Possible involvement of Philip Morris statistical consultant in analysis of Enstrom/Kabat data
The newly revived discussion of the Enstrom/Kabat California passive smoke study and conflict of interest led me to cite some addiional documents. The results of the study, as Bero et al show, were presented at a Philip Morris gathering, June 5-6 2000:
This seminar was part of WSA/INBIFO programming, viz. to "organize two-day seminar with Mr. Peter N. Lee and Dr. James Enstrom to discuss CPS -I and CPS-II results and develop possible approaches to analyzing the data." PN Lee is a long time statistical consultant to the tobacco industry (Philip Morris, BAT, German Verband). One may fairly ask if Lee helped with the analyses as they finally appeared in the BMJ article:
PN Lee had sharply criticised two of Enstrom's earlier papers on cessation and mortality, one of which used the same California CPS-I and CPS-II data sets. In the first instance, Lee criticised Enstrom's failure at risk-averaging over time, and in the second (the CPS data sets) called Enstrom's analysis a "totally inappropriate age-adjustment procedure."
Since the BMJ article conclusions depended on both kinds of analyses (age-adjustment wiped out the effects of passive smoke; declining exposure occurred with time in California) one wonders if PN Lee did help develop analyses more to his liking this time around.
Norbert Hirschhorn MD