Scene: Almost 30 years after their TV series was canceled, Bruce Wayne and his not-so-youthful ward, Dick Grayson, languish in stately Wayne Manor. Suddenly, the Batphone buzzes. A call for help from Gotham City! Sliding down the Batpoles to the Batcave, Batman and Robin find Alfred the Butler... playing solitaire on the Batcomputer.
Later, in Commissioner Gordon's office in Gotham City Police headquarters, Batman and Robin learn that the RADdler (a radon epidemiologist and a descendant of the notorious criminal "Riddler") is on the loose, leaving only a note that read:
Riddle me this Dynamic Dunderheads: How do you make something out of a lot of nothing? Robin: Holy hotcakes, Batman? What's the RADdler up to?
Batman: Well, Robin, since the mid-1980s the RADdler has been trying to trick the public into believing that radon causes 40,000 lung cancer deaths every year.
Robin: Oh yeah, that bogus weak association epidemiology stuff!
Batman: Precisely. Now, I recall that in 1995, Jay Lubin, John Boice and Jonathan Samet published a paper in Radiation Research [144:329-341] that stated "to date, epidemiologic studies of risk from residential radon have not convincingly demonstrated an association with lung cancer."
Robin: Right! After looking at seven major case-control studies of lung cancer and residential radon, they concluded that, if there is any risk from radon, it is so small that it cannot be measured because of the large errors involved in estimating exposure to radon.
Batman: Exactly, Boy Wonder. That's the "lot of nothing" in the RAD-deer's riddle. Now, how do you make "something" out of the "nothing" we call radon epidemiology?
Robin: Look Batman! In my new issue of The Journal of the National Cancer Institute... a new radon study! And it's by Lubin and Boice... with an editorial by Samet! Holy radioactive coincidence!
Batman: Good work Robin. Let's see...a-ha...Lubin and Boice have concluded that by combining the radon epidemiologic studies together through a meta-analysis, "a lot of nothing" can become "something!" Lubin and Boice used the meta-analysis technique to report that higher levels of radon exposure increase the risk of lung cancer by a statistically significant 14 percent. Then Samet, the King of Radon, blessed the report in his editorial!
Robin: What... what's a meta-analysis?
Batman: As you will recall when the U.S. EPA cooked up the risk assessment for environmental tobacco smoke in 1992, meta-analysis is a STATISTICAL technique in which the data from multiple small studies is combined to make one large study. Theoretically, a meta-analysis is a pooling of the data of the small studies to get an overall picture of all the data.
Robin: But Batman, that's only theory. Meta-analysis of epidemiologic studies is nothing more than a mean trick played on the unsuspecting. It's sole purpose is to transform meaningless small studies into a (supposedly) meaningful large study. Meta-analysis may be appropriate for combining carefully controlled animal experiments or clinical studies, but not epidemiology!
Batman: You're right my boy! Epidemiologic studies usually differ greatly from one another. Different study population characteristics. Different methodologies. Different assumptions. Different strengths. Different weaknesses. Combining them is like adding apples to oranges to make "apple" pie.
I recall serious criticisms of the technique in major scientific journals [i.e., "Meta-analysis in the Breech", Science 1990;249:476-480; and "Can Meta-analysis Be Trusted?", Lancet 1991;338:1127-1130].
Robin: Holy junk science! So the answer to the RADdler's riddle is meta-analysis! What do we do now Batman?
Batman: To the Batmobile old chum...
What will the RADdler do with this new junk science? Is the Dynamic Duo too late? How do Samet, Lubin and Boice sleep at night? Will they ever get a life and forget about radon? Tune in sometime in the future for the answers to these and other questions. Same Bat-home page. Same Bat-URL. Material presented on this home page constitutes opinion of the author.
Copyright © 1996 Steven J. Milloy. All rights reserved. Site developed and hosted by WestLake Solutions, Inc.
