Schanler's study was a clinical trial in which babies who were as equal as
possible in all other respects, were randomly assigned (using some accepted
method of assignation) to receive one treatment or another.  That is what an
RCT is about - keeping all other factors as similar as possible between
groups, and manipulating one single factor, in hopes of making visible the
effects of that factor alone.

One of the key things in an RCT is that assignment to a group is permanent
for the purpose of statistical analysis of the data.  That means that data
about all babies who were meant to be fed donor milk, will be analyzed in
one group, and the meant-to-be-fed-formula ones will be analyzed in another.
This ensures against skewing of the results from there being less adherence
to one treatment protocol than to another, and is key to the stringency of
an RCT.  The fact that many babies needed to be fed differently from how
they were randomized, is an important finding in itself, and it bears
witness to the honesty of the researcher if it is reported and accounted
for.

Obviously it would have been unethical to allow babies to stagnate in growth
simply because they were assigned to one group or another.  But it would be
just as unethical in a scientific sense to then analyze data as though they
were originally randomized to the control group, because they weren't.  

I still haven't read the article, but I would expect to find some mention in
the 'discussion' portion thereof, about what factors (bias, attrition, need
to switch babies from experimental to standard treatment) weaken the
findings and what factors might strengthen them.

Nancy included some discussion in her post on the article, and I urge you
all to read and re-read that post.  Failure to maintain randomization for
purposes of data analysis is a cardinal sin in RCTs.  It is possible to look
at the data again, looking at what actually happened to each baby, and it is
probably of interest, but it is not acceptable in the writing up of the
results of this study to analyze these events as though the babies were
randomly assigned to have the actual outcome they had.  

This isn't an easy concept but it is really important to understand.  There
may be lots of other things to question in this study for all I know, but
analysis of data by 'intent to treat' definitely isn't one of them.

Rachel Myr
Kristiansand, Norway  

             ***********************************************

To temporarily stop your subscription: set lactnet nomail
To start it again: set lactnet mail (or digest)
To unsubscribe: unsubscribe lactnet
All commands go to [log in to unmask]

The LACTNET mailing list is powered by L-Soft's renowned
LISTSERV(R) list management software together with L-Soft's LSMTP(R)
mailer for lightning fast mail delivery. For more information, go to:
http://www.lsoft.com/LISTSERV-powered.html