The wide variability of bodies and injuries is a great illustration of heterogenous treatment effects, and why estimating these "scientifically" is probably impossible in this case.
Body types and injury types plausibly interact with the treatment, but this Gelman post from years back (https://statmodeling.stat.columbia.edu/2018/03/15/need16/) shows, with some assumptions, that you need about 16 times the sample size to estimate an interaction than the main effect. Higher order interactions would blow this up even more. Given the already tiny sample sizes (31 people on the study you cited on Monday's post), there probably won't ever be enough data to know if something works given the widely accepted practices, even assuming a perfectly run experiment.
I'm loving your discussion of comparing the general (RCT) with the personal. I think about a lot of the same issues with trying to reconcile the fact that my PhD topic (causal modelling, causal inference) which seems like it should universally help reasoning is actually hard to apply to my own health
You have focused on the experience of craft and personalization side of PT.
However, one seems to need something more for the PT standard of treatment to progress.
Here's one example:
- For frozen shoulder (resulting from a lack of fluid between the cartilage sheath and the bone) the primary treatment method until relatively recently had been extended PT sessions, typically weekly or biweekly over 10-12 months that focused on slowly regaining range of motion by stretching the cartilage to "unbind" it, so that fluid can enter between the cartilage and bone.
- A newer treatment, fluid injection directly between the cartilage sheath and the bone, often only one shot, can completely relieve the issue over the course of a few weeks. It is a huge savings in quality of life for the patient, and removes the time and expense of a long series of PT sessions - which of course is a lot less work for PTs.
What is the process by which a personalized craft-focused field develops these insights? I don't see how to do it without something external, that is then argued using the kinds of methods you have been suggesting are over-glorified... I gather that this treatment was pretty controversial until there were RCTs showing its value.
Can you point me to the studies you are thinking about here with respect to frozen shoulder? It’s my understanding that fluid injection consistently tests worse than steroids + PT, which is a low bar to clear given the nice simplicity of the hydrodilatation intervention. See, this review, for example: https://academic.oup.com/bmb/article/147/1/121/7231525
Physical therapy is essentially RCT on yourself. I am an avid breakdancer and was an avid olympic-style weightlifter (even competed this in college). I have had tendonitis on my knee, herniated disc, wrist tendonitis, tennis elbow, sprained elbow, sprained wrist, strained neck muscles, ankle impingement… all from weightlifting and breakdance. By now, I know my body the best and generally follow a similar routine because I have gotten injured so often. These injuries are all soft tissue injuries and taking ibuprofen will do nothing to resolve the core issue with it. And I have done enough RCT on myself to resolve those injuries to know what muscles need to be trained such that I can avoid those injuries again and keep training. How to go slower when I need to. And how to rest. In general, if the joint can be moved, it’s likely not something that a doctor can do anything about except give you ibuprofen and tell you to rest (because it’s hopefully nothing catastrophic like a tissue tear or a fracture).
Also a random plug, but if a PT is too expensive, you can do your own research and RCT some treatments on yourself assuming the injury isnt catastrophic. There’s tons of resources online.
The wide variability of bodies and injuries is a great illustration of heterogenous treatment effects, and why estimating these "scientifically" is probably impossible in this case.
Body types and injury types plausibly interact with the treatment, but this Gelman post from years back (https://statmodeling.stat.columbia.edu/2018/03/15/need16/) shows, with some assumptions, that you need about 16 times the sample size to estimate an interaction than the main effect. Higher order interactions would blow this up even more. Given the already tiny sample sizes (31 people on the study you cited on Monday's post), there probably won't ever be enough data to know if something works given the widely accepted practices, even assuming a perfectly run experiment.
I'm loving your discussion of comparing the general (RCT) with the personal. I think about a lot of the same issues with trying to reconcile the fact that my PhD topic (causal modelling, causal inference) which seems like it should universally help reasoning is actually hard to apply to my own health
That reminds me of some animal /human behavior research. I am not sure how statistical/quantitative methods should be applied in those cases. For example, mice can do CPR on their companion https://www.science.org/content/article/scienceadviser-mice-employ-aggressive-first-aid-resuscitate-unresponsive-companions . I think the tendency of applying those methods is correlated with "dehumanization". How do they reproduce those kind of research ?
You have focused on the experience of craft and personalization side of PT.
However, one seems to need something more for the PT standard of treatment to progress.
Here's one example:
- For frozen shoulder (resulting from a lack of fluid between the cartilage sheath and the bone) the primary treatment method until relatively recently had been extended PT sessions, typically weekly or biweekly over 10-12 months that focused on slowly regaining range of motion by stretching the cartilage to "unbind" it, so that fluid can enter between the cartilage and bone.
- A newer treatment, fluid injection directly between the cartilage sheath and the bone, often only one shot, can completely relieve the issue over the course of a few weeks. It is a huge savings in quality of life for the patient, and removes the time and expense of a long series of PT sessions - which of course is a lot less work for PTs.
What is the process by which a personalized craft-focused field develops these insights? I don't see how to do it without something external, that is then argued using the kinds of methods you have been suggesting are over-glorified... I gather that this treatment was pretty controversial until there were RCTs showing its value.
Can you point me to the studies you are thinking about here with respect to frozen shoulder? It’s my understanding that fluid injection consistently tests worse than steroids + PT, which is a low bar to clear given the nice simplicity of the hydrodilatation intervention. See, this review, for example: https://academic.oup.com/bmb/article/147/1/121/7231525
Physical therapy is essentially RCT on yourself. I am an avid breakdancer and was an avid olympic-style weightlifter (even competed this in college). I have had tendonitis on my knee, herniated disc, wrist tendonitis, tennis elbow, sprained elbow, sprained wrist, strained neck muscles, ankle impingement… all from weightlifting and breakdance. By now, I know my body the best and generally follow a similar routine because I have gotten injured so often. These injuries are all soft tissue injuries and taking ibuprofen will do nothing to resolve the core issue with it. And I have done enough RCT on myself to resolve those injuries to know what muscles need to be trained such that I can avoid those injuries again and keep training. How to go slower when I need to. And how to rest. In general, if the joint can be moved, it’s likely not something that a doctor can do anything about except give you ibuprofen and tell you to rest (because it’s hopefully nothing catastrophic like a tissue tear or a fracture).
Also a random plug, but if a PT is too expensive, you can do your own research and RCT some treatments on yourself assuming the injury isnt catastrophic. There’s tons of resources online.