There are a lot of difficulties with doing a study the way you suggest, many of which would prevent it from being either conclusive or definitive.
One practical problem is that just getting the medical records for a large number of people scattered across the globe would be a huge undertaking, and not always possible even if you tried. There are various privacy laws that would have to be worked through, plus "simple" problems like records being stored in different formats at different medical centers, different MRI protocols, etc. So such as study would still be fairly costly and time-consuming, though obviously not as much as a prospective study.
Another problem is that CCSVI intervention isn't yet standardized, nor are the procedures used to even identify blocked veins. Did all of the patients have MRIs done just prior to the procedure, and again at roughly the same interval afterwards? Comparing 6--month MRIs for some patients with 12-month MRIs for others could be difficult.
Another really big problem is that the MS patients who have already had CCSVI intervention are almost by definition not "typical". That's why well-controlled clinical trials usually start by collecting a large group of "similar" patients, and then randomly assigning them into treatment and control groups, to avoid leting patients self-select. How does this mattter? Well, what kind of MS patient signs up for an invasive, experimental, unproven therapy, especially if it involves paying a lot of money out-of-pocket? I would expect many of these patients have already tried existing medications and found them ineffective. They may have more advanced or intrusive MS symptoms than "typical", leading them to be more willing to try something adventurous. Point is, they aren't typical by any stretch of the imagination, and these differences make it nearly impossible to find a truly comparable control group to compare them against.
berriesarenice wrote:is there an issue with taking 100 patients who got their veins opened on Jan 8 (one month ago), and counting their 100 total relapse-free months with open veins, the same way that you would count the 10 relapse-free months of 10 patients who got their veins opened in April (10 mos ago). Does that make sense? Are all relapse-free months equal?
No, not all months are equal. As a simple example, post-surgical complications are likely to be most prevelant right after a procedure, so combining data from 100 1-month intervals would grossly overstate the surgical risk. I think there is also some debate over whether or not the stress of surgery itself has any impact on relapse rate, so again you'd risk adding that confounding factor to the analysis.
I think it would
be very interesting to do a retrospective study as you suggest, but I think the best you could hope for would be for it to add a bit more interest among the medical community. Due to the reasons listed above, such studies tend to be subject to a lot of (often valid) criticism, which prevents them from being accepted as very conclusive.
Note that there is a third very important question to be answered before CCSVI treatment becomes widespread, and that is whether or not CCSVI can be safely
and effectively treated. It's tough to answer that question without long-term studies.