Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: sampsi

From   "Eurich, Dean" <[email protected]>
To   "[email protected]" <[email protected]>
Subject   Re: st: sampsi
Date   Fri, 19 Aug 2011 13:09:37 -0600

Hi Phil,

Sorry for the confusion. We are not really interested in the longitudinal
changes over the longer term (I.e., 5 years). What we want to do do is
basically - Does very short term changes in brain function at time of
diagnosis (I.e., 8 weeks) predict 5 yr mortality.  So in a sense the 8
week change score a "baseline value". We are not interested if the
patients continue to decline or not.

Hope that makes sense.

Thanks again.


On 11-08-19 12:57 PM, "Phil Schumm" <[email protected]> wrote:

>On Aug 19, 2011, at 9:07 AM, Eurich, Dean wrote:
>> The study is really two separate studies. The first is what sample size
>>do we require to observe a 2% change in magnetic imaging brain scans for
>>controls and ALS patients over an 8 month period.
>To clarify, the calculation you originally asked about involved detecting
>a difference in the mean change (in your imaging outcome) between ALS and
>controls.  IOW, it appears that you want to demonstrate that ALS patients
>on average decline faster than controls (who I presume are matched on
>age, gender, etc.).  This is consistent with the baseline measurements
>you showed (where the average for the controls was higher than for those
>with ALS), but I presume you want to show/estimate the difference
>longitudinally rather than just cross-sectionally.  You don't need to
>explain the rationale for doing this -- I just wanted to make sure that
>we were on the same page (your sentence above is a bit ambiguous WRT
>-- Phil
>*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index