Down the rabbit hole or jumping over it?
What researchers see as important is different from what clinicians need
– Written by Jill Cook, Australia
Alice in Wonderland fell down the rabbit hole, at least that is what I am told. I never did like that book. The analogy is perfect though for much research, where pursuing the answer to a single question (and its related parts) might seem like a good idea. It is also appropriate for some aspects of molecular research aimed at discovering single-cause reasons for disease (for example tyrosine kinase inhibitors for chronic myeloid leukaemia)1.
Clinical musculoskeletal (MSK) conditions are not like that, they have a multifactorial aetiology complicated by individual factors such as genetics and co-morbidities and it is unlikely that pursuing any one question about pathology will impact on the condition across the population. So resolving the minutia is not only unlikely to inform the clinician about the condition, but more importantly, will not help them manage the condition in the complex body in which it exists. Clinicians need to stay above the ground and out of rabbit holes, taking key messages from each rabbit hole and integrating the research into a management strategy for MSK pain.
What is most critical is that pain (the key clinical musculoskeletal domain and discussed here in a clinical context, not from a pain science perspective) is poorly related to pathology and structural change (although it remains an important risk factor for pain). Often, the primary (misguided) aim for both researchers and clinicians is to restore normalcy, most commonly thought of from a structural or pathoanatomical perspective. We want to measure everything to within an inch of its life and then ascribe benefit to restoring normal levels. But most MSK pain is poorly related to pathology, making the pursuit of normal structure and absence of pathology senseless.
If we only consider MSK pain, normalcy (for the patient) is the absence of pain. But MSK pain can be ameliorated by reducing use – impossible to do in many organic diseases but certainly a common response in MSK conditions. Reducing use may be beneficial for MSK pain, but it comes with a range of negative health consequences – because of physical inactivity – that are likely worse in the long-term than living with the pain (for example increased risk of non-communicable disease). Clinicians are paid to manage MSK pain but the aim should be to do that in the context of full function for each person. Reducing use is likely to negatively impact on the strength and capacity of tissues and therefore will affect function, further reducing ability to exercise and quite possibly reducing the threshold for pain. All this is considered without accounting for the central contributions to the pain experience.
So where does research sit in this complex clinical environment? There are those who preach that most research provides meaningless findings2 with small effect sizes3, and many clinicians would agree that most is not helpful for their patients and practice. What seems apparent is that much research pursues single ideas to death (down the rabbit hole), whereas clinicians need a broader perspective and helpful evidence-based ideas for their patients. If research is not helpful, clinicians are vulnerable to falling down clinical rabbit holes, where complex conditions are managed with simplistic and unimodal treatments.
Let’s examine one condition – tendinopathy – to illustrate that there are innumerable rabbit holes in research and clinical practice. These have direct implications (not always positive, in fact mostly negative) for clinical management.
COMMON RABBIT HOLES IN TENDON RESEARCH AND THEIR COROLLARIES
There are 28 matrix metalloproteinases (and several inhibitors of them) that are responsible for the turnover of the tendon matrix4. There are also 36 interleukins, all active across a range of tissue responses to disease and injury5. It is crazy to think that fixing levels/expressions/half-life of one will immediately have a positive effect on the tissue as a whole and, of course, the complex individual with the condition. Of course in this rabbit hole, it is also essential that there are no interactions between any of these molecular substances.
The clinical corollary of this is injecting tendons with these magical mixtures and expecting that not only will it normalise structure, but that in turn it will fix pain and therefore will return the tendon/kinetic chain/person to full function. We have spent a lifetime looking for simple cures for complex problems, it is time to let it go and move on to flexible, multimodal research that allows us to investigate MSK conditions in a real world environment.
We can image pain
I was brought up on the pathoanatomical model and it is so hard to let go of the relationship between structural pathology and pain. Surgeons may struggle more than most as their treatments are directed at ‘normalising’ structure. Interestingly, the injured population is also wedded to this idea, having been exposed to multiple scans and reports of their abnormal structure and struggle with the concept that function and pain can be corrected without structural change.
We also have to be so careful of using pejorative words in radiology reports; tears and degeneration imply poor outcomes and have been shown to influence response to treatment6. More and better imaging is unlikely to change this, any imaging that may shed light on local drivers of pain will be invasive and expensive and are unlikely to become feasible in the near future. In tendons, where the local nociceptive driver is not known and potentially cell-based, this is even less likely to be possible regardless of advances in imaging.
The corollary of nociception is that the brain is responsible for the generation of pain without a tissue-based nociceptive driver. This is currently fashionable (another bandwagon), but loses sight of the concept that a key driver of MSK pain is tissue-based nociception. In tendon, athletes can have 20 years of patellar tendon pain, that hurts when they jump (and only then) and for a short (tissue) time after, making the likelihood of a nociceptive driver, although not yet found, overwhelming. Ignoring the nociceptive driver and only addressing the central modulation is unlikely to have a positive long-term effect.
The rabbit response is exactly the same as the human response
Research in humans is complex and research in complex conditions in humans is impossible, especially if the comparison is not drug A to drug B. It is easy to control animal experiments perfectly and perhaps the military model (where many factors that might influence outcomes are more controlled than in general population research) is the best in humans.
Even in animal models the study design may not reflect human disease. The Backman model of inducing tendon pathology has been used extensively to investigate tissue responses7. The model is a repetitive plantar and dorsiflexion movement in an anaesthetised rabbit, so does not involve any energy storage and induces peritendinous inflammation rather than overuse tendon pathology we see in humans.
PUTTING THE BUNNIES IN WITH THE HARES AND GUINEA PIGS
It may be surprising to some that there is limited agreement on the diagnostic criteria for tendinopathy. In 2007, it was first recognised that set research diagnostic criteria were needed, and that is also true for clinicians. In tendons, if imaging and palpation tenderness are not always associated with tendon pain, then these criteria are not sufficient to diagnose tendinopathy and include participants in a study. Today, we still have no consensus on inclusion criteria of participants in studies. If there is no agreement on the inclusion criteria – how can the researcher know if the findings reflect a response of the true condition to the treatment? Following on from this, how can the clinician know the treatment is applicable to the person in front of him or her?
WHAT DRIVES RESEARCHERS DOWN RABBIT HOLES?
Pragmatic clinical research that allows for individual choices and clinical prescriptions based on presentation is not considered worthy of publication because of inherent biases in the study design. This results in effective interventions in trials that lack effectiveness in the real world.
Grant and publication imperatives
Simple questions with clear answers get funded and published, or at least get through the review process without a mauling. Introduce biases (such as co-morbidities in a real world population) then the numbers necessary for a research study would have to be huge to satisfactorily account for them.
It is easy to follow on from previous research
Question A leads to question B which leads to question C, and so on. The researcher does not care if it is not helping the clinician, it is being funded and published; end of story.
WHAT ABOUT CLINICAL RABBIT HOLES? (OR IS THAT A BANDWAGON?)
Applying a single, simplistic treatment is an easy response to the complexity of clinical practice and the research results that are not easily translated into meaningful clinical practice. There is also an imperative to use the newest and most exciting treatment – what I would call a bandwagon. These treatments become ‘hot’ in the media, the demand for them skyrockets and too often the research community is stranded years behind. By the time research catches up, that bandwagon has gone by and a new one is passing for clinicians to jump on.
Many bandwagons have come (and mostly gone) in my time as a MSK clinician: micro-current, hyperbaric chambers, isokinetic machines, injections (pick your poison) and most electrotherapeutic modalities. What has not gone, (or been fully embraced by many clinicians) is the use of structured exercise as a complete management – not just an adjunct to management after some magic intervention. This doesn’t mean applying a recipe. Clinicians jump on the latest bandwagon instead of treating the person in front of them, preferring to use a standard treatment for the condition or disease, rather than assessing and treating the individual with their nuances.
Recipe that is perfect for everyone
It is really simple: disease X needs one treatment Y that is effective across the population. Low back pain research is attempting to sub-group populations so that treatment can be directed at the most relevant structure/pain/function. For tendons, the past few decades have focused on eccentric exercise programmes applied to every patient with tendon pain. In some tendon populations, where tendon pain is seen in a homogeneous group (patellar tendon pain occurs, nearly all the time, in young men doing jumping activities), that approach may be appropriate. Achilles tendon pain is a condition across the activity and temporal lifespan, affecting the young sprinting athlete and the older sedentary person. There is no way that a single recipe treatment can effectively manage this condition in these varied groups – recipes do not account for co-morbidities, strength, activity and function.
The backbone is/is not connected to the neck bone
Treating things in isolation is rarely effective – even a simple ankle sprain results in ligament, muscle, sensory perception and brain changes8. Tendon pain also causes widespread changes that requires exercise that affects the kinetic chain and the brain9. Poking the sore tendon and directing (obsolete) treatments at it might give short-term relief but does not affect outcomes in the long term.
The corollary is treating the finger to fix the toe. Core stability appears to have the ability to fix all that ails us in the MSK system, regardless of clinical assessment findings and individual factors. It would be nice to pay the relevant attention to the injured structure, the central manifestations and the functional deficits in the right amount and get the balance right.
NASTY RABBIT HOLES FOR BOTH RESEARCHERS AND CLINICIANS
Who cares if a treatment does not have evidence of its efficacy? If it can be charged for (and the higher price the better), then it needs promoting through channels other than academic science, such as websites and social media. Testimonials, especially from elite athletes or important people, count the same as (or more than) any well-conducted randomised trial. This, in turn, drives interventions where science is falling behind clinical use or more commonly where clinical proponents do not care about science as the economics are more compelling.
THREE SOLUTIONS TO THE DISCONNECT BETWEEN CLINICIANS AND RESEARCHERS
Improve imperfect links between researchers and clinicians
Communication and collaboration between these groups, as well as people who have a foot in both camps are essential. It is likely the science-based research where a clinical link is not clear or at least not imminent, is where most rabbit holes likely occur. Researchers should have a clinical hat on at least some of the time. This is easy for clinical researchers, but harder for basic science researchers. However, exposure to the complex clinical presentation is often what is absent in basic science research and perhaps we should encourage more interaction here.
There are few ways around this impasse that will be acceptable to the academic industry. We then have to accept that research is not having the impact on clinical practice we would like. It takes 17 years for new research to be adopted into clinical practice10 and that is despite much money being poured into translation research.
Clinicians have a responsibility to utilise research in clinical practice. This involves knowing the research, but at the same time jumping over research rabbit holes – staying above the ground, so they can relate complex research to the patient in front of them.
Research has specific drivers and it is not always what is going to help the clinician. Researchers in clinical medicine should have a perspective on clinical practice. Similarly, clinicians need to determine the direction of research. Only when both aspects align, will we have meaningful clinical research.
Jill Cook Ph.D., Grad. Cert. Higher Ed., PG.Dip.Manips, B.App.Sci (Phty)
Deputy Director, Australian Centre for Research in Injury in Sport and its Prevention
Sport and Exercise Medicine Research Centre
La Trobe University
Hughes T, Deininger M, Hochhaus A, Branford S, Radich J, Kaeda J et al. Monitoring CML patients responding to treatment with tyrosine kinase inhibitors: review and recommendations for harmonizing current methodology for detecting BCR-ABL transcripts and kinase domain mutations and for expressing results. Blood 2006; 108:28-37.
Ioannidis JP. Why most published research findings are false. PLoS Med 2005;2:e124.
Walsh M, Srinathan SK, McAuley DF, Mrkobrada M, Levine O, Ribic C et al. The statistical significance of randomized controlled trial results is frequently fragile: a case for a Fragility Index. J Clin Epidemiol 2014; 67:622-628.
Davis ME, Gumucio JP, Sugg KB, Bedi A, Mendias CL. MMP inhibition as a potential method to augment the healing of skeletal muscle and tendon extracellular matrix. J Appl Physiol 2013; 115:884-891.
Uchida H, Tohyama H, Nagashima K, Ohba Y, Matsumoto H, Toyama Y et al. Stress deprivation simultaneously induces over-expression of interleukin-1beta, tumor necrosis factor-alpha, and transforming growth factor-beta in fibroblasts and mechanical deterioration of the tissue in the patellar tendon. J Biomech 2005; 38:791-798.
Gillan MG, Gilbert FJ, Andrew JE, Grant AM, Wardlaw D, Valentine NW et al. Influence of imaging on clinical decision making in the treatment of lower back pain. Radiology 2001; 220:393-399.
Backman C, Boquist L, Fridén J, Lorentzon R, Toolanen G. Chronic achilles paratenonitis with tendinosis: an experimental model in the rabbit. J Orthop Res 1990; 8:541-547.
Bullock-Saxton JE. Local sensation changes and altered hip muscle function following severe ankle sprain. Phys Ther 1994; 74:17-28.
Rio E, Kidgell D, Purdam C, Gaida J, Moseley GL, Pearce AJ et al. Isometric exercise induces analgesia and reduces inhibition in patellar tendinopathy. Br J Sports Med 2015; 49:1277-1283.
Grol R, Grimshaw J. From best evidence to best practice: effective implementation of change in patients' care. Lancet 2003; 362:1225-1230.