Reader Comments

Post a new comment on this article

Tone and complication

Posted by plosmedicine on 31 Mar 2009 at 00:31 GMT

Author: Thomas R. Knapp
Position: Professor Emeritus
Institution: University of Rochester & The Ohio State University
Submitted Date: October 16, 2008
Published Date: October 20, 2008
This comment was originally posted as a “Reader Response” on the publication date indicated above. All Reader Responses are now available as comments.

I am concerned about two things: (1) the tone of the article (apparently set by the three themes referred to on page 1215 that "the tobacco industry lies, nicotine is addictive, and secondhand smoke kills"?); and (2) the complicated analysis.

It is fine to compare the expenditures on health care for states that have a strong tobacco control program with those that do not. But what purpose is served by continuing to bash the tobacco companies (no, I am not a shill for them; I've never smoked in my entire life of 78 years), reminding readers that nicotine is addictive (of course it is), and claiming that secondhand smoke kills (the evidence for that claim is far from convincing)?

Were "cointegrating regressions" really necessary? And confidence intervals are used inappropriately throughout the article. They are relevant only for probability samples from well-defined populations, which is not the case here. Why wasn't a simple table provided that listed the health care expenditures for each of the 39 states (California plus the 38 control states) for the various years? The use of the term "control" in that sense is confusing in an article concerned with tobacco control. It is also misleading, since it implies that some sort of experiment was carried out. It was not. It also implies that causality can be safely claimed. It can not.

No competing interests declared.