Evidence based medicine: is it working in tendinopathy?

11th of October 2017home / blog / tendinopathy-updates / evidence-based-medicine-is-it-working-in-tendinopathy

 

Hi all,

Apologies for my silence over the last couple months. I fell off the wagon as family and research took over. I missed one and then it was so easy to miss another and another. But the blog is back (subscribe here) and what I plan now is shorter pieces to showcase interesting new research and a longer blog every now and then.

Part of the reason for the absence is that I am very excited to announce that the Mastering Lower Limb Tendinopathy Course is now fully online and will be launched in the next few weeks. Please stay tuned.

Hope you enjoy the blog this week, an example of evidence based medicine, or perhaps lack off, in tendinopathy.

See you next time

Peter

  

Evidence based medicine: is it working in tendinopathy?

This will highlight an example of evidence based medicine and how we should reserve excitement in science before we have multiple studies and sufficient evidence.

Evidence based medicine includes three steps 1) knowing all the evidence; 2) tempering the evidence based on your experience and the patient in front of you; and 3) the patients values and preferences. See this excellent EBM blog by Erik Meira for a more detailed run down.

About 4 years ago I was involved in a preliminary paper by Comin et al 2013 showing a link between central tendon hamstring pathology on MRI and delayed return to sport. We examined 62 hamstring injuries in a retrospective study and found that the 9 biceps femoris injuries with central tendon involvement took a median 72 days versus 21 days for the ones without central tendon involvement. Pollock et al. 2015 showed something similar. Wow you could say, but of course these are small retrospective studies open to bias and confounders (so a valid question is why are they being published in top journals?). Has this evidence made an impact? It seems so if you look at google scholar citations below – cited by 65 within 4 years.

 Picture1blog.png

A leading hamstring group including Bruce Hamilton, and Rod Whitely (I’m only mentioning these two because I know them, not because they are antipodean), has just shown that maybe we should adjust our beliefs on this issue (see paper here Made et al 2017). Although they found central tendon involvement was associated with increased time to return to play, but only by a week, and to quote the authors…

‘the considerable overlap between groups with and without intramuscular tendon involvement substantially limits the clinical (ie, predictive) value of intramuscular tendon involvement’.

The figure below gives an excellent graphical representation of this. The last group (100% disruption) had a significantly greater time to return to play than the group without central tendon involvement. But clearly the overlap is huge.

pic2.png 

So where does that leave us in terms of evidence based medicine, which paper do you trust? Well of course, the answer is the recent paper. But not because it’s prospective, blinded clinicians making decisions to MRI findings, and had a larger cohort (well partly due to these reasons, but also there is another reason). Bayes theorem (consider new evidence in light of ALL existing evidence). A systematic review of 12 studies concluded that there is currently no strong evidence that MRI parameters are not predictive of time to return to play. So, our prior trust in MRI predicting return to play should have been low, therefore, we should require strong evidence (multiple high quality studies) to change this belief. This is shown on the spectrum of uncertainty below – in science we ascribe probabilities to beliefs or theories being true or false and this shifts as new evidence comes to light.

 pic3.png

So in the case of central tendon injuries predicting time to return to play we should have been cautious and followed the evidence. The same can be said for many other beliefs in tendinopathy, e.g. heavy isometric is the best intervention for tendon pain. Based on current evidence is this certainly true? Read here for some discussion of this.

Well done to the authors for a quality paper that advances the field. As a scientist, you must put your hand up when someone does it better and accept that is probably the direction of the truth, rather than hold onto beliefs as the evidence crumbles around you.

See you next time

Peter