Introduction

Background

The Achilles tendon is the largest tendon in the human body and transmits the powerful contractions of the calf muscles that are required for walking and running. A rupture of this tendon is painful and has an immediate and serious detrimental impact on daily activities of living.1 In the longer term, tendon rupture results in prolonged periods off work and time away from sporting activity (average time away from work is between 4 and 8 weeks and time away from sport is between 26 and 39 weeks).1 This results in lost income and restricted daily activities in the early phase and reduced physical activity, with associated negative health and social consequences, in the long term. For high-level sportsmen it is frequently a ‘career-ending’ injury.

Achilles tendon rupture affects > 11,000 people each year in the UK, and the incidence is increasing as the population remains more active into older age.2 It affects all age groups in a bimodal distribution, with the first peak in patients aged 30–40 years and the second in patients aged 60–80 years.2 The first peak in incidence is often associated with participation in sport, such as football and racquet sports, whereas the second peak often occurs during normal daily activities, such as climbing stairs.2,3 However, all Achilles tendon ruptures are associated with a pre-existing ‘tendinopathy’, which is attributed to failures in the protective/regenerative functions that respond to repeated microscopic injury.4,5

Historically, the main question in relation to the management of patients with rupture of the Achilles tendon has been whether or not to perform a surgical repair of the tendon. In 1981, Nistor6 designed and published the first randomised controlled trial (RCT) to address this clinical question. This study was followed by a series of RCTs that were pooled in a meta-analysis by the Cochrane review group in 2004.7 The results suggested that surgical repair reduced the risk of re-rupture, but this came with an increased cost and a greatly increased risk of other complications, most of which were associated with infection and wound healing. There were few data on functional outcome at the time of this review. More recent trials comparing surgical repair and non-operative treatment have found no difference in functional outcome.8,9 As surgery carries considerable costs, and carries considerable risks to the patient in terms of complications,7 there is an increasing trend towards non-operative treatment. However, some surgeons have been reluctant to advocate non-operative treatment because of concerns about the lack of evidence to guide early rehabilitation for this group of patients,10 specifically whether or not functional bracing is safe and effective if the tendon has not been not surgically repaired.

Traditionally, patients have been treated in plaster casts after rupture of the Achilles tendon, with the cast immobilising the foot and ankle while the tendon heals.11 However, there are potential problems with this approach. First, there is the immediate impact on mobility for a period of around 8 weeks, affecting activities of daily life. Second, there are the complications and risks associated with prolonged immobilisation: muscle atrophy, deep-vein thrombosis (DVT) and joint stiffness.12,13 Finally, there are the potential long-term consequences, which include prolonged gait abnormalities, persistent calf muscle weakness and an inability to return to previous activity levels.14 Functional bracing, involving immediate, protected weight-bearing in a brace, was designed to address these issues.

In patients having a surgical repair, seven RCTs,1,15–20 directly comparing plaster casts with early movement and/or weight-bearing in a ‘functional brace’, had been conducted at the time that the protocol was developed for the UK Study of Tendo Achilles Rehabilitation (UKSTAR). The results favoured functional bracing in terms of re-rupture rate, functional outcome and quality-of-life (QoL) measures. Therefore, in the first guideline produced on this topic in 2009,21 the American Academy of Orthopaedic Surgeons recommended functional bracing for patients undergoing surgical repair of their tendon.

What about patients managed non-operatively?

Although there are clear guidelines for rehabilitation of patients who have a surgical repair, there is no clarity about the use of functional bracing in non-operatively managed patients. Does functional bracing provide improved function and QoL if the tendon is not surgically repaired? Or, in the context of a tendon that has not been stitched together, does a plaster cast provide greater protection and therefore improved healing? Does functional bracing facilitate faster return to work and is this cost-effective? Or, is the tendon more vulnerable to re-rupture in a brace, with the subsequent risk and cost of reconstructive surgery?

At the time that UKSTAR was developed, we supplemented the 2004 Cochrane review7 with an updated literature search and found that in total only two additional studies22,23 had been performed that compared functional bracing with plaster cast for patients managed non-operatively following rupture of the Achilles tendon. Both studies suggested potential benefits from bracing. However, the data from the studies should be interpreted with caution because patient numbers were small (90 in total), patients received different functional bracing regimes and the reporting of outcomes was minimal.

This gap in the evidence was recognised by the American Academy of Orthopaedic Surgeons in their 2009 guideline,21 which stated that they were unable to make a recommendation with regards the use of immediate functional bracing. With the incidence of Achilles tendon rupture on the rise, and in the light of the high associated personal and societal cost, this evidence gap is a clear priority. A Versus Arthritis (formerly known as Arthritis Research UK) multidisciplinary ‘Think Tank’ (Arthritis Research UK, Birmingham, 2013) on tendon injuries reported that rehabilitation following non-operative treatment of acute Achilles tendon rupture was ‘the top research priority’ in this area.

Since the start of the UKSTAR, a number of small randomised trials have investigated both the mechanistic and the functional effects of early weight-bearing in a brace compared with cast immobilisation. A trial of 56 patients indicated that tendon healing at a molecular level may be enhanced by early mobilisation, but, given the small number of participants, there was no difference in objective functional outcome (heel raise testing in this study).24 A second trial investigated the biomechanical properties of the healing tendon in patients randomised to early weight-bearing or delayed weight-bearing.25 The investigators noted that there was less tendon stiffness in the group treated with early weight-bearing. However, in terms of functional outcomes, the authors reported no evidence of a difference in Achilles Tendon Rupture Score (ATRS), although they did report a statistically significant improvement in health-related quality of life (HRQoL) at 1 year in the group treated with early weight-bearing.25 Another trial included 47 patients treated non-operatively for an acute Achilles tendon rupture. Half of the patients were treated with partial weight-bearing beginning on the first day of treatment and the other half were treated with non-weight-bearing for the first 4 weeks.26 The authors concluded that early weight-bearing was ‘safe’ in terms of the incidence of re-rupture, but there was no evidence of a difference in functional outcome (ATRS or Physical Activity Scale) in the first 12 months after the rupture. Finally, another trial compared two types of cast immobilisation of Achilles tendon rupture.27 Half of the patients wore a traditional cast, which restricted weight-bearing, whereas the other group wore a modified cast, which included a heel ‘iron’ to facilitate weight-bearing. The authors found no evidence of a difference in functional outcome (Leppilahti Score), but there were only 84 patients in the trial. One further study, published very recently, randomised patients to cast immobilisation or ‘early controlled motion’ and involved 130 patients at a single centre. The authors found no evidence that early controlled motion was of benefit compared with immobilisation in any of the investigated outcomes.28

Pre-pilot data

Before UKSTAR, we completed four phases of pilot and preparatory work to establish the following.

External pilot study

We randomised 48 patients receiving non-operative treatment for acute rupture of the Achilles tendon to either functional bracing or plaster cast. This trial1 showed that patients and clinicians had equipoise for this question and were happy to take part. However, the trial identified that although plaster casting was a mature intervention, the important facets of the complex intervention, namely functional bracing, were inadequately defined, and that this needed to be addressed before a larger trial was undertaken.

Defining the functional brace intervention

In keeping with the Medical Research Council framework for developing complex interventions, our group and collaborators performed a UK survey of current practice, a systematic review of published rehabilitation methods, gait analysis experiments using different functional brace and heel wedge combinations, and qualitative interviews to define the optimal functional bracing regime and refine the trial design.29,30 The rehabilitation strategy proposed in UKSTAR was the summation of that work that identified the optimal type of orthosis (brace), the optimal foot position within the orthosis and the duration of application of the orthosis.

To investigate the number of patients potentially eligible for UKSTAR, we carried out a UK-wide survey of orthopaedic trauma clinicians.10 This clearly showed that clinicians were enthusiastic about the study and that the number of eligible patients was large enough for a full trial.

Research objectives

The primary objective was to quantify and draw inferences about observed differences in ATRS between the trial treatment groups at 9 months post injury.

The secondary objectives were to:

• quantify and draw inferences about observed differences in ATRS between the trial treatment groups at 8 weeks and 3 and 6 months post injury
• identify any differences in HRQoL between the trial treatment groups in the first 9 months post injury
• determine the complication rate of the trial treatment groups in the first 9 months post injury
• investigate, using appropriate statistical and economic analytical methods, the resource use, costs and comparative cost-effectiveness of the trial treatment groups in the first 9 months post injury.

Patient and public involvement

We have been working with and listening to the views of patients with Achilles tendon injuries for many years. However, as well as this informal contribution, a series of formal qualitative interviews with patients and clinicians were carried out in the development of the UKSTAR (ISRCTN68273773).11 The views of patients were used to inform and refine the trial interventions and processes, in particular the development of the trial information and materials. The patient perspective was key during the development of the trail protocol to ensure that the interventions, and participation in the trial, would be acceptable.

Two of the patients who contributed to our development work agreed to act as lay representatives on the Trial Management Group (TMG) and as co-applicants on the research grant award. Mrs Richmond later had to leave the research team for personal reasons, but Mr Grant attended TMG meetings throughout the trial and contributed to all trial process and paperwork, with particular input on patient information leaflets. Mr Grant will be crucially involved in the dissemination of the study findings to the wider public. He will lead the development of any materials, leaflets and website information to be used for this purpose. Mr Grant has reviewed the Plain English summary of this report.

Mr Grant was supported by the chief investigator and the trial co-ordination team. He had peer support from the UK musculoskeletal trauma patient and public involvement (PPI) group, hosted in Oxford. He also had access to and support from the University/User Teaching and Research Action Partnership network through University of Warwick, an organisation that promotes the engagement and involvement of service users and carers from the local community in research and teaching in health and social care.