Welcome to This is MS, the leading forum for Multiple Sclerosis research and support. Join our friendly community of patients, caregivers, and researchers celebrating over 20 years of delivering hope through knowledge.
https://www.thisisms.com/forum/
Cece clarified the issue by adding...NHE wrote:If not, it seems to be a flaw in the design of the study. A good analogy might be something like this... Suppose you were tasked with determining the mileage of a car by using the average of a fleet of 20 identical cars to be tested. Now, suppose that half of the cars had properly inflated tires and the other half had grossly under inflated tires say around 5 psi. Would it be correct to include the low mileage from the cars with the flat tires? No. No company would do this. But that's just what happened when the data from the people with the successful angioplasties were mashed together with the other half of the data set with unsuccessful angioplasties. Moreover, the study was grossly under powered statistically due to very low recruitment into the trial. If n is too low, a small treatment effect is unlikely to be significantly different. I suspect that if this had been a drug study by a pharma company that the results never would have seen the light of day.ThisIsMA wrote:What I want to know is if they did a separate analysis of improvements for the 53% of participants who achieved increased blood flow?
Cece wrote:...I believe they did not. They looked at the whole group as "treated" whether or not the blood flow improved.