Customised foot orthoses for Achilles tendinopathy RCT: responding to the critics

By Dr Shannon E Munteanu, Lisa A Scott, Daniel R Bonanno, Dr Karl B Landorf, Dr Tania Pizzari, Prof Jill Cook, Prof Hylton B Menz
-achillesOur group’s randomised controlled trial (RCT) that evaluated the effectiveness of customised foot orthoses for Achilles tendinopathy was recently published in the British Journal of Sports Medicine. The abstract can be viewed here, and the full text of the study protocol can be viewed here.
As with any RCT that provides unfavourable results for a commonly-used intervention, our trial has attracted some degree of criticism on blogs and social media. Rather than entering into ongoing debates across multiple platforms, we have instead summarised our responses to these criticisms below.
Criticism #1: The sham orthoses were not ‘biomechanically inert’
In order to measure the treatment-specific effectiveness of an intervention, controlled trials often use placebo groups to compare the experimental intervention to – the best example being the use of inert pills in pharmaceutical trials. However, this is not possible when evaluating interventions which have a physical or mechanical effect (i.e. those commonly used for musculoskeletal disorders). In these trials, the control group is provided with a ‘sham’ intervention, which can be defined as “a treatment or procedure that is performed as a control and that is similar to but omits a key therapeutic element of the treatment or procedure under investigation” (note the added emphasis). Put simply, a sham intervention is not a placebo.
Our trial published in the British Journal of Sports Medicine compared the effectiveness of customised orthoses to sham orthoses for people with Achilles tendinopathy undergoing an eccentric exercise program. The sham orthosis was a contoured, vacuum-moulded device constructed from 4.0 mm thick ethylene vinyl acetate (EVA) with a density of 90 kg/m3 (i.e. very soft) and had an identical covering fabric to the customised orthosis. The sham device was selected based on a previous study which showed that it produced only small effects on plantar pressure (10% reduction in peak pressure under the heel) while still being considered a credible intervention by participants.
After this study was published, the following comment was made in relation to the selection of control interventions in orthotic studies:
“…the best you can do is to minimise the influence that the control orthoses have upon the variable of interest” (link).
We agree. However, since the publication of our RCT, some have argued that the sham device was not actually a sham because it was not biomechanically ‘inert’, for example:
“There is no such thing as a “sham” foot orthosis, and the sooner this is recognised the better” (link)
“Not sham orthoses at all. This study actually compared two different types of orthoses” (link)
“Does this mean they are completely “inert” and will not change any kinetic parameters at the foot-orthosis interface?” (link)
These comments represent a misunderstanding of what a sham intervention is. It is obvious that no device placed in the shoe can be truly biomechanically inert. However, at no stage have we claimed that the sham device used in our trial was inert – we have simply argued (and have evidence to show) that they have as minimal effect as possible while still being considered a credible intervention. Credibility of the sham is critically important when evaluating a real intervention against a sham intervention, otherwise resentful demoralisation comes into effect in the sham group, which might bias the findings.
Interestingly, a previous RCT that evaluated the effectiveness of customised orthoses for pes cavus also used a sham device for the control group (an insole made from 3-mm latex foam). This device had similar effects on plantar pressure to our sham device (reductions in pressure-time integrals of 9%, 11% and 6% in the rearfoot, forefoot and midfoot, respectively). However, the response to this trial could not have been more different, and no criticism has been made of this sham device. The fact that this study found customised orthoses were more effective than sham devices may explain the different responses to these two trials.
If there are any lingering doubts that the sham orthoses had minimal mechanical effects and were markedly different to the customised orthoses, we provide the following movie for your perusal:
<iframe width=”420″ height=”315″ src=”//” frameborder=”0″ allowfullscreen></iframe>
Criticism #2: The customised orthoses were not appropriately prescribed
Researchers conducting trials of customised orthoses are faced with the unavoidable dilemma of how to individualise or ‘customise’ the orthoses. This is a long-standing problem and arises because there are currently no evidence-based or consensus guidelines in relation to the prescription of foot orthoses – an issue that is explored in this commentary paper. An ambitious attempt to try and achieve this commenced in July 2009 but to date has produced no outcome (and has been described as an ‘epic fail’ by one of its proponents).
In our trial, the customised orthoses were individually prescribed based on an assessment of participants’ Foot Posture Index and bodyweight. At the time the study protocol was published, there was some concern regarding the lack of a heel lift (this is discussed later), but our approach attracted the following positive comments:     
“I am excited to see Shannon, Karl, Hylton and company trying to tackle this important subject since Achilles tendinitis is a very common injury in my sports medicine practice and it would be nice to see how the individuals all respond to the different protocols” (link)
“…they have made the effort to make the devices more custom than any other study that I have seen…If they have a pronated FPI, they get the medial heel skive, if neutral it is a basic shell, and if a supinated FPI they get a device to control supination as described in Josh Burns paper…the devices are made at a laboratory from slipper casts, so should be close to what many Podiatrists would prescribe” (link)
However, on the publication of the RCT results, the mood changed considerably:
 “Is giving everyone the same prescription even a true definition of a custom device? Not for me…” (link)
“It’s really hard for me to get excited about custom foot orthosis research unless the custom foot orthoses used in these studies are the types of foot orthoses used by those who are most experienced and expert at foot orthosis therapy” (link)
To reiterate – there are no evidence-based or expert consensus guidelines in relation to the prescription of foot orthoses. Until such guidelines are produced, there is no basis upon which to argue that our prescription approach was inappropriate.
Criticism #3: Why no heel lift?
Heel lifts are commonly recommended as a treatment for Achilles tendinopathy, based on the assumption that elevating the heel decreases Achilles tendon loading. In designing the trial, we considered adding a heel lift to the orthoses but eventually decided against this. This was criticised when the study protocol was published and again when the RCT results were published:
“I would have thought that if one was to design an experiment on custom foot orthoses for Achilles tendinopathy that one would have used a custom foot orthosis with a heel lift added to the orthosis since this is the type of custom foot orthosis that nearly every good sports podiatrist that I know of uses in their orthoses for patients with Achilles tendinopathy” (link)
“Of course there should have been heel lifts on the foot orthoses” (link)
“Clinically speaking I think it’s reasonable to say that most practitioners would incorporate a heel raise” (link)
So why no heel lifts?
Put simply, the evidence to support the use of heel lifts in Achilles tendinopathy is extremely limited. First, our systematic review revealed no trials of heel lifts as a treatment for AT. Second, the biomechanical evidence that heel elevation decreases Achilles tendon loads is not at all convincing. At the time the study was designed, three studies had examined the biomechanical effects of heel elevation on Achilles tendon loading:
         Reinschmidt and Nigg (1995) found no difference in plantarflexion moments when subjects ran in shoes which differed in heel height (2.1-3.3 cm), and concluded that “the results of this study did not support the speculation that changes in heel height would reduce the plantarflexion moments of the ankle joint, and thus, the Achilles tendon forces…the treatment or prevention of Achilles tendonitis with a raising of the heel is based on anecdotal evidence and not on research”
         Dixon and Kerwin (1998) assessed the effects of 7.5mm and 15mm heel lifts on Achilles tendon force during running, and found that Achilles tendon forces increased with greater heel elevation. They concluded that “the finding that increased heel lift may increase maximum Achilles tendon force suggests that caution is advised in the routine use of this intervention”
         Farris et al (2008) assessed the effects of 12 and 18mm heel lifts compared to a no-lift control condition in 6 female runners, and found no difference in peak Achilles tendon forces or maximum Achilles tendon strain between conditions, concluding that “Heel lifts alter ankle mechanics during running. However, this appears not to affect peak AT force or strain”
In the time between the commencement of our trial and its publication, one study has been published that suggested that heel lifts may have beneficial effects on Achilles tendon loading:
         Farris et al (2012) assessed the effects of 12 and 18mm heel lifts on Achilles tendon strain in ten female runners, and found that, compared to the control condition, strain reduced (by 1.9%) with the 18mm lift but did not change with the 12mm lift.
Would our prescription protocol have changed if this study had been published prior to the commencement of our trial? No – on balance, the evidence is not overly supportive of the use of heel lifts in Achilles tendinopathy. Interestingly, a recent biomechanical study found that contoured orthoses without a heel lift reduced Achilles tendon load in 12 runners, so it cannot be argued that orthoses must have a heel lift in order to be effective in the treatment of Achilles tendinopathy. Furthermore, even if there was good evidence to suggest that heel lifts reduce Achilles tendon loading, the question then arises: why not simply use a heel lift rather than incorporating a heel lift into a customised orthosis?
In summary, our trial is the most rigorously designed study so far undertaken to assess the effectiveness of customised orthoses in the treatment of Achilles tendinopathy in individuals undergoing an eccentric exercise program. We stand by our conclusion that foot orthoses, prescribed according to the protocol in this study, are no more effective than sham foot orthoses for this condition. We welcome further high-quality RCTs evaluating foot orthoses by other investigators, particularly if they believe that they can prescribe more effective foot orthoses.
Originally posted:


(Visited 1,786 times, 1 visits today)