-major bleeding were not adequately reported in the study and were therefore excluded from the current review. Overall, it was decided that the trial could be included on the basis of providing effectiveness information for the VTE outcomes for aspirin when combined with other prophylaxis, but not for aspirin alone, and that its effect on bleeding was still unknown.

Pharmacological and mechanical interventions versus no VTE prophylaxis

The committee discussed the need for prophylaxis in this population and appreciated that in a majority of the evidence where pharmacological or mechanical prophylaxis was compared with no prophylaxis, there were better outcomes in the group receiving an intervention. The committee noted that people with fragility fractures of the pelvis, hip and proximal femur tend to have a longer length of hospital stay; around 21 days for acute spells and 23 for super-spells (may include hospitals differential capture of rehabilitation length-of-stay).²²⁷ Patients have reduced mobility whilst in hospital, a factor that contributes to risk of VTE.

General consensus was that IPCD seemed effective as the clinical evidence presented showed clinical benefit for DVT (symptomatic and asymptomatic) and a possible clinical benefit for PE, although there was uncertainty associated with the PE result. The orthopaedic subgroup advised the committee that some hospitals use IPCD routinely in orthopaedic theatres and wards. The use of pharmacological interventions alongside IPCD is common practice but appreciated that there is an absence of RCT evidence evaluating the clinical effectiveness of this combination intervention in this population. It was therefore suggested that a research recommendation be proposed in order to encourage this evaluation.

Some members of the subgroup were of the view that the use of IPCD may discourage mobilisation. Therefore the subgroup and committee agreed to recommend IPCD only when pharmacological prophylaxis was contraindicated and only until people are able to mobilise themselves. Mechanical prophylaxis is recommended until the patient is back to normal mobility as the committee believe that mechanical prophylaxis offers little benefit once a patient is mobile.

LMWH at a standard dose for a standard duration versus other pharmacological interventions

The committee considered that the evidence sufficiently supports the use of LMWH and fondaparinux. It was discussed that UFH is not commonly used in current practice. It was previously recommend for patients with renal failure, but low doses of LMWH are currently used in practice instead for these patients.

The committee discussed the evidence presented for LMWH versus fondaparinux and noted that the clinical evidence suggests a higher clinical benefit of fondaparinux over LMWH, as seen in moderate quality evidence for a clinically important reduction in the rate of DVT with fondaparinux compared to LMWH. The committee considered other aspects of the interventions that were not listed as outcomes in the review, such as the half-life of each, with regard to considering situations where prophylaxis would need to be reversed. Fondaparinux has a half-life of 17 hours whereas LMWH has a much shorter half-life ranging from 2–5 hours depending on which preparation is used (according to summary of product characteristics). The committee decided to also recommend LMWH based on the effectiveness evidence showing a possible benefit when compared with no prophylaxis for DVT and PE, although there was uncertainty around these effect estimates. Recommending LMWH is in line with current practice as it is already widely used in this population and is not associated with a high bleeding risk, as is the case with fondaparinux. The committee discussed the major bleeding risk associated with fondaparinux and suggested that it only be used once haemostasis has been established and there is no risk of bleeding. The committee discussed the duration of prophylaxis and noted that the duration of VTE prophylaxis identified in the studies ranged between 28–31 days. The committee acknowledged that recommending VTE prophylaxis for a month is more pragmatic. The committee noted the increased benefit of an extended duration of fondaparinux as reported in one of the studies included in this evidence review.

Aspirin (± other prophylaxis) versus no aspirin (± other prophylaxis)

The PEP trial was discussed at length. The committee were aware that some of the orthopaedic community believe aspirin is an appropriate form of prophylaxis, and that the PEP trial provides evidence for its use in this population. The committee were also aware that aspirin is recommended in the American College of Clinical Pharmacy (ACCP) as a method of VTE prophylaxis in this population. The orthopaedic subgroup considered the evidence showed that aspirin alone is an effective method of prophylaxis and advised it should be recommended for this population. However, the committee was concerned about the lack of evidence for aspirin alone particularly around bleeding that is commonly associated with the use of aspirin. Therefore they did not consider that it should be recommended in this population. A research recommendation was proposed to investigate the effectiveness and safety of aspirin compared with the other routinely used pharmacological prophylaxis – LMWH, in people with fragility fractures of the pelvis, hip or proximal femur.

Combination comparison: UFH + AES versus AES alone

The committee noted that combination prophylaxis has limited benefit so suggested that the CG92 recommendation which recommends combined prophylaxis should not be adopted unless mobility is reduced. The committee expressed concerns about the overuse of AES in current practice within this population with little evidence of clinical benefit. It was also noted that AES are difficult to fit, applying them can be painful to the patient and they are not always worn properly. Therefore, it was agreed that the use of AES should not be specified in the recommendation. Although the committee believe that AES should not be routinely used they noted that they may be effective for patients with a high risk of bleeding.

Two economic models were developed for this population in CG92 and were included in this review. The first model compared all standard duration prophylaxis strategies. This analysis showed that fondaparinux (2.5 mg) was the most costeffective strategy, with an incremental net monetary benefit (INMB) of £2,148. This analysis was assessed as partially applicable, with potentially serious limitations.

The second model compared fondaparinux initiated post-operatively and continued for 10 days to no post-discharge prophylaxis. This analysis showed that fondaparinux was cost effective compared to no prophylaxis, with an INMB of £239. This analysis was assessed as directly applicable with potentially serious limitations.

Additionally, two studies were identified but were selectively excluded due to the availability of the more applicable models from CG92.

The committee discussed the relevance of the clinical evidence used in the CG92 model to the evidence included in the current review. It was acknowledged that there were differences between the interventions included in the model and those included in the current clinical review, where aspirin (high dose) is not used in clinical practice in the UK.

The committee also highlighted that there was no evidence to support the use of AES for lower limb fragility fractures and that they are difficult to fit, necessitating time from the nurses to ensure they are properly fitted and monitored. Hence, it was concluded that the routine use of AES in this population represents a financial burden on the NHS without evidence of cost effectiveness. The committee discussed the evidence available for the use of IPCD and concluded that this is the only mechanical prophylaxis method that has clinical and cost-effectiveness evidence to support its use in the early post-operative period until mobilisation. It was acknowledged that although there might be an upfront cost of providing IPCDs in hospitals, this is likely to be offset by the saving achieved from not using AES and the standardisation of practice. It was also highlighted that, in most cases, IPCDs are provided rent-free to hospitals and the only cost involved would be that of the sleeves. Additionally, IPCDs are used for a shorter period of time until mobilisation.

The committee discussed the evidence for pharmacological prophylaxis in this population and noted that the CG92 model showed the cost effectiveness of LMWH (standard dose) and fondaparinux compared to no prophylaxis. Based on the clinical evidence in this update and the trade-off between clinical benefits and harms, the committee decided to retain the CG92 recommendation of these options, giving clinicians the ability to choose between them based on clinical and individual factors.

The orthopaedic subgroup discussed the evidence for aspirin, all of which came from the PEP trial and considered its lower cost compared to LMWH and fondaparinux. They concluded that it is very likely to be a cost-effective option in this population. However, the committee considered the PEP trial to show evidence of clinical effectiveness of aspirin as an add-on prophylaxis option rather than stand-alone, and its cost effectiveness should be considered in this context. Hence, the committee determined that the pharmacological options that could be recommended should be limited to LMWH and fondaparinux. However, the committee acknowledged the potential value for money that could be achieved if aspirin is proven to be effective as a stand-alone prophylaxis strategy. Hence, the committee made a research recommendation to assess the clinical and cost effectiveness of aspirin in this population.

There are 70,000 hip fractures a year in England, Wales and Northern Ireland (National Hip Fracture Database; http://www​.nhfd.co.uk/). This population is associated with older and frail people, with the mean age of patients being 82 years (http://www​.nhfd.co.uk/). Age is a significant risk factor for VTE and bleeding, thus it is important that prophylaxis is provided for these patients. There is an increasing trend to mobilise patients post-operation from day 0 in this population, which can reduce the risk of VTE.

There was a lengthy discussion about the lack of evidence evaluating DOACs in this review population. DOACs are currently licensed in the orthopaedic populations of elective hip replacement surgery and elective knee replacement surgery. The subgroup understood that the absence of evidence about these interventions in this review population prohibited a suggested recommendation but appreciated that there may be some clinical benefit and cost saving from these interventions.

The committee made a high-priority research recommendation on aspirin alone, and a research recommendation on IPCD, in this population group; see appendix R for more details.