If You Torture Data Enough It Will Speak To You!

 FDA has commissioned a number of studies of menthol based upon secondary analyses, meaning that the survey had been completed in past years and the data is then reexamined by researchers, often against study objectives that do not match the intent of the original survey.  For example, in the studies by Okuyemi (2003 and 2007), National Survey data was re-categorized into menthol and non-menthol sub-groups which became disproportional and not adequate for computing statistics.

CRE is not alone in this conclusion in that authors of one of the FDA commissioned studies state:

“All the studies examining the relationship between menthol smoking and cessation have important limitations that may explain the inconsistent findings. In particular, many of these studies were based on convenience samples, or secondary analysis of clinical trials data or case control studies. (1, 3, 4, 6) First, cessation has not been operationalized consistently across these studies. Second, these studies are unlikely to be representative of the general population of smokers who made a quit attempt. Only one paper published to date has utilized nationally representative population level data”

(Cristine D. Delnevo, PhD, MPH, Daniel A. Gundersen, MA & Mary Hrwyna, MPH}

To the credit of a number of authors, they do not make definitive conclusions; instead they state  the study “suggests” a particular conclusion.

Consequently although the presentations to be presented to TPSAC may be a guide for additional research; they are insufficient to make definitive conclusions regarding smoking initiation/cessation related to the use of menthol.

We refer our readers to two studies that arrive at  conclusions opposite that of the FDA sponsored reviews at  http://www.thecre.com/tpsac/?p=892

In those  few instances where the FDA commissioned reviewers use  mathematical models developed to process information having several orders of magnitude more accurate input data  then the data being analyzed, the reviewer is conducting  a project in sub-optimization—doing something well that should have not been done in the first place; others might call It torturing the data.

N. B.  CRE does not have the resources to analyze each new round of studies that FDA surfaces—a competitive advantage FDA has due to the size of its budget, an advantage which is compounded if in fact the budget of the Center for Tobacco Products is not subject to the scrutiny of the federal appropriations process but  is instead financed by user charges levied on the tobacco industry.

Leave a Reply

Please Answer: *