Dr. Jared Powell, PhD:
Chad, I want to open with a series of quotes. Happy is the one who has been able to know the causes of things. That was said by Virgil, who was a Roman poet more than 2,000 years ago. And he, in fact, was echo- echoing Aristotle, who two and a half thousand years ago said, "We don't really think we know a thing until we've grasped the why of it." Humans have pondered on the question of cause and effect probably since the dawn of time.
Dr. Jared Powell, PhD:
We've tried to explain long winters, droughts, the storms, the plagues, and when we couldn't find the causes, we conjectured them, we invented them, gods, the humeral theory of disease, then whatever the best story of the day was. Physiotherapy and musculoskeletal rehab is probably no exception. Patients ask us why and how every single day. How did this injury happen? Why will this exercise help? Why didn't this exercise help? Why does pressing on my spine make it feel better? And we've traditionally had a confident answer at the ready, strength, posture, mobility, unlocking a joint. These are all mechanisms that we've now started to explore and ultimately challenge. Chad, you've spent a sizable chunk of your very distinguished career exploring these questions. And in your latest editorial published in JOSPT, you suggest the puzzle may never be fully solved. Professor Chad Cook, welcome to the show.
Prof. Chad Cook:
Jared, thanks for having me again. It's been several years.
Dr. Jared Powell, PhD:
It has been several years. We've, we've done a little bit of work, , together in between, which has always been fun. Chad, , for the listeners, who are you and what do you do before we get into the meat of this conversation today?
Prof. Chad Cook:
Well, I am a physical therapist by clinical training. I am a researcher at Duke, Duke University in the United States. , Primarily, , do research on orthopedics. , Lately, it, I've kind of morphed into a translational researcher. So you'll see me do clinical work, some mechanisms-based work, even going into looking at different biomarkers. I have been doing this now for 27 years of research, so having a lot of fun along the way, extremely curious. I like to research things that have always bugged me in the clinic.
Dr. Jared Powell, PhD:
Yeah. And that's why I love some of your research, Chad. Not some of it, most of it, because it always seems clinically applicable and it's written... You have a really nice way of writing, so I wanna commend you for that because some academic writing is a bit of a grind to read and I don't find yours is like that. So kudos to you. Chad, what do you do for fun? I know you're based over there in North Carolina. What are your hobbies and interests?
Prof. Chad Cook:
Well, I'm a, a big hiker now. , We bought a cabin in, , Western North Carolina in the mountains and, , we spent the weekend there. In fact, it was amazing. , Do hiking, I do mountain biking. , Used to run, like to go to the gym just to stay fit. I watch a lot of YouTube. I'll be honest, that's one of my favorite things to do is watch YouTube and then figure out how to fix things that I could never sort out myself. So - Good. I do a lot. Try to stay active.
Dr. Jared Powell, PhD:
Love it. What's your thoughts, and this is just a bit of a curve ball for you, what's your thoughts on AI in terms of, and how its impact on physical therapy and knowledge work more broadly? Do you see it as a threat? Do you, do you think we're not there yet? Do you have any sort of vague insights into AI in terms of its, , impact on physical therapy? So I
Prof. Chad Cook:
Have a lot of thougts and I'll, I'll keep them brief. I actually did a certification through Harvard's AI program for healthcare. I am pro-AI. I actually think it's gonna make things better. I don't see it as a threat. I see it as support. To me, it's no different than when we started using the internet or when we started using handhelds. It really gave us access to more information. , I think we need to learn how to use AI effectively. So that piece of it will be a, a learning experience for a lot of individuals, but I actually think it's gonna lead to better outcomes.
Dr. Jared Powell, PhD:
Cool. So you see it as a, a normal technology that, like the internet, like electricity, that we can use it as part of co-intelligence to, to help our research and to help our, our treating of patients?
Prof. Chad Cook:
I do. And partly because it's been around since 2019. They've incorporated these into healthcare organizations. They're making, , they're doing predictive analytics with these, , very well right now. We're learning things like non-linear, , decision making, which, you know, it's very exceptionally difficult to do in statistics, but AI can do it, it can do it quickly. And so we're learning so much from this already, and it's already incorporated. It's already shown that it can be very beneficial. Yeah. ,
Dr. Jared Powell, PhD:
I'm with you. I think it's, , generally cause for optimism, although the rhetoric out there is, , both positive and negative, which is probably a normal thing in the, in 2026 in terms of a binary political landscape. Anyway, let's go into, Chad. We're gonna move into mechanisms. It's a, it's a passion of mine. I know it's a passion of yours. We're both involved in this research at the moment. I'm looking at shoulder pain specifically. You've published some recent work looking almost at the philosophy of causation, which I find interesting, and also testing it in musculoskeletal rehab with a recent clinical trial that you've published. Chad, can I ask you first, like, how do you think about, how do you define a mechanism? And maybe we can go into these first order, second order, third order questions here. What is a mechanism and why does it matter?
Prof. Chad Cook:
That's a great question. You would think that everybody would know. I asked this question during NIH meeting one time and everybody had a different answer. So I see mechanisms to fall in three buckets. I think there are treatment mechanisms. This is the reasons our treatment works. This is the why. This is the why our treatments influence outcomes. There are disease mechanisms. And again, these are the, why are we, why is the body suffering from this disease either physiologically or psychologically? There's the underlying factors that influence that. And then there's one that kind of falls in a bucket. Those are pain mechanisms. And we've seen a lot of people, , sort these into three primary categories and the reasons they fall within those clusters. The biggest area we study, I think, is treatment mechanisms. And if we think of the treatment being X and the outcome being Y, all of that, what happens in between the Y are the mediating components, those are the treatment mechanisms.
Prof. Chad Cook:
Those are the reasons why we see this treatment influence this outcome.
Dr. Jared Powell, PhD:
Cool. So strength exercise for knee pain is purported to work via getting stronger, for example. Yeah. Would that be an example you would use, maybe manual therapy, some traditional mechanisms might be blah, blah,
Prof. Chad Cook:
Blah. Change in mobility, change in inflammation, change in fear, improvement in coping. I mean, there's so many different potential mechanisms that mediate that pathway that are difficult to study, can be drilled down even further down into biomarkers or molecules or cell level if a person wants to do that. I would argue that that's important for some areas of disease work, but not all. So when, when somebody asks me, you know, how do you measure the effectiveness of a mechanism? , The first question I ask is, "What's your outcome measure?" Because that's gonna dictate a lot on which mechanisms you plan to measure. Yeah.
Dr. Jared Powell, PhD:
Okay. I think that ties or, or leads nicely into this concept of first order questions, second order questions, third order questions. And I even hear, Chad, there's a fourth order question. Do you mind running us through these, , order of questions,
Prof. Chad Cook:
Please? I, I would love to because I wish everybody thought this way because it explains things so well. A first order question is, does a treatment have an effect? And we compare this against a placebo or no treatment or usual care or something. So if it has an effect, we know that that is a first order question. We just know that we're not dealing with snake oil or a useless intervention. It has an effect. A second order question is why does it work? And this is the mechanism piece. You would use some sort of mediation analysis or some sort of path analysis to determine why this particular treatment, which has an effect from that first order, why it actually leads to the outcome it leads. Very important, I think. It gives us a perspective on why our treatments work. A third order question is who does it help?
Prof. Chad Cook:
And Jared, you and I both know this is the least studied out of all of the work that is done in MSK. We, we have thousands upon thousands of, of randomized controlled trials, but exceptionally few moderator analyses which fit within that, who does it help? And that way at your baseline, you know who should get manual therapy, who should get exercise or who should get a talk-based therapy. Fourth order questions are, okay, we've got this innovation or intervention. We know it's gonna work. What can we do around it to set it up so that it's even more effective? Is there a particular language that we can put in that place? Is there an environment where this person is more likely to excel? Should we see them quicker or should we give it a little bit more time? What's the cadence of the treatment? What kind of follow-up do we need?
Prof. Chad Cook:
All of these are fourth order questions, which are also exceptionally beneficial too.
Dr. Jared Powell, PhD:
Okay. So in summary, , first order question is, is there an effect? Second order question is, how is that effect generated? Third order question is, in for whom may this effect manifest or not? And fourth, can we optimize the whole clinical scenario perhaps in order to, to make that effect amplified, better, or, or how is it getting dampened even? When you're... , Do you still treat patients, Chad?
Prof. Chad Cook:
Just in the trials. So I'm not in the clinic, but I do all of the care for the people in the trial.
Dr. Jared Powell, PhD:
Do you think clinicians should be thinking, maybe not formally, but along these lines? So first they're thinking of, "Here is a treatment that I've read from voluminous randomized control trials that is slightly better than usual care. Then, you know, there's less, there's less mediation analysis out there, but I know that maybe it works, , this intervention works through enhancing pain self-efficacy. Then I know that it works for this person with these baseline characteristics." Then do you think they should be thinking about, "Well, how am I going to optimize the clinical scenario here in order to make this treatment as effective as possible?" Do you think that should be rippling through the minds of clinicians when they're seeing a patient or do you think that would make it a bit too formulaic, algorithmic, and perhaps, you know, divorced from seeing the patient in front of them?
Dr. Jared Powell, PhD:
What are your thoughts around that rambling question?
Prof. Chad Cook:
I have a lot of thoughts. I, I think the first is we expect a lot about our, out of our clinicians who are overworked, asked to do more every year without resources, and they need help. So I think that kind of load upon the mental processing might be a bit much without some sort of external help. And that's where I think AI might actually step in and make suggestions and, and that sort of thing. We also know that people respond differently to treatment. So even though they may fit this pattern or this pathway may not respond well, and that's where a really high level clinician can recognize that, "Hey, you know, that we should be seeing a selected type of response. We're not, we're gonna pivot and we're gonna move on to something different." I think clinicians are asked to do a lot. I think it's time for us to ask our researchers, our funding sources, and our people who write clinical practice guidelines to help us out in that area, maybe make things a little bit clearer and easier for clinicians.
Prof. Chad Cook:
I, I don't think clinicians should be asked to, to be the, the lead pool in this.
Dr. Jared Powell, PhD:
100% agree with you. Chana, I could talk about this all day, but I've gotta, I've gotta r - I've gotta remain slightly tethered to some questions that I want to ask you so I have some sort of a structure to today's conversation. So why do we need to understand mechanisms? All right? Why isn't, why isn't the observation that a treatment works enough? Why do we need to understand the second order question? How is that helping us when we know that if you just give someone a course of manual therapy or an exercise program, it'll make them marginally better than if they hadn't had that? Why isn't that sufficient, Chad?
Prof. Chad Cook:
Okay. There's, to me, there are layers to that question, and I'm gonna break it apart. The first one is, I think in some treatments, understanding mechanisms is going to be more important than in other areas. So for example, if the disease mechanism is life-threatening or if it's a really progressive disorder that can lead to massive disability, then we need to understand mechanisms so we reduce the potential harms of not giving them the right treatment, but also to better understand that how this influences the disease, how there's that match between the treatment and disease. So I think of things such as neurological disorders or cancer where they're really looking at precision-based components. And that's, that's a why question. So in that population, it's essential that we understand it. We deal in the world of MSK where our outcomes measures are essentially proxies of a concept. So it's a proxy of a disability, a proxy of a person's pain experience.
Prof. Chad Cook:
There's variability in that, in the way that a person reports that or experiences that. So in some cases, the why we see a change in that particular area where there's a lot of gray, one could argue it's not as important. To me, it still has important. And, and if you don't mind, I'll give you a scenario. In our mediation analysis in our SSMEK trial, we actually found that changes in fear and changes of coping were the two areas that mediated the outcomes the most. So if you reduced fear and if you improved a person's coping, it led to greater outcomes. And we know there are treatment strategies that people can do that they can incorporate into their treatment to actually improve those areas. So if you understand the why behind it, that it's often psychological components that are mediated, then you can incorporate those into your treatment approach and that's, that's a fourth order piece, and then you can improve that overall intervention.
Dr. Jared Powell, PhD:
So knowing how may lead to the tailoring or the optimizing of a particular treatment to, to target that particular, , causal mechanism that actually influences the outcome rather than just throwing a bunch of whatever, poop at the wall and hoping something sticks, right? It, it makes us a little bit more targeted in our approach. Is that, is that an okay way to surmise what you just said? It
Prof. Chad Cook:
Could potentially do that, yes. We're not near there yet. I, I think we're, we're scratching the surface. You and I both know second order, third order questions are... There are many. There are, there are more fourth order questions than second and third order questions studies combined. , And more work needs to be done in that area.
Dr. Jared Powell, PhD:
So I wanna, I wanna now generalize or extrapolate this talk of mechanisms into the stories that we tell our patients, because this is what every clinician probably cares about the most. So I did a scoping review a number of years ago looking at the causal explanations for shoulder pain that are proposed in clinical research, and we found 30 to 40 odd causal explanations for how an exercise program or a set of exercises was going to help a person with shoulder pain. I know there's a lot more mechanisms out there for manual therapy and dry needling or acupuncture. Insert intervention, there's going to be dozens of, of proposed mechanisms that will explain that treatment's effectiveness. When we look at the traditional narratives that we tell our patients in physical therapy, strength comes up a lot, and that's something I've researched, and that really doesn't seem to be a tenable causal explanation for, for many of our treatments, especially even resistance exercise.
Dr. Jared Powell, PhD:
Posture is now one that we've, I would say, refuted manual therapy, we've had to reconsider how manual therapy might be working. Are these explanations that we know to be incomplete or even wrong still useful? Should we still say them to patients? Or do you think we're h - they're harmful and we should give them the truth even though that truth may be a pretty tough thing to hear that we actually don't know how this intervention is going to help you?
Prof. Chad Cook:
There's a lot there. Let me try to unpack it. First of all, I'll start by saying, I think we went a little overboard on the be careful what you say thing. I mean, there were papers published on this, and I appreciate those papers, but I don't think it necessarily derails a person's recovery process if you use the wrong word in your description. Now, if you catastrophize, if you tell somebody that, you know, this, you've got this degenerative process, it's unlikely to improve unless you get surgery, I think that leads a person down the wrong path, but I don't think a few words that fall into the don't say list is gonna necessarily derail a person's outcome. , I'll start with that. Is it important to the way you communicate the telling a person, , potentially why these interventions work? I think you have to be able to have a conversation with the patient with some knowledge on what we know so that you don't lose that patient's faith in you as a clinician.
Prof. Chad Cook:
Now, you had mentioned there are dozens, , potential mechanisms for manual therapy. Our umbrella review in plus one found hundreds of mechanisms that were theoretically affiliated with manual therapy, hundreds. Most of those were neurophysiologically based. So I think saying some kind of conversation around, we know that this has a fairly strong influence on the pain pathways of the body, and some people really experience improved pain and improved movement with this. , Multiple studies have shown that. In some cases, range of motion improves, et cetera. I think it's okay to say that. What we don't know, Jared, is whether these are the reason we see improvements with manual therapy. We just finished a systematic review and we found two papers who claimed to find causal mechanisms, but there was not a temporal sequence, so there wasn't a causal pathway. So there are actually zero papers that take a mechanism and measure it in a clinical outcome trajectory longitudinally to explain what happens with manual therapy.
Prof. Chad Cook:
We really don't know, but the best thing we can say for the patient is it's been studied a lot and there's a lot of preclinical work that suggests it's this. It's probably something around here. It... We generally know if patients are gonna respond fairly quickly from this, so we'll be looking for that, but that's essentially what we know about this condition, or this treatment, excuse me.
Dr. Jared Powell, PhD:
So manual therapy has no real mediation analysis that supports a particular mediator that is responsible, partially at least, for a change in a particular outcome. Is that, is that what you just said?
Prof. Chad Cook:
There are no unique mediators to manual therapy that aren't also present with another treatment approach.
Dr. Jared Powell, PhD:
Why do you think our field of physical therapy hasn't explored with gusto causal mechanisms up until the last, let's say, decade, if we're being generous? Why do you think it has been something that we've neglected? Is it hard to get funding? Do we not care about how and why have we been more concerned with doing a large scale randomized controlled trial that shows a one point difference in this intervention versus that intervention? Why do you think we're lagging behind here versus other profession like in psycho - I look at psychology a fair bit here and they've been thinking about how and why of cognitive behavioral therapy in their treatments for quite some time. Why are we struggling here?
Prof. Chad Cook:
Well, I'm, I'm gonna say what I know you think too, all right? So I know you're asking, you're setting that up for me. First of all, it is a funding issue. Funding agencies fund people who already have received funding, there's an old boys at work, and they fund trials. I mean, you, you're more competitive if you put together a trial. And so that's one of them. We also know that mediation analysis or any of these causal pathway studies, they need a large sample size. So this is an expensive study that's more expensive even than a randomized trial. And the average trial here in the United States is $3 million. So if you're going to double or triple the number of individuals that are needed to, with realistic precision, do a mediation analysis, you're talking about a major cost. And, you know, we've only really been seeing, , funding agencies, and it's primarily PCORI will require moderator funding, treatment, , heterogeneou treatment effects built into the trial, but they don't require that for a mediation analysis or for a second order question.
Prof. Chad Cook:
So it's cost, it's funding agencies, and you also know that it's also very difficult to do these well. , Because they're longitudinal, you have to bury the mediators in between the first stage and the final stage of the outcomes so that you have a temporal sequence associated with that because so you can really tell if those mediators are truly influencing the overall outcomes. So I think it's all of those things. And I, I think if you take the average good researcher, effective researcher, and you sit down and talk to them, they don't understand causal mediation analysis that much. It's complicated.
Dr. Jared Powell, PhD:
Yeah. It's, it's really complicated. I've, - Yeah. Had to, had to bury my head in the data a few times and, you know, I was already bald, but if I wasn't bald, Chad, I reckon my, all my hair would've fallen out again. It's, it's pretty dense stuff. Do you think this has... Do you think that this reluctance to embrace these more complex questions, such as second, third order questions, has hampered our progress as a profession? Or do you think that we're just humming along and that this interest in how and why is a little bit more abstract and philosophical and it's not, it's not really changing clinical outcomes that much? What are your thoughts around it from a, you know, a higher order level if we zoom out a little bit? Do you think that we're, we're actively obstructing patients in getting a slightly better outcome by this ignorance of how and why and who?
Prof. Chad Cook:
So you read a lot of psychological literature. I do as well, but I've also lately on a number of trials had worked with psychologists and they think very differently. And I've heard one of my favorite psychologists who does mediation work, he once told me, "Wouldn't it be interesting if everything is purely just a therapeutic ritualism or a therapeutic ritual effect, that it's just that encounter that makes the biggest difference, and it doesn't really matter necessarily what the treatments are, that's the weeds." He goes, "Wouldn't it be funny if it fell to that?" And I won't lie that there's a small part of me that actually believes that's a big part of what our benefits are. The fact that we give people the permission to move forward, we give them the permission to improve. I mean, we're dealing with MSK, right? MSK is, there's so many factors that are wound up in the outcomes with MSK.
Prof. Chad Cook:
So giving a person the opportunity to, "Oh, I can get back to being normal. I can go back to work. I can walk again. I can do these things." I mean, maybe that's a part of it. , The other part of it is, and you ask us, did, do, have we done ourselves a disservice? I think we have. And I think of clinical practice guidelines. I massively support practice guidelines, but these only involve first order questions. So you're giving people guide rails, and these guide rails are very wide, and you got all these interventions that could potentially work, and here's an A or a B or a C, a D, or whatever, and you go in there and you sort that out yourself, and, you know, which person's gonna benefit from this? We don't really give a lot of guidance on that part, and I think that's a mistake.
Prof. Chad Cook:
We had this big passion for clinical prediction rules for a while with... That we're theoretically supposed to g - to help on that. They were all derivation studies, and you and I both know they were basically prognostic studies. That's not going to drive a treatment choice. A moderator analysis is a very sophistated... Excuse me, sophisticated study that involves a different type of analysis, and we just don't do those. So yeah, I think we've done ourselves a disservice. The jaded part of me thinks, okay, maybe it's just that interaction, but the hopeful part of me actually sees moderator analyses, yeah, and says, "Okay, this is different." Like the, the study that looked at CFT that Mark Hancock led that showed high disability should receive CFE - CFT versus people with low disability. , Makes total sense to me. Even in our SSMET trial that we just finished, we found a number of moderators that suggested someone might benefit from manual therapy, but not exercise or benefit from exercise, but not manual therapy.
Prof. Chad Cook:
Those are the sort of things that give me hope that we'll get to a point where we start... Where the clouds start to clear on that.
Dr. Jared Powell, PhD:
Is this where we go down a pathway of precision medicine, which is you've done a lot of work in with Amy Devitt and, and others. How do you see precision medicine fitting into this research program, and how do you see that having a positive effect on patients' clinical outcomes?
Prof. Chad Cook:
Well, within it, by its nature, I think precision medicine is a third order question, is, you know, how can we tailor this treatment to the person who's likely to benefit the most from it? Unless we do that moderator work, we're not gonna know what that is. , So I do think we can get to a point where we can potentially do that. The other thing, Jared, you and I know that our outcomes measures are imperfect. , There's a lot of error that is built into one of these measures. We also know that there are a lot of factors outside the primary disorder that influence those measures, including just overall health. So any model that we build is competing against a lot of confounding areas, and that, that really erodes our ability to be as precise as we need be. I think, you know, there's stories ahead of us, cancer is, but cancer's a clearer type of disease process than our musculoskeletal disorders are, which are so heavily affected by psychological components, comorbidities, et cetera.
Prof. Chad Cook:
And there... It's apples and oranges. So it's gonna be difficult, I think, to, to ever make the headway that we see in some of these disease processes in MSK with precision medicine.
Dr. Jared Powell, PhD:
Perfect. This leads very nicely into my next question, and I wanna push back a little bit here, because I believe it as well, so I want you to help me. I love complexity, and there's a line in your recent editorial in JOSBT that I read multiple times, and I've written it up on my whiteboard as well, because I often find myself thinking about it. So you wrote verbatim, "Therapeutic interventions ripple across multiple pathways simultaneously, producing effects that interact, which is key, amplify or dampen one another in ways that are often unpredictable and hard to measure." Now, if we take this to its logical conclusion, if we think pain is a complex experience, and, and I do, then how can we be precise with our treatments if it's unpredictable and hard to measure? Are we just making assumptions about probability based on what we know about the moderator analysis that we have, but we're always gonna be slightly detached from predicting with certainty how this person is going to go with this particular treatment?
Dr. Jared Powell, PhD:
Can you run me through your thought process around complexity, which means things that are interacting and emergent, and by definition, impossible to predict, and reconciling that with this research program of precision medicine, which is trying to almost collapse all of that complexity down into something that we can feel and is tangible and that we can predict with some certainty?
Prof. Chad Cook:
Yes, first of all, thanks for reading that. And yeah, I mean, you've written a number of editorials too. These things marinate over a long period of time, and this was one that we started writing this, and I think it was eight months later when we finally decided to submit it. And, and it's because of what we were experiencing. We kept finding studies that would look only at cortisol or look only at oxytocin or, or some sort of, you know, ILA-3, a- assuming that this thing functions in a vacuum and doesn't interact, and doesn't interact even at a systems level. And we had written a, , previous paper when we were defining mechanisms and explaining that, "Hey, you, you know, there is... There's never gonna be this linear path that you have a treatment that affects one biomarker, and that biomarker's gonna then run by itself to the, the finish line.
Prof. Chad Cook:
It's gonna cascade around like a pinball, and it's gonna influence a lot of different things." So with that said, our reasoning behind when we wrote this in the editorial was, "Hey, the simplistic, I'm looking at two biomarkers or one biomarker or five, or even eight biomarkers which I can afford to measure and s - when I send it in and they do all the assessments, it's very expensive," that even those in a comprehensive biomarker, , assay, it's gonna be so limited in really what it tells us, because we won't be able to measure all these other things, and we won't be able to measure them in a clinical trial. We'll have to control things so tightly. That it's gonna be preclinical or we're gonna have to do an, an animal study, and that transferability is almost nothing. The sentence that you quoted is almost the essence of the paper that we wrote, that it's just so complex that we're probably never gonna fully understand it, especially for MSK.
Prof. Chad Cook:
So the complexity, I, I think in my mind, we're never fully gonna understand what happens with our treatments because we don't completely understand what's happening with MSK. We don't even really understand the disease mechanisms of MSK. So we don't understand our treatments and completely what they do, and then we don't understand the disease mechanisms completely. And then we're doing a study where we push in a cat's back and assume that it, this is the reason. I mean, it's just like, no, no, it's n - we're not there. It's, it's so much more complex than that.
Dr. Jared Powell, PhD:
It's a bit, it's a little bit embarrassing when you articu - articulate it like that. It's hardly the rigorous science that we want it to be, right? But to give us some, you know, to maybe take the heat off us a little bit. I think you've said this quote, Chad, I think I heard it a lot, , quite a while ago on a podcast, something about we were building the plane as we were learning to fly it or something like that. Can you give me that quote again?
Prof. Chad Cook:
Yeah. We were, , building the plane as we were flying it. Okay.
Dr. Jared Powell, PhD:
Yeah. And do you think that runs through this conversation that we're having here about, you know, we haven't explored how and why. So do you think that is a reasonable excuse because we're this young profession, evidence-based medicine is only 30-odd years old. You know, let's not forget that up until the mid - 90s, it was, , the wild, wild west out there, , when it came to medicine. , I've got an interest in the, the statistician and epidemiologist, , Austin Bradford Hill, who was involved in one of the first randomized controlled trials in the, in the 1950s. And he proposed, you know, these nine tenets of causation to assess smoking as a cause of lung cancer. You know, that's not that long ago. 70 odd years ago isn't that long ago. So do you, do you feel like that is a... Can we explain away our current failure in looking at these things as we're a young profession?
Dr. Jared Powell, PhD:
Or do you, do you reckon we could have done better?
Prof. Chad Cook:
So I appreciate you outlining that. And by the way, if you ever wanna set off reviewer two, talk about the Bradford Hill criteria. They just take... The people either hate it or they love it. And it's just -
Dr. Jared Powell, PhD:
I'm in the process, Chad. I'm in the process. I just stay away from it. , Sorry, I'll interrupt for one second because I have a, an anecdote on that and then you can get into yours. My last, , paper in my PhD, my last chapter was applying the Bradford Hill criteria to exercise therapy for, , shoulder pain. And tried to get... I've been trying to get it published for about three years now, and it's been an absolute nightmare. So anyway - Yep. As you were.
Prof. Chad Cook:
Yeah, there's nothing that sets a reviewer off any more than discussion around causation, because everybody tr - is so trained differently on that. So we are a young profession. I actually did a talk on this in which there were multiple, I, I'll call them pioneers in manual therapy that were in the audience. And my comment was what, what sort of lazy argument is it to sit back and point fingers at someone who didn't have the technology or the science or even the understanding to truly measure what, what they were seeing in the clinic? They saw value in the clinic and they proposed a framework on why this was potentially beneficial. There is nothing wrong with that. That should not be made fun of whatsoever. As science builds, some of it will be refuted, some of it will be supported. It becomes a problem when the stuff that is refuted is, is still supported.
Prof. Chad Cook:
And, and we know that. That's an issue with many philosophies, especially those that have, have a market in that area and have a profit to make. So that part of it is certainly some of it. I think we've become more sophisticated as a profession in how we're investigating things. I think all of medicine has done that, and we will slowly move in that direction. Someone will listen to this podcast in 10 years and say, "That dude was an idiot. I can't believe he said those things. He's right on a few things, but some of the stuff he's saying, he doesn't know what he's talking about." That's a natural progression, I think, on our knowledge a- around our treatments. We don't know that much about them yet. , We know what works in the clinic. We've seen that. We've been misguided, I think, historically from some of, of this, but we're getting better as we move forward.
Dr. Jared Powell, PhD:
Chad, do you think that sophisticated clinicians who are on top of the latest research should call out peers or other clinicians who are using narratives or stories about how their treatment is working, let's say for manual therapy, that it's simply breaking up adhesions or doing... I don't even know. I don't keep up with the manual therapy stuff, but it's doing something that's being disproven. Do you think that we should call out our peers in order to progress the profession quicker? Or do you think we should respect their right to believe what they wanna believe and just worry about ourself? I have an opinion on this. I, I wonder what your opinion is.
Prof. Chad Cook:
So I, I will speak for myself and how I have managed things. I don't wanna say anything or act in a way that I would embarrass my mother moving forward. So the way I interact with individuals is always with the benefit of doubt. And I don't always assume that I understand what a person is thinking. You know how many... Because I refused to agree with everybody when they were dumping on manual therapy, you know, seven, eight years ago. I was brutalized on social media. People were saying, "Well, he believes this and this." And it was nothing of what I believed in. It... All it would take is somebody to do a five-minute Google search to see that I had questioned all of these things in my first textbook that I wrote basically saying, "I'm sorry, but a lot of these philosophical frameworks are not correct, and here's why." I mean, I've been saying this for 20 plus years, so I'm very cautious about claiming someone believes this or someone suggests this.
Prof. Chad Cook:
The other thing is I got better things to do than break down somebody's though process. I don't necessarily think it derails our profession. We got other things that are more important. I, I'll tell you this. I... We're seeing a shift finally from this traditional approach of front-loaded care for our patients with chronic pain to more of an extended, back-ended supported approach. I should have been available 20 years ago, and we're just now seeing it. I would complain more about that than I would about what some clinician in some nowhere else place is saying to a patient. I, I... To me, that's less of an issue.
Dr. Jared Powell, PhD:
Yeah, good. I agree. Chad, can you run us through your SSMEK trial? So this is one of the best trials I've read. I love how you've actually embedded mediation analyses within it, a moderator analysis as well, which is great. I don't know. I mean, I can't think of a trial that's done all three. Maybe the, , maybe the... , What's the trial that was led by Kent and Hancock that you said a moment ago, Peter O'Sullivan - Oh,
Prof. Chad Cook:
Yeah. What is that trial called? It's. It's a really good one.
Dr. Jared Powell, PhD:
- Is it RESTORE or is it something else like that?
Prof. Chad Cook:
It is RESTOR. Yeah, I think it's RESTOR.
Dr. Jared Powell, PhD:
Yep. Okay. Maybe they've done it as well. They've certainly done a mediation analysis, moderator analysis. So can you run us through, set it all up from the trial to the mediation analysis to moderator analysis, please?
Prof. Chad Cook:
Well, I got interested in this concept called shared mechanisms, and it was basically from talking to John Burns, who is a psychologist out of the United States. And, and John has found that there was really no superior type of psychologically informed approach, that they all basically have a common pathway. They have shared mechanisms. They influence the mechanisms affiliated with psychoec - psychological change.
Dr. Jared Powell, PhD:
Sorry, Chad, can you just, , define shared mechanisms for everyone, please, upfront?
Prof. Chad Cook:
Yeah, you got it. So we know that there are specific mechanisms of our, , treatments. Like with exercise, we assume that it creates larger torque, , that you'll see cross-sectional fibers increase if you do it for a while. Manual therapy, we assume that it is a pain modulator, that it has a central mediated effect, , because we've seen this in FMRI studies, et cetera. So th- those are specific to the intervention. A shared mechanism is where both interventions influence a third variable, which in turn influences the overall outcome. And that third variable, in many cases, is psychological. So coping, self-efficacy, reduction of fear, you know, all of these things. And then you can design a trial to actually see if it's those things that are influencing the outcomes more so than what we think the specific mechanisms are. So I was interested in that because if you really looked at the literature, the outcomes for neck pain and comparative trials of manual therapy versus exercise are basically the same.
Prof. Chad Cook:
The average treatment effects are the same and there's no superior group. So I said, "I wonder if it's because of shared mechanisms. So I'm gonna make this a mechanistic clinical trial, and we're gonna measure mechanisms as our primary outcome, and our secondary outcome will be outcomes measures." So we captured a lot of mechanisms, proxy mechanisms, direct mechanisms, and to see if they were different between the two groups. Interestingly enough, they weren't. The specific mechanisms were even the same across the groups, but the shared mechanisms were the same as well. So we looked at that as a first order question. The second order question, we were interested in were there, were there any factors that mediated group-related responses? There were not. And then we looked, okay, then if we lump these together as a group, exercise and manual therapy together, are there any mechanisms that actually mediated the overall outcome?
Prof. Chad Cook:
And there were, and it was fear and coping, so psychological variables. The work we just finished was a moderator analysis, and that's where we looked at baseline characteristics and whether or not those baseline characteristics suggested that a person would benefit more from exercise but not manual therapy or manual therapy, but not exercise. And what we did is a type of analysis that actually told us where the thresholds were too. For example, if it was range of motion, we knew that if a person had X amount of range of motion at baseline, they were gonna benefit from exercise. And if they had limited range of motion at baseline, X amount, they would benefit from manual therapy. We found a number of them that actually predicted that. The most interesting one I, I though was fear. People who had high levels of fear at baseline actually did better with manual therapy than they did with exercise.
Prof. Chad Cook:
Completely the opposite of what we've reported. But in the same token, those with low levels of fear, start them with exercise, you'll do better with exercise, but not manual therapy. So there, that paper will be coming out very soon. Again, musculoske - skeletal science practice. They are doing a dang good job. , That will be the home for that paper. And there are, , several moderators that suggest manual therapy, but not exercise or exercise, but not manual therapy.
Dr. Jared Powell, PhD:
Love it. That's gonna be such a valuable contribution. Can I ask about the fear, , moderator? Do you hypothesize that someone who had a high level of fear, , who improved with manual therapy, that that was mediated by a reduction in fear, fear change, , with manual therapy or d- did something else change? Were they still highly fearful at the end?
Prof. Chad Cook:
So we did see a reduction of fear quite a bit with manual therapy, more so than exercise, but only on people with higher levels of fear. That's what's cool about a moderator analysis. Those individuals with low fear at the start, if they received manual therapy, they actually improved less than if they received exercise.
Dr. Jared Powell, PhD:
L- less in total in terms of the outcome or less with a fear reduction? Less
Prof. Chad Cook:
In total. Yeah, less in total.
Dr. Jared Powell, PhD:
How do you propose that manual therapy is reducing fear? What are your speculations there?
Prof. Chad Cook:
Yeah, what we put in the discussion is, is that if you watch a good manual therapist, they don't just do something to a patient and then run away. They basically take the person through the movements and then talk to them about... Now you can, you can see that you can start moving in this position now. Let's move on to... It's very similar to what you see with CFT. They do more of an active approach with the talk-based overlay. A manual therapist will just do it passively with a talk-based overlay. So we proposed that what likely happened is they had a pain controlled situation in which the manual therapy was applied and the person could see that they can move through these movements at the neck and reduce their fear. So, oh, I can move, reduces... And their fear is reduced from this. That's what we proposed within the discussion.
Prof. Chad Cook:
What we really don't know unless we followed up with some sort of qualitative analysis and just ask, "Why? Why do you think this helped? Why do you think you reduced your fear of movement?" What is that called? A sequential explanatory design, which I think are brilliant designs, by the way, and we need more of those as well. But we, you know, ours was completely hypothesized at the end. This is what we found. Will it carry over to the back? I don't know. Shoulder, knee? I don't know. And we found it at the neck, and that's it.
Dr. Jared Powell, PhD:
I don't think the neck should be any different, in principle. , I, I have such a, ... I love these... We... I did a, , qualitative study asking patients how and why they perceived, , an intervention worked for them. And it's such a cool study to do when you, when you ask the patient, "Why do you think manual therapy helped you?" And the, the stories that they come up with are so compelling. And m- mine was on exercise, and a, a particular lady described how simply being told that she was allowed to lift a two kilogram weight into abduction or scaption plane broke down her entire fear of that she was never a- a- allowed to lift her arm in the air and that her rotator cuff was gonna explode or that her acromion was gonna rip into a supraspinatus and cause a, an irreparable rotator cuff tear.
Dr. Jared Powell, PhD:
When she was told by the clinician that she was allowed to do that and in fact was going to be helpful for her, that what she described was like, you know, brought a tear to my eye almost without being overly pathetic. , Reading it, it was such an amazing thing to observe and to read. So the paper's called Restoring That Faith in My Shoulder. It was published in PTJ, and it was the fa - it was the best study I've ever done in terms of its impact on me. So I really love how you're, how you're thinking about embedding the patient's perspective into that as well. Do you think that's something we should do more of, Chad?
Prof. Chad Cook:
100%. And so that's the depth behind the breadth of what we find, right? Qua- quantitative will give you a breadth of findings, but it doesn't tell you exactly why at a deep, deep level to the patient. Only the patient can do that. And sometimes they will bring insight in that you're like, "I had no idea that that was meaningful. That is stunningly interesting." We've done a few of these now because we were, we were interested in why people were coming to therapy for one time and then never showing up again. We were interested in why people were either adherent to their home exercise program or not. So we, we did qualitative analysis on that and really found some fascinating information that could be very helpful to a clinician, just a few words, just a few recommendations to really tie it in to make a big difference.
Prof. Chad Cook:
I also have found these to be the most rewarding. , One of the, one of the grants that I haven't got that I still think has s... We submitted a grant to the McKinsey Foundation, or it's called the MDT Foundation. And what we wanted to do is look at patients who received a, a very strict McKinsey-based approach. We wanted to ask them a question, "Why do you think this helped you? Or why did... Why didn't it help? What happened that, that made you think that this wasn't helpful?" We need that type of information. And, , and the only way to do is through a, a qualitative analysis.
Dr. Jared Powell, PhD:
Yeah. No, I agree. Can I ask you quickly about, , patient expectations? So that, that's been the typical moderator that we've, in physical therapy, been... We've thought of as being maybe an inhibitor or a facilitator of an outcome. So if somebody comes in, they have low expectations of physical therapy for whatever reason, and that tends to be associated with a wo- worse outcome with physical therapy. In the shoulder, that's led to an earlier decision to undergo rotator cuff repair surgery as opposed to going down a 12-week program with a physical therapist. Expectations are now being thrown around as also being a mediator of an outcome, but I'm not aware of any research that shows that we can actually change expectations very well. I think, have you published a study adjacent to this or on this? And would you like to expound on that?
Prof. Chad Cook:
Yeah, it's exactly on this. And it's the PERC study that we did, and then we published about three years ago in BMC Musculoskeletal Care. And we attempted to change expectations indi - on individuals with shoulder pain who were receiving care under a physical therapist. Now, this was during COVID, so we ran into all kinds of issues with finding a, a, a decent population for this. What we found was this. There really weren't that many people that had low expectations. We had a ceiling effect of really high expectations. We also didn't really like the tool that we used, the expectation tool we used. So I'm, ... And so we didn't find an effect. We could, we couldn't change expectations because they were already high. We didn't find a baseline to long-term outcome difference in expectations because we didn't have enough data spread. In my experience, I've had six studies that I've been part of with patient expectations.
Prof. Chad Cook:
Big piece of it is how you ask the question. You know, if you just ask a person, "Hey, you think you'll have a, you know, a, a positive outcome with this particular intervention?" It, it's really important when you ask the question, and you ha - they have to understand what you're going to give them so they can truly base their expectations based on that. If you ask them before you actually prescribe a treatment, they're, they're going off guesswork. They're going off that visit, that encounter, and not actually intervention. And there has to be enough data spread, and it's really hard to get data spread on PTs because everybody loves their PT and everybody has high expectations and, and you end up with, , kind of useless information. In our systematic review, and gosh, I don't remember where we published it. It was out of Belling College.
Prof. Chad Cook:
They actually didn't find patient expectations to be as compelling as a predictor as what people though. So I put this in a bucket of this is meaningful and interesting. If there's a complete contradiction between where your patient's at and where you're at, that's gonna be meaningful. If a person says, "Yeah, I think this is gonna be beneficial," so it's probably not gonna matter much in that, in that individual.
Dr. Jared Powell, PhD:
Were you involved in that paper? I think Nathan Hutting led it with, like, the expectation, , like the red light, the orange light, and the green light with concordant expectations and then flat out wrong expectations. Were you involved in that paper or do you know what I'm referring to?
Prof. Chad Cook:
I do, and I don't think I was part of that. , I've been on several of Nathan's papers. , Giacomo and Nathan have teamed up, and they've pulled me in a few things, but I don't think I was part of that one.
Dr. Jared Powell, PhD:
Okay, cool. And it's kind of around, you know, if someone comes in with concordant expectations with what the guidelines say and what the clinician wants to give them, great, best case scenario. And then there's expectations that the patient has, which might be slightly different, but doesn't really matter. You know, you don't need to go in and go, "You're completely wrong with how you think your rotator cuff tear came on. You don't, you don't really need to have that conversation." And then there's ones that are, you know, completely false and at odds with what the evidence have. If they have an expectation that their pain is going to be better in one treatment, even though they've got fibromyalgia, you need to have a conversation about why that's perhaps incorrect. So I, I really felt like that was a nice way of looking at expectations. If they're concordant, if they're different, but it doesn't really matter, and then if they're completely opposed or at odds with the literature.
Dr. Jared Powell, PhD:
So that might be helpful for some, , clinicians to think about. It's led by Nathan Hutting, and I think it's published in MSK Science and Practice as well just this year. Chad, conscious of your time, we're gonna begin to wrap up. Where, where next for you personally, and where next for the field? So what study do you think could, in principle, move the field forward? So what studies should we do? And more importantly, so two part, what should clinicians do in the meantime with respect to explaining the mechanisms of our treatment to, to their patients? What, what should they do in the face of so much uncertainty?
Prof. Chad Cook:
So I think my hope is that we do moderator analyses or even a moderator meta-analysis that we can better understand a broad spectrum of baseline characteristics to allow matching of treatments. I think that's holy grail type work. I think it's needed so much in our profession. And I'm somewhat, , optimistic, especially after our moderator analysis found so many baseline characteristics in a population of 124 people. To have that sophisticated of an analysis pull out so many differences, it gives me a lot of optimism. And your second part of the question, if you don't mind again?
Dr. Jared Powell, PhD:
What, what should we tell clinicians? What... How do clinicians have the conversation about uncertainty? But we know from qualitative work, patients want confidence and competence out of their clinician. So how do we remain competent and confident seeming without hedging on, "Oh, we really have no idea. Don't worry about it. Let's just do this."
Prof. Chad Cook:
Yeah, I think the worst thing we can do is Golden Corral, which is a cheap smorgasbord restaurant here in the United States where basically everything costs the same and you can pick whatever you want. , That's the worst thing we can do with our patients. I think we have to show that, hey, this is what we know. And the first thing I would preface things with is based on average treatment effects, based on averages across people like you, here is what we've seen within the literature. People tend to benefit from exercise. It may not help overall with your pain, but it's gonna help with your health, and you may see an indirect effect from that. There are some individuals though that have a reduced pain response, not everybody, but we have seen it within the literature. I think telling that story when you're framing up a treatment intervention can be extremely valuable.
Prof. Chad Cook:
And also, and can indicate that we do know that individuals who have stuck with the program and are adherent to this and, and really look for, you know, the signs and symptoms of a good recovery, they do well. And then you can self-manage this and you will improve. That, that sort of message, messaging, I think, is accurate and beneficial. And just like what you mentioned earlier with the qualitative, that sometimes it's the little nuggets that we tell our patients that they latch onto that really make a difference. So framing things positively, indicating what we know, indicating that this is based on averages now, average people are different, but, you know, on average, this is what we know.
Dr. Jared Powell, PhD:
Perfect. Chad, thank you very much. Chad, where can people find you? Are you still on the socials with the collapse of Twitter? What, what are you doing now? Do you, do you spend time on any medium or platform?
Prof. Chad Cook:
So the collapse of Twitter is so real. , I, I hardly ever post on Twitter now, and, and I don't read it either because it's just kind of, it's so different. , I'm primarily on LinkedIn. , Linkedin appears to be a more gentlemanly environment to share information, and people tend to be positive. So I will post my information there, and I will post studies that I think are meaningful as well from colleagues or from, you know, individuals who do work that I really respect.
Dr. Jared Powell, PhD:
And you're just at Chad Cook, I imagine, on LinkedIn, easy to find in a sense?
Prof. Chad Cook:
I think so. I, I don't even know what my username is, but I think, I think it's... Is it, is it Chad Cook PT? I, I don't even know. It, but I'm, I'm on LinkedIn.
Dr. Jared Powell, PhD:
I'll find it and I'll link it. Thank you. Chad Cook, thank you very much.
Prof. Chad Cook:
Jared, thanks for having me again. It's been fun. I'll see you in another seven years.
Dr. Jared Powell, PhD:
Thanks, mate. Thank you for listening to this episode of the Shoulder Physio Podcast with Professor Chad Cook. If you want more information about today's episode, check out our show notes at www.shoulderphysio.com. If you liked what you heard today, don't forget to follow and subscribe on your podcast player of choice and leave a rating or review. It really does help the show reach more people. Thanks for listening. I'll chat to you soon. The Shoulder Physio Podcast would like to acknowledge that this episode was recorded from the lands of the Terribalang people. I also acknowledge the traditional custodians of the lands on which each of you are living, learning, and working from every day. I pay my respects to elders past, present, and emerging and celebrate the diversity of Aboriginal and Torres Strait Islander peoples and their ongoing cultures and connections to the lands and waters of Australia.
