Announcement

Collapse
No announcement yet.

General figure question for Dr. McIntire

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • General figure question for Dr. McIntire

    I finally found the entire Katz et al., 2010 article (http://www.ncbi.nlm.nih.gov/pubmed/20516309) and was just scanning it before running off to work. I'll look at it in depth later but I was struck by Figs. 4, 5, and 6 which only present the trend lines and not the data points.

    I don't think 100 points is so much that they can't show it on one (or a few) figures... if you reduce the size of the symbol then it can easily be done. It just seems to me that only presenting the trend line will hide important facts about the variance and imply the predictive ability is far better than it is. There is a reason why we don't just report means and why we report std also as you know.

    I recently plotted many hundreds of data points (in five series) on a graph (not meant for publication) to find a problem in an analytical method and also included trend lines. The trend lines were only one part of the story... the other was the HUGE variation in the parameter I was investigating that crunched down into those trend lines. In other words, you could never use that trend line to predict anything although it was clearly trending one way.

    I think if you are going to only show trend lines you need to show confidence envelopes. Just based on the one bar graph in the appendix, the mean is not the message, the variation is.

    And speaking of mean, do you think they should have shown median instead? Certain data I deal with are usually presented as medians and not means and I wonder if this is the case with these data. I wonder if they would have come to the same conclusions if they plotted median. In the scoliosis game, it seems the outliers are common which would drive selection of the median over the mean.

    Last, it seems like the kids who were braced before the growth spurt had more progression than the kids braced after. I am not understanding why this result doesn't undermine the claim that bracing is only or most effective during the growth spurt.

    I would like to see the individual data of total brace wear versus progression. I think it would show that some (many?) very compliant kids had progression and some (many?) non-compliant kids did not. It would almost have to given those error bars.

    One result that will stand in all these studies is the extreme variability of response to conservative treatment even after correcting for compliance as in this case. There might be a signal buried in there by it may be impossible to find using any methodological approach.
    Sharon, mother of identical twin girls with scoliosis

    No island of sanity.

    Question: What do you call alternative medicine that works?
    Answer: Medicine


    "We are all African."

  • #2
    Actually Sharon, my biggest problem with this paper is Figure 3. The plot percent of patients in each of their respective groups (>12 hours etc). This doesn't say anything about the relationship between groups or number of patients in each group.

    It is extremely misleading.

    Also, as an aside. . .I have a good friend who is a physicist and he is constantly and considerably frustrated by statistical data for people. It just isn't clean enough. The very uniqueness of an individual makes it very hard to get clear cut results. A model has to be very specific and the subject population very large to get a reliable result.

    p

    Comment


    • #3
      Originally posted by PNUTTRO View Post
      Actually Sharon, my biggest problem with this paper is Figure 3. The plot percent of patients in each of their respective groups (>12 hours etc). This doesn't say anything about the relationship between groups or number of patients in each group.

      It is extremely misleading.
      I agree. My eye glossed over that figure (I dislike bar graphs) and I suspected I would have a lot to say once I reviewed the article.

      What we have here in my opinion is that there is more than one valid way to show the data but that the different valid ways will (mis)lead to different conclusions. I will see if that is the case.

      The biggest potential issue that popped for me was means versus medians. That is a potentially a game changer.

      Boy would I LOVE to get the raw data on this one.

      Also, as an aside. . .I have a good friend who is a physicist and he is constantly and considerably frustrated by statistical data for people. It just isn't clean enough. The very uniqueness of an individual makes it very hard to get clear cut results. A model has to be very specific and the subject population very large to get a reliable result.

      p
      Yes exactly. I think these researchers should be using individual based models which I haven't used myself but what little I know about them is that they might be more appropriate than averaging what is known (or highly suspected) to be apples and oranges.

      I think if they just plotted the trajectories over time of all 100 patients (as very thin lines on a large graph) instead of just presenting the statistics, this point would become very clear. Even if the means are different to some specific point, if the variation is huge then you still have no predictive power in the end. I would not use these data to try to convince my kid to wear her brace more hours but YMMV.

      ETA: I would also like to see a tally of the number of very compliant kids who needed fusion and a tally of non-compliant kids who did not need fusion. I think there might be some good information in that in terms of saying:

      1. who is not likely going to be helped by bracing as opposed to trying to show who is most likely to benefit.

      2. who is most likely not to progress no matter what you do or don't do.

      Then compare that to what is already known and see if it agrees. If it doesn't then there is something different between this (small) patient population and the small patient populations in the other studies which is just to say this is a highly variable condition and that 100 patients is too small a study.
      Last edited by Pooka1; 08-12-2010, 06:01 PM. Reason: typos
      Sharon, mother of identical twin girls with scoliosis

      No island of sanity.

      Question: What do you call alternative medicine that works?
      Answer: Medicine


      "We are all African."

      Comment


      • #4
        Originally posted by Pooka1 View Post
        I finally found the entire Katz et al., 2010 article (http://www.ncbi.nlm.nih.gov/pubmed/20516309) and was just scanning it before running off to work. I'll look at it in depth later but I was struck by Figs. 4, 5, and 6 which only present the trend lines and not the data points.
        I won't try to defend their statistics but my interpretation of that data was that they were just showing the basic trend line of their data. Personally, I don't read it as them trying to be absolutely predictive but rather generally predictive. That is, when they speak with patients that they know aren't going to wear their brace for the 23 or 16 hours of prescribed wear, they can maybe say "If you only wear your brace XX hours a day, try to wear it during this time frame". Overall these trend lines just seem to be interesting things to look at in future bracing studies. Kind of like a proof of concept type of statistic. It doesn't strengthen or weaken the paper for me. Just an interesting way to stratify their data.

        Originally posted by Pooka1 View Post
        And speaking of mean, do you think they should have shown median instead? Certain data I deal with are usually presented as medians and not means and I wonder if this is the case with these data. I wonder if they would have come to the same conclusions if they plotted median. In the scoliosis game, it seems the outliers are common which would drive selection of the median over the mean.
        Comparing two groups they used a t-test with unequal variances assumed. I know for some of these types of non-parametric statistics the variance is calculated from the median. I'm unfamiliar with the t-test version but have worked with non-parametric ANOVA's as well as parametric tests that violate some of the assumptions made and the median is frequently used or there is some other type of adjustment made. It reads like they did a straight ANOVA, which if the groups had unequal variance (like the t-test) that would be an inappropriate statistic. That's not clear and is a bit if-y for me.

        Originally posted by Pooka1 View Post
        Last, it seems like the kids who were braced before the growth spurt had more progression than the kids braced after. I am not understanding why this result doesn't undermine the claim that bracing is only or most effective during the growth spurt.
        That is correct that kids before their growth spurt progressed more. I don't see it as going against their claim because the compliant braced group that had a growth spurt still had less progression than the non-compliant group. Also, overall there were fewer progressions in more skeletally mature patients. So I think they are acknowledging the risk for progression is lower in a more skeletally mature patient.

        Originally posted by Pooka1 View Post
        I would like to see the individual data of total brace wear versus progression. I think it would show that some (many?) very compliant kids had progression and some (many?) non-compliant kids did not. It would almost have to given those error bars.
        Table 2 can show you some of that information. I don't think it would be many but certainly what you're suggesting would be the case for some of them. Also the discussion says that only one of their compliant risser 0 patients experienced progression.

        Originally posted by Pooka1 View Post
        One result that will stand in all these studies is the extreme variability of response to conservative treatment even after correcting for compliance as in this case. There might be a signal buried in there by it may be impossible to find using any methodological approach.
        The way I read and approach these types of studies is with the assumption that nothing is going to work 100% of the time. There will always be large variance because the range of curves can be so large. Those that progress will end up with a 60° curve. So when you average that number in to a group, whose average is 35°, it'll shoot that variance through the roof. So I think your point about using a different type of statistic is a good one. Something based on the median would seem to be appropriate.

        Comment


        • #5
          Originally posted by skevimc View Post
          I won't try to defend their statistics but my interpretation of that data was that they were just showing the basic trend line of their data.
          Okay but I would like to send you that graph I referred to which had clear trends for all five analytes and yet nearly the entire graph space was filled with individual points. The predictive power is near zero. I haven't studied the article yet but I certainly hope they report the r^2 for the trend line fit. I'll bet it is much closer to zero than it is to one.

          Personally, I don't read it as them trying to be absolutely predictive but rather generally predictive.
          Yes of course you don't read it that way but almost every single parent will read it that way and we have seen some of that already. That's why I kept banging on about exactly what surgeons will say differently to their patients about bracing efficacy after this study compared to before.

          That is, when they speak with patients that they know aren't going to wear their brace for the 23 or 16 hours of prescribed wear, they can maybe say "If you only wear your brace XX hours a day, try to wear it during this time frame". Overall these trend lines just seem to be interesting things to look at in future bracing studies. Kind of like a proof of concept type of statistic. It doesn't strengthen or weaken the paper for me. Just an interesting way to stratify their data.
          I'd like to see the paper stratified other ways just to see if the conclusions hold up. They might.

          Comparing two groups they used a t-test with unequal variances assumed. I know for some of these types of non-parametric statistics the variance is calculated from the median. I'm unfamiliar with the t-test version but have worked with non-parametric ANOVA's as well as parametric tests that violate some of the assumptions made and the median is frequently used or there is some other type of adjustment made. It reads like they did a straight ANOVA, which if the groups had unequal variance (like the t-test) that would be an inappropriate statistic. That's not clear and is a bit if-y for me.
          I know only a minimum of statistics so I can't really comment other than the issue of using medians instead of means if outliers are seen as is likely the case.

          That is correct that kids before their growth spurt progressed more. I don't see it as going against their claim because the compliant braced group that had a growth spurt still had less progression than the non-compliant group. Also, overall there were fewer progressions in more skeletally mature patients. So I think they are acknowledging the risk for progression is lower in a more skeletally mature patient.
          I think we need to know exactly how large each of these sub-groups is.

          Table 2 can show you some of that information. I don't think it would be many but certainly what you're suggesting would be the case for some of them. Also the discussion says that only one of their compliant risser 0 patients experienced progression.
          How many compliant Risser kids were there total? Don't answer that! Maybe the paper says... haven't studied it yet.

          The way I read and approach these types of studies is with the assumption that nothing is going to work 100% of the time. There will always be large variance because the range of curves can be so large. Those that progress will end up with a 60° curve. So when you average that number in to a group, whose average is 35°, it'll shoot that variance through the roof. So I think your point about using a different type of statistic is a good one. Something based on the median would seem to be appropriate.
          Depending on the number and magnitude of the outliers, I am guessing using medians vice means might have lead them to different conclusions. Can't know without seeing the raw data. Since the study was self-funded, there is no obligation to publish or otherwise make available the raw data.
          Sharon, mother of identical twin girls with scoliosis

          No island of sanity.

          Question: What do you call alternative medicine that works?
          Answer: Medicine


          "We are all African."

          Comment


          • #6
            Originally posted by Pooka1 View Post
            Okay but I would like to send you that graph I referred to which had clear trends for all five analytes and yet nearly the entire graph space was filled with individual points. The predictive power is near zero. I haven't studied the article yet but I certainly hope they report the r^2 for the trend line fit. I'll bet it is much closer to zero than it is to one.
            OIC what you're saying. Agreed. I actually hadn't even thought about it in that way (because I didn't think of it as anything more than just an interesting tid-bit). But yes, I hate correlations that, while significant, have an r value of anything < .6 or .7. Studies will put forward data that reaches significance but is low to moderately correlated. I actually saw someone present an r-value of .2. When someone questioned him on that during the q&a he said "yeah, but it was significant". To quote Inigo Mantoya "You keep using that word. I do not think it means what you think it means".

            Originally posted by Pooka1 View Post
            Yes of course you don't read it that way but almost every single parent will read it that way and we have seen some of that already. That's why I kept banging on about exactly what surgeons will say differently to their patients about bracing efficacy after this study compared to before.
            I understand what you're saying. And just knowing how MD's work in the clinic across multiple fields, I doubt it will change much of anything in how they prescribe bracing or how they instruct. (Aside from the doctors at TSRH). If a doctor likes a particular type of brace they are going to stick with it. Like I've said before, for most people, papers that support their bias tend to be read more favorably. Papers that go against their bias tend to be discredited more easily. It takes an enormous study to significantly change the way doctors practice. There's still school screening even though it's been shown that it doesn't change the number of kids that go to surgery.


            Originally posted by Pooka1 View Post
            I'd like to see the paper stratified other ways just to see if the conclusions hold up. They might.
            I think another big thing with this paper was showing the proof of concept that implanting heat sensors in the brace is easy to do and gets reliable compliance results. This has always been a HUGE issue with bracing effectiveness. Now studies and doctors can monitor brace wear and be better equipped to say that the brace is working or it isn't working. As a treatment, this is HUGE. From a research point of view, this is also HUGE. Compliance is the thorn in every clinical trials side. If it proves not to be effective the argument is "well maybe they didn't do the treatment". If it proves to be effective the argument is "well how compliant were they. Maybe they wouldn't have gotten worse anyway". This concpet can begin to answer some of those questions and we might actually start getting an answer to the question of "is bracing effective". Right now, it's kind of like a new research toy and they are seeming to say "look at the kind of ways we can analyze data now that we know when they wear their brace." No doubt studies in the future will stratify results in several different categories and will continue to fine tune or refute the results presented here.


            Originally posted by Pooka1 View Post
            Depending on the number and magnitude of the outliers, I am guessing using medians vice means might have lead them to different conclusions. Can't know without seeing the raw data. Since the study was self-funded, there is no obligation to publish or otherwise make available the raw data.
            It's possible their results could have been different. The type of statistics used in all scoliosis papers is definitely a good point. Particularly when discussing anything that will have a huge range. The data is most likely skewed in one direction which would make most parametric tests inappropriate depending on how skewed it is. There are decent alternatives that use the median or rank.

            Comment


            • #7
              Originally posted by skevimc View Post
              OIC what you're saying. Agreed. I actually hadn't even thought about it in that way (because I didn't think of it as anything more than just an interesting tid-bit).
              Yes, again, you know how to interpret these things but the vast majority of parents out there do not. That is the danger in some of these papers, especially this one.

              But yes, I hate correlations that, while significant, have an r value of anything < .6 or .7. Studies will put forward data that reaches significance but is low to moderately correlated. I actually saw someone present an r-value of .2. When someone questioned him on that during the q&a he said "yeah, but it was significant". To quote Inigo Mantoya "You keep using that word. I do not think it means what you think it means".
              When should the public slappings start... claiming significance <0.4? How about <0.3? At some point, all we can do is snicker to ourselves and our colleagues.

              Yes I agree >0.6 (or so) is something to talk about. Actually I would say >0.7. Anything less that 0.6 is just ridiculous. They could have calculated an r^2 for those lines. If they didn't then there is likely a reason. Looking at the bar chart, I think I know what that reason is.

              I understand what you're saying. And just knowing how MD's work in the clinic across multiple fields, I doubt it will change much of anything in how they prescribe bracing or how they instruct. (Aside from the doctors at TSRH).
              I agree. But this point will be entirely lost on the great run of parents who hear about this study.

              If a doctor likes a particular type of brace they are going to stick with it.
              I agree. I hope in the future, we will hear about what surgeons say if asked about this study to see if they changed anything in how they approach cases.

              Like I've said before, for most people, papers that support their bias tend to be read more favorably. Papers that go against their bias tend to be discredited more easily. It takes an enormous study to significantly change the way doctors practice. There's still school screening even though it's been shown that it doesn't change the number of kids that go to surgery.
              Do you think BrAIST is big enough to change the approach of any surgeon if the results don't support bracing over watching/waiting to avoid surgery?

              I think another big thing with this paper was showing the proof of concept that implanting heat sensors in the brace is easy to do and gets reliable compliance results. This has always been a HUGE issue with bracing effectiveness. Now studies and doctors can monitor brace wear and be better equipped to say that the brace is working or it isn't working. As a treatment, this is HUGE. From a research point of view, this is also HUGE. Compliance is the thorn in every clinical trials side. If it proves not to be effective the argument is "well maybe they didn't do the treatment". If it proves to be effective the argument is "well how compliant were they. Maybe they wouldn't have gotten worse anyway". This concept can begin to answer some of those questions and we might actually start getting an answer to the question of "is bracing effective". Right now, it's kind of like a new research toy and they are seeming to say "look at the kind of ways we can analyze data now that we know when they wear their brace." No doubt studies in the future will stratify results in several different categories and will continue to fine tune or refute the results presented here.
              Yes it's a good first step. I think BrAIST is using heat sensors also.

              It's possible their results could have been different. The type of statistics used in all scoliosis papers is definitely a good point. Particularly when discussing anything that will have a huge range. The data is most likely skewed in one direction which would make most parametric tests inappropriate depending on how skewed it is. There are decent alternatives that use the median or rank.
              Yes if there is going to be a skew then you might need one-tailed tests or some such. I hate statistics and tend to stick with very basic summary descriptors of the data. And I try to always show the data, especially if I will be showing a trend line. I don't think I have ever shown just the trend line, and certainly not without an r^2.
              Sharon, mother of identical twin girls with scoliosis

              No island of sanity.

              Question: What do you call alternative medicine that works?
              Answer: Medicine


              "We are all African."

              Comment


              • #8
                Originally posted by Pooka1 View Post
                I agree. But this point will be entirely lost on the great run of parents who hear about this study.
                I'm biased towards conservative management options (even though I also don't particularly like bracing and don't have anything against surgery in severe cases). So I don't really see that much of a problem with parents or patients reading this study and thinking "If I wear my brace, I have a better/good chance of it actually working." The primary results are pretty well presented without anything tricky being done to the data. In curves that didn't progress, patients, on average, wore their brace more hours. I understand your concern about some of the studies and claims being made on the forum. I don't see that anything particularly bad would happen even if you believed everything written in this study. It's taking an accepted type of treatment and saying, if you do the treatment you have the best chance of it working. Am I oversimplifying or is there another concern you have that I'm not thinking about (which is certainly possible)?



                Originally posted by Pooka1 View Post
                Do you think BrAIST is big enough to change the approach of any surgeon if the results don't support bracing over watching/waiting to avoid surgery?
                Considering the number of sites, I'd definitely think that would be large enough to make a serious impact on standard of care. This is the size of study that I think is needed. Certainly, no matter what the outcome, some people will disagree. But overall, I'd imagine that whatever the basic finding of the study ends up being will become the standard of care. Personally, I'm excited either way. I hope they've done a good job collecting the data.

                Originally posted by Pooka1 View Post
                Yes it's a good first step. I think BrAIST is using heat sensors also.
                Yeah, they're using sensors. If they end up with the mountain of data they are supposed to get... OMG... How much fun would that be to dig around in that for a few days? Maybe they'll post it online. Or at least hook it up to a serachable database. Probably not.

                Comment


                • #9
                  One of the biggest problems with Braist is they deliberately excluded patients who were doing stretching and exercise while using their braces, as I recall.

                  According to what I have learned, and many orthopedists also indicate, stretching and exercise helps to increase flexibility in the spine and therefore increases the chance of bracing success. Because of this, I think the results of the study are going to be lacking going into it. Somehow I doubt they'll start another trial to see if bracing with exercise and stretching cuts it afterwards.

                  Hopefully, it will just show overwhelmingly that braces are effective, otherwise you really need to do more studies. It's a shame they didn't include those patients in the first place and not waste government money.

                  Comment


                  • #10
                    Originally posted by Ballet Mom View Post
                    One of the biggest problems with Braist is they deliberately excluded patients who were doing stretching and exercise while using their braces, as I recall.

                    According to what I have learned, and many orthopedists also indicate, stretching and exercise helps to increase flexibility in the spine and therefore increases the chance of bracing success. Because of this, I think the results of the study are going to be lacking going into it. Somehow I doubt they'll start another trial to see if bracing with exercise and stretching cuts it afterwards.

                    Hopefully, it will just show overwhelmingly that braces are effective, otherwise you really need to do more studies. It's a shame they didn't include those patients in the first place and not waste government money.
                    When we were trying to decide what we should include in our little rotational strength training study we discussed several other things. In the end we only included the rotational strengthening because you have to start somewhere. If you have too many extra things involved it's hard to know what works. Do the one thing really really well. Once you find out that one thing works/shows promise you add another thing.

                    We intentionally excluded patients that were wearing a brace. If they included patients involved in stretching and exercises then that can be used as a reason for or against a result in either direction. It's frustrating because it can move so slowly at times. But the end result is worth it. We either find out that bracing works and we can look at ways to improve that result (exercise or stretching for example). Or we find out that it doesn't work and we try to find something else (bracing plus exercise or even something completely out of the blue).

                    Comment


                    • #11
                      Originally posted by skevimc View Post
                      We intentionally excluded patients that were wearing a brace. If they included patients involved in stretching and exercises then that can be used as a reason for or against a result in either direction.
                      And rightly so. I think we have enough evidence in hand to show that PT can temporarily slow or halt curves such that including it in a bracing study would completely confound the results in the short term at least. In separate papers, Dolan and Dannielson (IIRC) excluded studies that included PT in their analysis of bracing efficacy for a good reason. And if there is any evidence of long-term slowing or halting of curves through PT then it will completely confound the bracing study results in the long term also.

                      Bracing and PT are competing hypotheses at the moment and have to be separated because even if you show the combined approach works, the first question will be which one mattered more because patients are NOT going to want to wear a brace and do exercises both, even in the short term unless they damn well know both parts are absolutely necessary.

                      On a related point, bracing is a short-term (not life long) treatment and I know you have investigated torso rotation as a short-term treatment but I wonder if the other PT folks claim their treatment is short term or life long?
                      Sharon, mother of identical twin girls with scoliosis

                      No island of sanity.

                      Question: What do you call alternative medicine that works?
                      Answer: Medicine


                      "We are all African."

                      Comment


                      • #12
                        Originally posted by skevimc View Post
                        It's taking an accepted type of treatment and saying, if you do the treatment you have the best chance of it working. Am I oversimplifying or is there another concern you have that I'm not thinking about (which is certainly possible)?
                        Well, I'll quote you from upthread which I completely agree with:

                        And just knowing how MD's work in the clinic across multiple fields, I doubt it will change much of anything in how they prescribe bracing or how they instruct. (Aside from the doctors at TSRH).
                        I suggest the great run of parents here will assume this is a game changing pub because they will read this line from the abstract:

                        CONCLUSIONS: The Boston brace is an effective means of controlling curve progression in patients with adolescent idiopathic scoliosis when worn for more than twelve hours per day.
                        That's what I see as the issue.
                        Sharon, mother of identical twin girls with scoliosis

                        No island of sanity.

                        Question: What do you call alternative medicine that works?
                        Answer: Medicine


                        "We are all African."

                        Comment


                        • #13
                          Here's something that is conspicuous by its absence... the starting curve magnitude of the progressive and non-progressive patients.

                          The reason this is important is that Cobb at diagnosis has been claimed by many as an important variable in predicting progression. They have the data and can say if it supports or does not support this claim that Cobb at diagnosis is predictive.

                          At one end of the spectrum, we can imagine that the starting Cobbs are equally distributed in both groups (progressive and non-progressive) at the end of the study. At the other end we can imagine that most of the smaller curves did not progress and most of the bigger curves did irrespective of brace wear times.

                          Thus it is conspicuous by its absence.

                          Also, were all or most of the 8 L or TL curves in the non-progressive group?
                          Last edited by Pooka1; 08-14-2010, 08:10 AM.
                          Sharon, mother of identical twin girls with scoliosis

                          No island of sanity.

                          Question: What do you call alternative medicine that works?
                          Answer: Medicine


                          "We are all African."

                          Comment


                          • #14
                            Well written

                            By the way, as I read and re-read this paper, I think it is very well written and extremely clear in what they do say.

                            Just wanted to add that.
                            Sharon, mother of identical twin girls with scoliosis

                            No island of sanity.

                            Question: What do you call alternative medicine that works?
                            Answer: Medicine


                            "We are all African."

                            Comment


                            • #15
                              Originally posted by Pooka1 View Post

                              On a related point, bracing is a short-term (not life long) treatment and I know you have investigated torso rotation as a short-term treatment but I wonder if the other PT folks claim their treatment is short term or life long?
                              I see PT overall as being targeted at correcting neuromusculoskeletal dysfunction. If there are soft tissues that, through some type of dysfunction, are causing pain or possibly damage, then rehab has the ability to improve that condition. Correct the dysfunction and hopefully improve the pain. However, if there is damage, exercises can't do much to fix that.

                              So I would say that PT can be a long term fix just like it is for knee pain or back pain. If you can correct the dysfunction before there is any type of physical damage, then you've probably got a long term fix. (Although, tendinitis or low back pain can always return.) If you correct the dysfunction after there is damage, it will probably require more work. PT won't heal articular cartilage or straighten wedged vertebral bodies.

                              If exercises are able to stabilize a curve through the growth spurt and the curve eventually progresses as an adult, I wouldn't say that is a failure of PT anymore than I'd say PT failed to permanently fix my patellar tendinitis if it returns after a couple of years. Exercises can't change the laws of physics or permanently keep muscles strong and functional. If a joint is permanently altered due to an injury or chronic condition, exercise can only work from that starting point. Maybe the alteration won't dramatically affect the muscle's ability to work properly and maybe it will.

                              This is why I am strongly in favor of early treatment. Treatment has to start before there is permanent damage for the best chance at a permanent solution. If there is permanent damage then it's possible that the function/biomechanics of the muscle will be permanently altered. Whether this alteration would eventually destabilize the spine is a guessing game. Certainly, the larger the change/curve, the more likely it is that the muscles will be severely affected.

                              Again, I wouldn't say that PT is a failure if it doesn't permanently stabilize a curved spine with no additional work required. Physics still exists, muscles will still get weak, bones will still get softer and cartilage will still dehydrate. To say that any treatment (PT, brace, surgery) is a failure if it doesn't protect against these things isn't realistic.

                              Originally posted by Pooka1 View Post
                              Here's something that is conspicuous by its absence... the starting curve magnitude of the progressive and non-progressive patients.
                              That is kind of silly that they didn't include this info.

                              Comment

                              Working...
                              X