 |
 |
----- Original Message ----- From: "rayilynlee" <[log in to unmask]> To: <[log in to unmask]> Sent: Thursday, May 18, 2006 11:18 AM Subject: Re: Correction on Selegeline's neuroprotection > Selegeline was the first med prescribed for me after > diagnosis. I never > took it. How would one know if progression of the disease > is slowed when > everyone is so different? To my knowledge there is no > standard rate of > progression. > Ray This is exactly what statistical analysis does. Assuming that there is a valid way to measure disease progression (not a trivial issue), the investigators randomly assign the new treatment or the best existing treatment to each of a group or research subjects. Ideally, they should all be at the same stage of the disease. You should also make sure that neither the patients nor the doctors know which individuals are getting which treatment. After the treatment period is over and each patient's progression has been measured, you then determine whether those getting the new treatment not only do better but do better enough that the difference is not likely to be due to chance. What does it mean to say that "the difference is not likely to be due to chance"? Some people are naturally going to progress more slowly or react more favorably to the medication. If there happen to be more of such folks in the treated group, then the treatment will appear to be better than it is. But since the groups are assigned by a formal random process, the mathematics of probability lets us calculate how likely it is for a difference as large as that observed to have occurred as a result of random variation. The more variable an outcome, the larger the group that needs to be studied to insure that we are not just seeing a chance fluctuation. If there is data available before the study showing how large this variability is, then the investigators will devote considerable effort to try to make sure their sample is large enough. If such data does not exist, then you do the best you can. If it turns out that the outcomes are more variable than was thought beforehand, then you likely end up with an inconclusive result. ---------------------------------------------------------------------- To sign-off Parkinsn send a message to: mailto:[log in to unmask] In the body of the message put: signoff parkinsn