If work done by Steven Milloy is indicative, parents may be able to relax about letting their children watch "Baby Einstein" videos, and the Journal of Pediatrics and the AAP may have some apologizing to do to Disney.
Apart from illustrating the limitations of peer review and applicability of data, it unfortunately illustrates something I've noticed with the American Academy of Pediatrics. Specifically, they've been willing to publish articles that probably should never have been published--at least without major revisions and qualifications of their conclusions.
Some examples? Sure. On the basis of a discredited study claiming you were more likely to be killed than saved by your own gun (the victims in the Kellerman study were actually killed with someone else's gun, by and large), they've come out against firearm ownership by parents. The proper conclusion of the Kellerman study is that you shouldn't live in a bad neighborhood and have criminals among your friends.
Also, on the basis of a few dozen reports of "failure to thrive" among babies subjected to Gary Ezzo's "Babywise" feeding schedule, they've issued a warning about this program. I am no fan of Ezzo or "Baby-Foolish", but I took a look at the actual data, and there is no significant difference; the overall % for "failure to thrive" is about 1%, so a few dozen cases would be statistically significant only if a few thousand people used his schedule.
Finally, the AAP came out recently with a recommendation against parents sleeping with their babies on the basis of one study that did not attempt to separate different variables related to SIDS and baby smothering. Unfortunately for the AAP's credibility, there are dozens of other studies that do control for these variables (smoking, couch vs. bed, intoxication, etc..) and demonstrate that co-sleeping is not a significant hazard.
Now don't get me wrong; I appreciate pediatricians and their work, and I'm very glad that they are good at recognizing diseases common to children. I just think that when they start telling parents how to live, they're in deeper than they can swim.
Which is, by the way, one statistically significant thing they can tell parents that is more significant than the four things I've mentioned combined. Be careful around the water.
Hot Gear Friday: The Swiss Army Guitar
-
The word “iconic” is overused these days. I try to avoid it. It’ll be hard
in this next bit. If you are not a guitar player, and someone says
“electric gui...
9 hours ago
7 comments:
When I was having babies, I was always baffled by the advice books that told you to "ask your pediatrician" about behavioral issues. I'm not talking about strange things, or things that were so baffling as to be worth checking out for some physical cause, or anything like that. More the run of the mill, "How do I get Johnny to eat his Cheerios and pick up his toys?" kinds of issues.
In my experience, it makes much more sense to ask a parent of three or so children whose abilities you believe are tested, than someone whose parenting status you know nothing about, on such issues. Sure, they study child development, but that's no substitute for experience. The reason I say "three or so" is simply that I've observed that parents of one or two children (myself included back in those days) tend to assume that the answers are pretty uniform and simple. Once you hit three, you're usually confronted with the fact that kids do vary enough that "one size fits all" answers don't work, but general principles do. However, that doesn't mean that parents of one or two kids don't often have sensible advice to offer.
The idea that pediatricians automatically know more about gun control than the rest of us (insofar as the rest of us are able to read up on the subject) is just silly.
...they're professionals! Scientists! of COURSE they know more! :)
man...I'll have to find the awesome quotes from 'that hideous strength' and 'road to serfdom' about that topic...
I'm with ya, Bert. We're a co-sleeping, gun-owning, Baby Einstein-watching family who doesn't like the Babywise books ... I've found, even as a first-time parent, that I'm better off relying on my own common sense and research than relying on the blanket statements from most pediatricians and medical journals. Don't even get me started on the misinformation being passed along by many doctors regarding breastfeeding/breastmilk ...
"Also, on the basis of a few dozen reports of "failure to thrive" among babies subjected to Gary Ezzo's "Babywise" feeding schedule, they've issued a warning about this program. I am no fan of Ezzo or "Baby-Foolish", but I took a look at the actual data, and there is no significant difference; the overall % for "failure to thrive" is about 1%, so a few dozen cases would be statistically significant only if a few thousand people used his schedule."
Actually, the AAP warning wasn't based on the FTT reports, but on the medical misinformation about infant feeding practices--Babywise was not named directly, but the practices warned about are taught by Babywise.
And, as a mom who had a son who was diagnosed FTT as a direct result of implementing "parent directed feeding" (with guidance from "Contact Moms" and an Area Coordinator) I don't find the possibility of FTT related to BW to be statistically insignificant.
Especially since within the BW subculture there are many, many times that FTT is not diagnosed but BW moms reassure one another that their babies were "small," too, and "someone has to be below the 50th percentile" and that it is "normal" to have significantly different percentiles for height and weight--all signals that would alert a non-BW mom to be concerned about growth.
(Btw, when you say "but I took a look at the actual data, and there is no significant difference"--which data sets are you referring to? Thanks!)
Tulipgirl, thanks for the visit. Part of your question has to do with the difference between personal significance and statistical significance. To use your example, the trouble for your son was personally very significant to you, but no statistician can do much with a sample size of one. He can only offer sympathy.
In the same way, imagine a lactation consultant has 12 Ezzo FTT children in her practice. She says "hum," and issues a warning, right? Medically, that's OK, but statistically, it doesn't wash; you need relative rates of FTT between Baby-foolish kids and demand-fed kids to make a contrast.
Now here is valid statistics; imagine she rebukes all 12 couples and only six repent. However, the repentant parents escape FTT, but those who continue either go on formula or continue in FTT. That's statistically significant due to the structure of the experiment.
In short, I agree with the conclusion and most of the methods, but simply citing FTT doesn't appear to be statistically significant.
Thanks for explaining. . . To my knowledge there have not been any actual studies done to support or not support Ezzoesque scheduled feedings--so I was wondering whether you knew of any?
I agree with what you are saying about statistical significance.
At the same time, those who have been within the circles of those who support and implement PDF/BW/ETC, there is a pattern that can be observed of milk supply issues, slowed and delayed growth, signs of dehydration, and on occasion actual FTT diagnoses. I know one has to be very cautious about citing personal examples--over the long term I hear more stories of problems than I hear "It worked for me!", but those abound as well. However, with what is known about how milk production works, what infants need for optimal growth and development, and the patterns for slowed growth--warnings are warranted, even if not at this point.
Agreed--I've actually sent a note to Dr. Aney. I'm not a pediatrician, nor do I play one on TV, but I am trained in stats and would love to see whether someone could make sense out of what's there.
Whether I would be that person or not remains to be determined, of course, but I'd love to have a chance to be personally reviled on Ezzo's site. :^)
Post a Comment