For years, one of the great controversies in the treatment of scoliosis has been the use of bracing. Despite almost a hundred years of (mixed) experience, the scoliosis world divided itself into three camps. On one side, were the Ardent Bracers – “Give me enough Plastazote and I will conquer any curve!” The rival camp, the Nihilists, would rejoin, “It’ll never work!” like the character in the child’s cartoon. Somewhere in the middle was a pragmatic group which would use the braces with modest expectations, sometimes because they knew that a brace must work in some settings or no one would have invented in the first place and sometimes because they had nothing better to do.
In October 2013, a landmark article was published in the prestigious New England Journal of Medicine titled “Effects of Bracing in Adolescents with Idiopathic Scoliosis” with the lead author Dr. Stuart Weinstein from University of Iowa. The article reported on the findings of a multicenter trial (the BrAIST trial) involving both randomized and preference groups of braced vs. observed patients. 72% of the braced group did not progress to surgical curves by maturity vs. 42% in the observed cohort. The trial was in fact so successful that the group applied the research equivalent of the Mercy Rule – they stopped the trial well before its scheduled time because the numbers were so one-sided.
This is a landmark for several reasons. For generations, the holy grail of bracing literature was how to do a suitably ‘controlled’ study with the devices. After all, you can’t fool a patient into thinking they are wearing a brace when they are not (and vice versa). And you can’t hide the brace from the doctor. Combining the randomizing with a preference group would seem to balance out some of the inherent bias in any brace study. Although I believe that there is still an ‘apples vs. oranges’ effect inherent in the fact that every curve has its own personality, the relatively large groups involved would tend to even this out (although I am surprised that the authors could only come up with 242 patients for this study – why not add another few centers and get up to four- or five hundred?). So here at last is the first recognizably scientific paper on bracing, and it is in favor of the devices.
The favorable outcome for bracing is in itself landmark in a surprising way. Dr. Weinstein, while not a true Nihilist has been known in the community to be skeptical to say the least about braces. I was involved in discussions with him and Lori Dolan, PhD at a national meeting not long ago, and they seemed convinced that the study would go quite the opposite way, citing the difficulties in getting kids to wear the braces and the somewhat suspect effectiveness as a reason why bracing is not cost effective. The numbers in this study must be truly convincing if Dr. W has come around to the pro-brace side.
The third and most astonishing thing about this study is that it was published in NEJM. The Journal as a rule is the bastion of Internal Medicine and Public health studies. It is unusual to say the least for them to publish a surgical paper at all. The last time they even acknowledged the existence of orthopaedics was to publish a very negative article concerning the use of a specific knee arthroscopic technique. I’m not sure they even know how to spell ‘pediatric orthopaedics.’ For them to feature a work on a mechanical treatment in Pedi Ortho is as remarkable as when the Red Sox first won the World Series. But there it is, in proverbial black and white. Which means that the paper must be remarkably rigorous from a statistical point of view (and it seems to be) and it must be of really significant clinical importance.
So does Dr. W’s paper solve the question? Are the Nihilists now going to throw down their standards and take up Boston braces with a hearty “Huzzah?” Probably not. There are still important questions about which curves do best with bracing, when to start (age and magnitude of curve) and when to stop. The paper rather tantalizing brings up the amount of time in the brace as an important variable but falls short of any clear recommendation. What is the effect of genetics of the curve? Will the ‘successful outcomes’ in this study hold onto their success or will they progress even after maturity?
Most importantly, what kind of brace is best? I will discuss the types of bracing a future post, but there are rigid and flexible, daytime and nighttime types. Until the community can come to a real consensus as to who to brace, how long to brace and what brace to use, the controversy will rage on and on.