Average Progression rates

[Pooka1]

I was referring to success in the short-term. Kids with 30 degree curves usually get great correction if the brace is fitted well. I'm not talking about once brace treatment is discontinued, or success in the long-term.

--Linda

Ah okay. I see. We need studies on the mid-term and out years for everyone, treated and untreated.

[hdugger]

In the case of scoliosis bracing, being familiar with the literature is synonymous with realizing how much is unknown still. Claims that more is known than has been shown is to be unfamiliar with the literature.

I'm still not clear why that would be useful.

So, a doctor does a bunch of bracing and has his/her own impressions of the results - seems to work with these kids, not so much with these, and so on and so on. IMO, that working knowledge comprises almost the whole ground of what is known and unknown. The literature contributes very little knowledge, and likely will go on contributing very little knowledge. IMO, what the literature can do is slightly inform the anecdotal knowledge. But it in no way contributes "knowledge."

So, what's "known" lives almost entirely in the anecdotal realm. This discussion keeps coming up with the bracing literature, but it's equally relevant to just about everything concerning scoliosis. What cause PJK, for example, and how can it be avoided? The literature is all over the map - my only hope is that our surgeon can sort it out based on his experience. Should people have physical therapy right after surgery? I think the literature is mute on this topic - have any studies been done? - but doctors seem pretty certain. I'm guessing this comes from some oral history of trial and error which we're not privvy to.

You sometimes discuss the failings in the bracing literature as if it means anything at all about bracing. But, the exact same failings are seen in the research supporting almost every decision related to scoliosis. Either we can rely on the people who actually work with patients every day to sort their way through what they see and draw conclusion about our treatment, or, really, we're just totally sunk. Science - at least the perfect science that you rely on - is almost completely mute and of no help at all.

[Pooka1]

I'm still not clear why that would be useful.

So, a doctor does a bunch of bracing and has his/her own impressions of the results - seems to work with these kids, not so much with these, and so on and so on. IMO, that working knowledge comprises almost the whole ground of what is known and unknown. The literature contributes very little knowledge, and likely will go on contributing very little knowledge. IMO, what the literature can do is slightly inform the anecdotal knowledge. But it in no way contributes "knowledge."

So, what's "known" lives almost entirely in the anecdotal realm. This discussion keeps coming up with the bracing literature, but it's equally relevant to just about everything concerning scoliosis. What cause PJK, for example, and how can it be avoided? The literature is all over the map - my only hope is that our surgeon can sort it out based on his experience. Should people have physical therapy right after surgery? I think the literature is mute on this topic - have any studies been done? - but doctors seem pretty certain. I'm guessing this comes from some oral history of trial and error which we're not privvy to.

You sometimes discuss the failings in the bracing literature as if it means anything at all about bracing. But, the exact same failings are seen in the research supporting almost every decision related to scoliosis. Either we can rely on the people who actually work with patients every day to sort their way through what they see and draw conclusion about our treatment, or, really, we're just totally sunk. Science - at least the perfect science that you rely on - is almost completely mute and of no help at all.

Anecdote gathered wihtout controls is as useless as the literature without controls. Some surgeons may not accept this.

[hdugger]

Anecdote gathered wihtout controls is as useless as the literature without controls. Some surgeons may not accept this.

Unless they want to put down their scalpels, they really have no choice. How are they choosing which levels to fuse? Deciding when the sagittal balance is correct? Like it or not, it's not strictly controlled scientific studies which are forming the bases of those decisions. Surgeons are not, as you suggest for bracing studies, dividing their patients by initial curve, marfans symptoms, type of curve, etc and trying out randomly assigned fusion levels across those groups. Instead, they're looking at each individual and figuring out how to precisely align and fuse their spine based largely on anecdotal knowledge and trial and error(!).

I suspect that surgeons are pretty comfortable with the "informed intuition" that forms the basis of most scoliosis decisions. Or they'd find some more "known" area of science to settle into. And I also suspect that they feel that they "know" these things (levels to fuse, etc) just as well as they "know" whether to brace or not, even though their knowledge is not largely based on rigorous scientific studies.

[Pooka1]

Well, with surgery, they get feedback in the form of patient data. For example, if they choose a lowest instrumented level using Method A, they will quickly amass data on if that method produces stability and balance going forward. This doesn't mean that there is only one "best" approach only that there is at least one.

One potential problem here is the pediatric guys who I don't think follow patients out past 18 but rather hand them off. These guys have to keep data themselves on their patients and analyze it. I think they all do this. And they have to confer with colleagues on their data, usually at meetings.

In the case of bracing, they can't know in principle if most of their braced patients weren't unnecessarily braced. As someone said recently, the actual successes are swamped out by the apparent successes of curves that wouldn't have progressed anyway. This appears to be the majority of braced kids.

All they can note is the bracing failures but even in those cases, they really don't have any handle on compliance nor do they necessarily spot all connective tissue patients. So many of those "failures" are only apparent and might just be artifacts of compliance and condition.

So with bracing, they are dealing with a huge false positive rate and also a false negative rate. Bracing is simply NOT like surgery where you can document all results and therefore have a much better sense of what's what, albeit entirely empirically. There is no comparison that I can see in terms of knowing outside controlled studies.

[LindaRacine]

There's a basic problem with the way kids with scoliosis are treated is that the vast majority are treated by pediatricians, who do not see them after around the age of 18.

[Pooka1]

There's a basic problem with the way kids with scoliosis are treated is that the vast majority are treated by pediatricians, who do not see them after around the age of 18.

Excellent point.

[hdugger]

So with bracing, they are dealing with a huge false positive rate and also a false negative rate. Bracing is simply NOT like surgery where you can document all results and therefore have a much better sense of what's what, albeit entirely empirically. There is no comparison that I can see in terms of knowing outside controlled studies.

I'm just saying that their method of "knowing" is the same in both cases - they do something on an individual and they look at the results. That's in contrast to the idea that "knowing" is only available through controlled research studies. Were that the case, surgeons couldn't "know" anything.

As to seeing the results, it's somewhat similar to bracing in that they see the immediate results of the surgery but don't have a clear sense of how their surgical choices map to long-term life quality. How did their particular choice of fusion level/method turn out 5 (or even 2) years down the road? I suspect they don't know enough about that, even anecdotally, and I'm certain they don't know it through well-controlled scientific studies.

So, to draw the parallel

- surgeons don't know whether bracing maps to patients final curve when the brace is removed (the important outcome for bracing)
- surgeons don't know how their choice of levels fused (or any other surgical decision) maps to quality of life (the important outcome for surgery)

In both cases, they have no solid research to guide their most critical decisions.

[LindaRacine]

I'm just saying that their method of "knowing" is the same in both cases - they do something on an individual and they look at the results. That's in contrast to the idea that "knowing" is only available through controlled research studies. Were that the case, surgeons couldn't "know" anything.

As to seeing the results, it's somewhat similar to bracing in that they see the immediate results of the surgery but don't have a clear sense of how their surgical choices map to long-term life quality. How did their particular choice of fusion level/method turn out 5 (or even 2) years down the road? I suspect they don't know enough about that, even anecdotally, and I'm certain they don't know it through well-controlled scientific studies.

So, to draw the parallel

- surgeons don't know whether bracing maps to patients final curve when the brace is removed (the important outcome for bracing)
- surgeons don't know how their choice of levels fused (or any other surgical decision) maps to quality of life (the important outcome for surgery)

In both cases, they have no solid research to guide their most critical decisions.

For the most part, I agree with this. But, non-pediatric surgeons have been doing some (though not enough) long-term followup on surgical patients. As far as I know, no one other than the Swedish group has done any long-term peds followup (surgical or non-surgical).

--Linda

[hdugger]

My sense is that the major weakness in scoliosis surgery research isn't the lack of long-term follow-up (although that's pretty bad!), but the lack of a meaningful outcome measure.

<rant on>

Surgeons tend to fall back on the Cobb angle, I assume because it's so gosh-darned easy to measure. But just reading this forum, I've seen very little correlation between Cobb angle and quality of life in adult patients. Although the Cobb angle seems to be pretty good for measuring outcome in kids, it just seems to fall apart completely for adults. There has to be some better way of getting a complete picture of an adult - how are they balanced, overall? did their initial complaints resolve? did they end up with new pains? are they able to get back to what they used to do? The survey measures just don't seem accurate or complete enough. There needs to be some combination of actually *looking* at them and judging their overall balance, and then some set of physical tests they try to complete. Without that, you're stuck judging outcome from a single number which really says nothing at all about the things you're interested in.

<rant over>

[Pooka1]

I'm just saying that their method of "knowing" is the same in both cases - they do something on an individual and they look at the results. That's in contrast to the idea that "knowing" is only available through controlled research studies. Were that the case, surgeons couldn't "know" anything.

No I think there is a real difference here. Surgeons have a certain small number of "methods" for doing different things (like identifying the lowest instrumented vertebrae) and then someone like Lenke can amass huge numbers of patients in a short time who were fused with that method. He knows the results (balance) very soon afterwards but can't know about things like pseudoarthroses right away. Things like adjacent level disease down the road are not complete black boxes; something is known about how low or high you fuse, balance and intrinsic bone properties drive the likelihood of that outcome.

As to seeing the results, it's somewhat similar to bracing in that they see the immediate results of the surgery but don't have a clear sense of how their surgical choices map to long-term life quality. How did their particular choice of fusion level/method turn out 5 (or even 2) years down the road? I suspect they don't know enough about that, even anecdotally, and I'm certain they don't know it through well-controlled scientific studies.

I think folks who do a million surgeries like Lenke and Bridwell know more about the particulars than you think. And they are publishing them, albeit relatively slowly. Surgery, because the outcome seems to be largely driven by variables under the surgeon's control like how to select LIV, is more tractable because you can simply hit it with so many patients while holding other variables constant that an empirical answer emerges in a given, well-defined surgical scenario. You can take a million 60* Lenke Type 2b curves and chose one method for something and see what happens.

With bracing, there are too many variables not under the surgeon's control and known huge false positive rates that are in principle not identifiable empirically. And then there is the unknowable false negative rate which I suspect exists but may not. It seems very few things can be held constant while varying others in bracing studies.

So, to draw the parallel

- surgeons don't know whether bracing maps to patients final curve when the brace is removed (the important outcome for bracing)
- surgeons don't know how their choice of levels fused (or any other surgical decision) maps to quality of life (the important outcome for surgery)

In both cases, they have no solid research to guide their most critical decisions.

Again, I think more is known about surgical outcomes than you think. For example I think they know good balance in all planes usually maps to a good result radiographically and quality of life-wise and in the long term, at least for T fusions. Equally, they know fusing into the lumbar with H rods can be problematic. Pedicle screws aim in part to solve those problems. It is a constant honing and perfecting using the previous generation instrumentation. Surgery, unlike bracing, is a very long process of perfecting past technique and constantly seeking to build a better mouse trap. Although of course it started some where out of nowhere but that was a long time ago and it has gotten down the road in many ways. And disastrous techniques are identified and abandoned.

In contrast, we see bracing going off in new directions (e.g., Spinecor) because there is no there there to build on because of the huge false positive and potential false negative rates. It is intrinsically harder to study than surgery because you don't really have a clear shot at holding at least some variables constant while varying one. The combination of patient plus brace seems far wider than the combination of patient plus fusion with kids. The results of bracing are all over the map and known to be incorrect at times whereas the fusions in kids tend to group pretty well and positively, at least in the short to mid term. Adult fusion is more all over the map but still probably groups more positively than bracing when correcting for false positives in the bracing.

[hdugger]

We're really just trying to pin down the definition of "known." So, when you say:

"Again, I think more is known about surgical outcomes than you think. For example I think they know good balance in all planes usually maps to a good result radiographically and quality of life-wise and in the long term, at least for T fusions. Equally, they know fusing into the lumbar with H rods can be problematic. Pedicle screws aim in part to solve those problems. It is a constant honing and perfecting using the previous generation instrumentation."

you're talking about knowing as a kind of trial-and-error anecdotal process, guided mostly by surgeon's shared experience and informed intuition.

That's really all I'm getting at. It's *all* that kind of knowledge - none of it is what anyone would call clearly researched and replicated.

I understand your "black box" argument about bracing, but I'd suggest that surgeons are tracking their bracing results intuitively in a way that isn't represented in the literature, in much the same way that they track their surgical results. I suspect it's that shared intuition that serves as the basis of knowledge about the kinds of patients who benefit from bracing. So, although they don't understand the inner workings of the box, they do see into it well enough to recognize the kinds of patients who aren't benefiting from bracing. If it were entirely a black box, they'd be bracing every patient who comes in.

I'm not necessarily pro-bracing (nor necessarily pro-surgery ), but I don't think the arguments against it involves the lack of clear scientific proof. There isn't clear scientific proof for any of this stuff.

[Pooka1]

I understand your "black box" argument about bracing, but I'd suggest that surgeons are tracking their bracing results intuitively in a way that isn't represented in the literature, in much the same way that they track their surgical results. I suspect it's that shared intuition that serves as the basis of knowledge about the kinds of patients who benefit from bracing. So, although they don't understand the inner workings of the box, they do see into it well enough to recognize the kinds of patients who aren't benefiting from bracing. If it were entirely a black box, they'd be bracing every patient who comes in.

No, intuition about what works and doesn't work about bracing is not useful because there is no way in principle to intuit when you have a huge false positive and potential false negative rate and no way to figure out when a patient is in one of those categories. This is where intuition is known to fail a person. It can't be used. There can never be even empirical (true) data in this situation.

Intuition is much less important in surgery now that there is a huge mass of empirical data on the relatively few methods and combinations of methods to skin the cat. And because it builds on itself.

I guess that is the point... bracing relies on intuition which must fail in that particular case and surgery relies on tons of empirical data where you can hold some things constant while varying others. Because of this, surgery is simply more tractable and therefore more can be known outside controlled studies.

[Ballet Mom]

I understand your "black box" argument about bracing, but I'd suggest that surgeons are tracking their bracing results intuitively in a way that isn't represented in the literature, in much the same way that they track their surgical results. I suspect it's that shared intuition that serves as the basis of knowledge about the kinds of patients who benefit from bracing. So, although they don't understand the inner workings of the box, they do see into it well enough to recognize the kinds of patients who aren't benefiting from bracing. If it were entirely a black box, they'd be bracing every patient who comes in.

I think you're absolutely right hdugger. When my daughter was diagnosed with a 35 degree curve, I thought the surgeon was nuts to have prescribed her a Charleston Bending Brace after I read the research studies on it. The TLSO braces worn continuously clearly get better results on larger curves than the nighttime braces and 35 degrees was the maximum size of curve recommended for this brace.

But in retrospect, I think the surgeon was simply highly skilled and attuned to my daughter's physical traits. He knew she was very thin, and was a ballerina and therefore likely very flexible, which he probably could figure out in the exam anyway, and he encouraged her to continue with the ballet. He also told me it was the brace he would use on his daughter if she had scoliosis, so he had obviously had good experience with it, at least for kids with the same physical attributes, and felt it had a good risk/reward trade-off.

[Ballet Mom]

For the most part, I agree with this. But, non-pediatric surgeons have been doing some (though not enough) long-term followup on surgical patients. As far as I know, no one other than the Swedish group has done any long-term peds followup (surgical or non-surgical).
--Linda

I think it's amazing the research studies that are relied on for scoliosis. I know it must be terribly difficult to get the numbers of patients needed for a good study, but when you realize that the progression numbers for those patients in the great Weinstein/Poncetti study "Curve Progression in Idiopathic Scoliosis" for those with thoracic curves from 50 to 75 degrees is based on eleven patients, it makes me wince. Surely they could be doing more studies following more patients over time to see how they do, both those who choose surgery and those who don't.

Thank goodness for the Swedes!