This is part 2 of a reflection on the Preventing Overdiagnosis Conference 2019 (PODC 2019). Read part 1 here.
Overdiagnosis is closely related to overtreatment. So much so, that the two often become conflated and confused. Overdiagnosis can often lead to overtreatment but not always, and overtreatment can occur without overdiagnosis (Ref). So as not to continue the confusion, in this second part of my reflection, I will refer to overtreatment.
Overtreatment in the musculoskeletal realm is a much more complex issue than many can appreciate. It is indeed the rest of the train – the bulk of it. It is unruly, cumbersome and just so happens to be where the majority of people spend their time. This part of the train holds the passengers – patients and clinicians alike. The issue with this train is that it has completely lost its way, incapable from deviating from a path so removed from the one it may have originally set off on. This runaway train may look like it’s on-track from the outside, but when you’re on the inside it’s terrifying.
It’s a serious problem
The musculoskeletal realm, as far as the media goes, is a nuisance not a real problem because most people don’t die from low back pain or arthritis. Except when people do die. Like they are dying in one of the more terrifying trends of recent times. The opioid epidemic is well known in the United States for killing more people than HIV, car crashes or gun violence (Ref). And it’s coming to Australia (Ref). The Australian problem is very much real and it continues to march towards us unabated (Ref). And it’s strange how little coverage this actually gets.
Guidelines specify that for chronic low back pain,
“Clinicians should only consider opioids as an option in patients who have failed the aforementioned treatments and only if the potential benefits outweigh the risks for individual patients and after a discussion of known risks and realistic benefits with patients.” Qaseem, A., Wilt, T.J., McLean, R., Forciea, M.A (2017)
Choosing Wisely has paved the way with opioid education programs, for both practitioners and patients (Ref, Ref). A researcher who has been a maverick in this space is Emily Karanges from Sydney. She presented some amazing work at the PODC 2019. Unlike the imaging issue, the opioid epidemic represents low hanging fruit; it’s obviously a case of ‘too much medicine’ and the lines in the sand are clear. We need to act fast to avoid more deaths.
It’s a wicked problem
If only winding back too much medicine in the rest of musculoskeletal care was low hanging fruit. The rest of the train is filled with a complex web of what may be termed ‘low value care’. Value as it pertains to healthcare, continues to be an incredibly hard thing to define; it is often discussed but poorly understood (Ref, Ref). Broadly, value-based care means care where the clinical benefits outweigh the risks and the cost (Ref). It is dynamic and mutable. Just as in any other field, value is defined around the customer. Therefore in healthcare, value is defined as patient health outcomes achieved relative to the costs of care. It is value for the patient that is the central goal, not value for other actors per se (Ref). Value is an inherently loaded term; not only does this present an ideological attack on the sanctity of patient-clinician interaction but also confronts larger socio-cultural ideas (Ref, Ref). If value is defined by the “customer”, shouldn’t we let the free-market decide what this proves to be? What if you think we have patients not ‘customers’ and this whole discussion smacks of neoliberal progression? If universal healthcare is something you believe in, then should we be much more punitive around the types of care available? Importantly, practitioners may wonder, ‘If I’m getting people better, doesn’t that mean I am providing value to them?’ This much maligned term creates confusion; confusion creates inertia; and inertia keeps this train rolling. The problem of low value overtreatment is a really wicked problem to solve.
Dr Rachelle Buchbinder presenting a list of factors in this wicked problem. Hint: it’s more wicked than this.
Patients on the train
It is a key fact that patient expectations are a central tenet to the confusion around the value discussion. It is clear that patient expectations are an extremely powerful predictor of treatment success; however, this presents yet another quagmire (Ref, Ref, Ref). A common misconception is that if you don’t do what the patient wants, then it will violate their expectations and the outcome will suffer, thus reducing the perceived value in care. But this has yet to be borne out by any empirical evidence. In fact, there may be nothing wrong with expectancy violation, as long as it is contextualised (Ref). And although it is clear that patients want a diagnosis, they don’t seem to have similar feelings about treatment (Ref, Ref).
“At such moments, talking about the moment of diagnosis, doubt is being smothered and certainty is manufactured”
This desire for certainty could be adding the problem of overdiagnosis, but it may not be adding to the problem of overtreatment. Patients do expect healthcare professionals (HCPs) to do something but these wants and desires don’t seem to extend to specifics for the most part (Ref). With the utmost respect, what we could be seeing here is a case of classical conditioning. Just as Pavlov’s dog expects food as the bell rings, patients may expect certain types of care because it is all they’ve ever received. So although patient expectations may add to the confusion around the value discussion, it’s probably not in the way commonly espoused. All the same though, the inertia increases.
I won’t pretend that in this age of insta-garbage and meta-misinformation patient expectations aren’t impacted upon by other, external factors. These Google and Facebook-led expectations add to one of the most important jobs of healthcare; to educate and inform patients of what can help and what can harm them. Indeed, this is one of the most problematic issues of the modern age. The pursuit of an appropriately informed and critically thinking patient has engendered many responses; one of them is shared decision making. Shared decision making is obviously a worthwhile endeavour, as patients not only should be involved, but in fact really want to be involved in their care (Ref). To have patients at the centre of information exchange really ought to be standard practice, but progress in this arena has been slow (Ref). It seems the ‘real-life’ efficacy of this process is hard to discern (Ref, Ref).
As part of the shared decision making process, decision making aids have proliferated throughout the health landscape. Producing these aids involves a lot of time and effort but also can involve unwanted and unscrupulous actors. There is now a copious amount of online decision making aids produced by pharmaceutical companies for syndromes you probably didn’t know you had (until you take the test) for which, conveniently, they will also sell you a remedy. It’s not only the obviously conflicted cases of decision making aids that are fallible. Legitimate, well produced aids can involve conflicts of interests beyond financial gain. Rachel Thompson from the University of Sydney is pioneering some fantastic work into rigorously evaluating these tools (Ref). Again this doesn’t take away from their merit, it simply exemplifies what a wicked problem we seem to have. What is clear is that a simple aid is not a magic bullet in undoing ‘too much medicine’ (Ref). The decision about what kind of treatment to receive is a dynamic and complex process. It is easy to talk about shared decision making, but harder to do and even harder to see what works, that’s if it works at all. It may be that in aligning the issues of value and patient centred care, what we may need to do is re-work the very fabric of our healthcare reality. A reality in which the train is going in a very different direction.
The tracks are set
Because the direction we are heading isn’t working and the tracks of the train are set. It keeps patients and clinicians from alighting (if they wanted to). We have a funding system that incentivises clinicians to do more at every level. The activity based funding (ABF) model is a proposed response to increased cost and accessibility of healthcare; the free market should improve efficiency, as it’s supposed to in other realms. It is now the dominant healthcare funding model locally (in Australia) and internationally. Although there are many guises to this model (synonymous terms in policy include – ironically – patient-centred funding and performance based funding), its primary function is to reward activity above all. A large systematic review of ABF models from OECD 10 countries demonstrate that when this experiment has been going on the longest it actually results in more episodes of care over time (Ref).
While this system may manage simple cases well, it completely mismanages complex patients (Ref, Ref). And it just so happens that complex health needs create the majority of healthcare use (Ref). The data show that people with musculoskeletal conditions are more likely to have co-morbid conditions, so the population we primarily deal with are very much included in this problem (Ref). One has to ask, has this experiment worked? If we want customers for life, then this could be a positive or a negative depending on your vantage. There may be movements internationally towards more nuanced funding models for hospital and public systems, but change is difficult and will be slow (Ref).
We also have a system that may not even fund the care our patients actually need. Our patients exist within a complex web of social and medical needs. Care that accounts for all of these needs, which provides the optimum outcome for the patient at least cost to the healthcare system should be employed by everyone. This is high value care and our funding system seems not to value it much at all. A good, workable example of this is government rebated management of chronic conditions under the Chronic Disease Management scheme. Patients who have complex care needs are provided a referral to a small list of allied health HCPs, who are remunerated at a staggeringly low $52.95 per session. A minimum duration of service is set to 20 minutes and patients can only have 5 of these appointments per calendar year. This doesn’t engender a lot of enthusiasm on the patient or practitioner’s behalf to meaningfully change any factors that contribute to these complex health needs.
Not only do funding systems provide incentives to do more of the wrong care, they provide disincentives to give enough of the right care.
This train is surely very off course.
This is not absolving responsibility for the practitioner; I will explore that next. It is, though, an explanation demonstrating that the very system we exist in makes high value biopsychosocial care a difficult choice. Akin to jumping from a runaway train. A train whose momentum is maintained by the people involved. ‘Too much medicine’ in the musculoskeletal realm is a monumental problem. Like any wicked problem, the solutions need to be advanced, moveable, nuanced and laser focused. That’s probably not what we’ve tried so far. Stay tuned for part 3.