Lies, Damned Lies, and Statistics: Vol. 1

Mark Twain's famous quip probably has four hundred thousand instances prior to this one, but for the Billy Rubin Blog, this is Volume Number One.

Actually, the problem in this instance may not necessarily be in the statistics, just how the stats are portrayed. And the difference between one portrayal and another is the difference between a shrug and full-fledged panic.

In yesterday's New York Times there is a brief piece about a study of "near-term" infants, comparing them to those born full-term. The gospel when I was in medical school was that 32 weeks was something of a magic number--those babies who could make it to that time were believed to have no increased risk of abnormalities compared to babies who were not delivered until 40, i.e. the full term. The current study discussed by the Times article casts doubt upon that. The near-termers (born 3-6 weeks prior to due dates) were noted to have--and this is the key word--subtle differences in levels of disabilities or developmental delay. But before we get there let's trot the stats out. Assuming the research is good (which I do) both of these statements are true:

a. Near-term babies have a thirty six percent higher risk of having a disability or developmental delay.

b. Slightly more than four percent of near-term babies had these abnormalities, while just under three percent of full-term babies had them.

As you can see, statement "a" seems a good deal more troubling than statement "b," even though they utilize the same data. Interpretation is everything. And the Times does a nice job of explaining the nuance. Roni Caryn Rabin, the reporter, notes before laying down these two identical yet seemingly contradictory stats that "Over all [sic], the risk is small, and doctors emphasized that parents should not be alarmed."

The question is: will smaller print-news organizations that are likely to pick up the Times story, or the local TV media that have a habit of trying to tease viewers with the most sensational bit of a story to attract viewers to their local "Health Watch" sections of their news programs, relay the all-important subtlety that the overall risk of developmental delays and whatnot in near-term infants is still quite low?

I have my suspicions. Readers, help me--let's keep track of this story over the next week and see what play it gets, and if it becomes a sensational, 36 percent increased risk in abnormalities, New Threat To Civilization, or if it is described for what it really is. I hope to hear from you.
--br