How to stop a runaway train: Part 3

Updated: Jan 5, 2020

This is the final instalment in a series reflecting on my attendance at the Preventing Overdiagnosis Conference 2019 (PODC2019). Read part 1 here and part 2 here

There is no doubt that this ‘too much medicine’ train is off course in the musculoskeletal realm. No one is in control and this runaway train may even be speeding up. As our society improves in almost all other walks of life, we are getting sicker? We are becoming more painfully disabled year on year (Ref). In no uncertain terms, progress in stopping this train has been incredibly slow (Ref). We have been unable to muster much change from a key party in this conversation – the clinicians. Through this piece, I will discuss reasons why what we have tried hasn’t worked. And although this train seems completely lost with no chance of redemption, I will attempt suggestions to instil hope that we can get it back on track. 

Our runaway train has a complex web of factors driving it towards somewhere we never set out to be and adding to its inertia. We visited some systemic and patient factors in parts 1 and 2, these are mighty challenges. But solutions to this runaway train’s inertia thus far are akin to throwing boulders at the tracks and hoping for change. I will visit one type of boulder (of which there are more) in the form of Clinical Practice Guidelines (CPGs). It’s not that these boulders are a bad idea, it’s just that they empirically haven’t been successful for their intended purpose. So marked has this failure been, that there is now a dearth of work exploring it (Ref, Ref, Ref, Ref). These data reveal the extent to which CPGs can be misinterpreted and disagreed with in rich quotes like these (taken from systematic review by Slade et al 2016)

“ Some of it is a wee bit insulting.. It reduces medicine to… all these algorithms” Crawford (2007) “ They are setting the profession back 20 years” Parr (2014)

These are extreme quotes, but they do provide a proxy for the context in which CPG’s are supposed to work. CPGs can be defined as

evidence based statements that include recommendations intended to optimise patient care and assist health care practitioners to make decisions about appropriate health care for specific clinical circumstances.” (Ref).

Successful implementation of guidelines should reduce the gap between research and practice by expediting advances in every day practice and reducing inappropriate variation. Guidelines, then, are a tool for clinicians to use in situations where multiple therapies are available, or where uncertainty in terms of treatment options exists. They are ultimately one of the most important tools HCPs are endowed with and should promote all of our best rationalistic tendencies (Ref). Instead, they don’t seem to promote reason or science at all.  

Where is our reason?

This may be because CPGs represent a direct attack on something that every HCP is endowed with: an identity. Clinicians are fallible humans and reason goes out the window when our identity is involved. CPGs should be a neutral and innocent entity; they simply provide a reductionist solution to a complex problem. The outcome, however, couldn’t be further from neutral. CPGs cut deep into something primal; beyond the thoughts of ‘reducing my clinical practice to a recipe’, they uncover the embarrassed school child whose capacity to help people is suddenly threatened (Ref). It’s threatening to many other things, like livelihoods and money spent on extravagant courses, but mostly CPGs are threatening to what people do. Clinicians are largely defined by what they do and CPGs specific to low back pain specifically identify and set limits around these modalities (Ref). This would be akin to telling a leopard not to have so many spots. What I’m not saying is that leopards shouldn’t be told this, simply that it’s hard telling a leopard to have less spots; you should try it. Clinicians can also be cruelly tribal amongst each other, bitterly divided by these CPGs amongst other issues. These issues, like the manual therapy versus no manual therapy debate, are not trivial by any means as they deeply impact the outcomes of patients. They do, however, quickly descend into mud slinging, and tribal psychology may be at the heart of this. Evidence based practice is part of the enlightenment of medicine, but this levity above our basic human instincts has yet to extend into cross-HCP communication. The number of issues HCPs get tribal about is mind boggling; is another ‘us and them’ scenario really the thing we need? 

Another well-established ‘us and them’ scenario is the gap between research and practice. The ‘know-do’ gap is a complex and dynamic place, one which has spawned a flourishing scientific practice (implementation science). A gap, which no CPG, no matter how nuanced in its language, could effectively close. The epistemology involved in implementing guideline based care by orthopaedic surgeons has been robustly studied by Amy Grove and colleagues across three hospitals in the UK (under the NHS).  When discussing knowledge translation, a first salient question needs to be, where does the knowledge come from? Made clear from her analysis, it can come from many places indeed.

Once again, simple low hanging fruit in the ‘low value care’ discussion might be ensuring practitioners are aware of the CPG. But following this mammoth task, you still have a complex array of knowledge, from which the practitioner can weigh up and choose from when making decisions about care. One has to ask; is this even about knowledge? If it is, what knowledge do we value above others? Once again, human psychology could be at the heart of the matter. More fantastic thematic analysis by Amy Grove is presented by the below hierarchy.

Moving past the individual level of implementation, we get to groups. It has long been known that people who belong to a group will vehemently defend a concept that is demonstrably untrue (see Flat Earth Theory for background) because refuting it will mean a painful ostracism. When we belong to groups, our knowledge can be occluded from view because something much more powerful takes over: beliefs. Beliefs are richly entwined in our identities, more linked to our need to belong than knowledge. Certainly, the social construct in which decisions take place seems to be more predictive of what kind of care people receive, not, as would seem intuitive – the specific provider (Ref, Ref). Groups are incredibly important and much work has been done into finding out what clinicians state as barriers, but we have only just begun to scrape the surface of uncovering clinician beliefs and how to change these surrounding CPG implementation (Ref).     

A key clue to changing behaviour around evidence-based, clinical guideline based care is provided by this snap taken during Teppo Jarvinen’s keynote at PODC2019. It’s called GOBSAT (Good Old Boys Sitting Around a Table).

Involve me and I learn

On a more serious note, the problem of changing clinician behaviour around guideline based care is a question which continues to perplex even the best researchers. There is a derth of literature looking into what successful CPG implementation strategies should look like (Ref, Ref). Thankfully CPGs aren’t simply parachuted in from a height in isolation (tell me and I forget) and gone are the days of traditional guideline education strategies (tell me and I forget) (Ref). It is clear that multifaceted strategies are more successful than single component ones, but we have no real idea about what components may be most effective (Ref). There is high level demand from clinicians to be actively involved in implementing guideline based care (Ref). One reason for this might be that each strategy has to be both tailored to the context in which care decisions are made (including multiple levels of health service, hospital, ward, team) and tailored to the individual that makes them. This level of complexity is hard to garner an appreciation of, unless you’re there on the ground making decisions every day. Indeed, the peer review and auditing systems of implementation show promise (Ref).

A great example of this was presented by Adam Elshaug during PODC2019. His work with the Medicare Benefit Scheme Review Taskforce and Australian orthopaedic surgeons is an interesting take on this issue. He produced ‘dashboards’ available to all surgeons operating in a particular health district. Dashboards are populated with all of the ‘billable items’ a surgeon can perform (different surgeries), stratified by hospital. This notifies surgeons, in real time, about their volume of specific surgeries, which provides some objective data into performance. Although this may be at risk of adding more ‘noise’ to a chaotic decision-making framework, it does another crucial thing. Besides, most clinicians get used to using a computer system in this way (Ref). These dashboards also include peer data, which creates an altogether different effect. 

How does behaviour change? Slowly, incrementally, probably randomly. But it seems that groups, peers and tribes are important.

Do allied health HCPs need this kind of system? Some may see this as another invasion into the clinical sanctity, but Adam’s results seem promising so far (unpublished data presented at conference) and if surgeons can change behaviour then… Well you know the rest of the joke. I think allied health HCPs have an inertia all their own when it comes to the low value care overtreatment conversation. Typically given a ‘bye ball’ because no one dies from too much of ‘our medicine’. But this is a moral hazard we need to shake, our patients’ deserve better.  


Another take on the wicked problem of social groups and behaviour change is depicted with the quote by Max Planck (also presented by Teppo Jarvinen at the conference), 

“ A scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die and a new generation grows up that is familiar with it.”

Rather than revert to nihilism about the current predicament we find ourselves in, we can instead take on this challenge with open arms. I would be the first to admit the preventing overdiagnosis and overtreatment conversation can tend to a very paternal and elitist one and this can add to the problem, but it doesn’t actually have to be this way. That’s why I’ve written this blog, that’s why countless people toil away getting information to where it needs to be. This needs to be a discussion that everyone is involved in. If we embrace our tribal psychology rather than ignore it potentially we don’t have to wait until the previous generation scuttle off this mortal coil.

Having trust in the slow, steady pace of change for the better is a start. Trusting in the next generation of clinicians to make better decisions is a necessary stance; because they inevitably will. The new generation of clinicians are already better at providing high value care and more receptive to guidelines (Ref). Stopping this runaway train, may require more boulders, or something more subtle. Whatever it is, it needs to involve everyone. We do need health system change, but after this happens, which it inevitably will, it just leaves us. In the words of Hans Rosling, 

“I’m not an optimist, I’m a very serious possibilist.’

So, how do you stop a runaway train?

38 views0 comments

Recent Posts

See All