I am finally getting a chance to read the abstract fully (not yet the full
article) and comment on a flaw that I noticed when I first skimmed the
abstract.

The study suffers from the same problem that the HIV/mothers mortality
study suffered from.  The so-called randomization was not randomization of
the actual behavior study, but the intervention to try to influence that
behavior.  Unlike trials of medications where you can actually control what
the subjects ingest, you cannot control behavior to the same extent (thank
goodness!).  In fact, 61.8% of the infants in the intervention group still
were given pacifiers!!!!

So, this study does NOT test randomized allocation of pacifiers versus no
pacifiers.  What this study DOES test is a randomized allocation of an
intervention to discourage pacifier use.  In this case, the intervention
did have a statistically significant impact on pacifier use, BUT as with
most behavioral interventions, not all of the subjects adhered to the
recommendation. This means that the effects of the control and intervention
are "attenuated".  The implications are that:

1) A better designed behavioral change intervention may have reduced
pacifier use even further and it is possible that a larger difference in
the behaviors (pacifer use vs. not) might result in a difference in weaning
at 3 months of age.
2) A larger sample size might have provided more statistical power to pick
up a difference in weaning at 3 months of age - compensating for the
attentuation of the behavioral intervention.
3) The many other behavioral factors sited in numerous postings by
lactnetters may have "overridden" a real effect of pacifer use had these
other obstacles to longer breastfeeding duration been removed.

I'm sure if I thought about it long enough, I could come up with even more
possibilities.

I do not agree with the authors that they have proven that pacifier use
does not cause breastfeeding problems. It is much more difficult to assess
causality from a negative study and the fact that it would be next to
impossible to truly randomize pacifier use makes it even more difficult in
this context.  I do think that, in some instances, their conjectures about
whether or not the infant's feeding behavior influences the use of
pacifiers shows merit and should be investigated.

I should point out that the abstract I read made it pretty clear that they
were talking about the "intervention" which was the discouragement of
pacifier use and not use of the pacifier itself.

Kramer is a solid epidemiologist who did a marvelous review in the 1980s on
the causes of LBW.  Nevertheless, epidemiology needs to be complemented
with those knowledgeable in the field of inquiry.  I've loved all of the
interesting interpretations of what could be going on behaviorally with
pacifier use!

Susan Burger, PhD, MHS
Once upon a time a nutritional epidemiologist

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