Originally posted by hdugger
View Post
So, if you tool a bunch of people with 25 degree curves and through a conservative treatment kept them all at 25, that would be of greater significance then just keeping them from advancing to surgery.
The Scoliscore results, if they stand, have raised the bar on conservative treatments really high, perhaps too high to show any clean results. They have to beat an ~75% "success" rate unless they can show those kids have a score >41.
But that's only part 2 of the question. Part 1 is, whether or not science can prove it, is it *desireable* to keep a 25 degree curve from getting to 35 for reasons of cosmetics/pain/further progression in adulthood.
The issue of measurement error that I mentioned really bothers me. That was perhaps the most significant criticism of Katz et al. as I understand the comments. There is something I'm missing here if measurement error is the potentially largest Achilles heel of that work. While that would naturally increase the variability, I still have a sense that the precision must be better than they think though it would be hard to show and I would never be able to prove it.
In my own field, let's say an analysis has a precision of 15%. If I subtract two numbers the precision on the difference is the square root of the sum of the squares as I am told which is about 21%. And the error keeps blowing up from there for other reasons. But to look at the data, it must be that some errors cancel because the data do not look like the total error blew up.
I understand the various sources of error in a Cobb measurement but I think the biggest ones are easy to control for... selecting the same end vertebrae and taking all radiographs in the early morning I think. And I don't actually understand how experts can even select differ end vertebra if they follow the simple rule.
The range of measurements in the paper on Hawes demands an explanation and none was forthcoming from the authors. That was beyond frustrating in trying to understand these issues. Some of it was different end vertebra but then why the heck aren't these experts using the simple rule? Her curve is not that complicated either as far as I can tell... single T. It is all very perplexing.
Comment