June 4, 1998
Devra Lee Davis, Ph.D. World Resources Institute 1709 New York Avenue NW Washington D.C. 20006
Dear Dr. Davis,
It is unfortunate that our letter to the editor comments on your global warming health impact report (The Lancet, Nov., 1997) did not get published in light of your totally misleading if not dishonest response sent to Stephanie Clark at The Lancet.
Your discussion points suggest strongly that you have never read our unpublished "monograph" nor did you give any detailed thought to our concerns. Let me first point out several of your misstatements.
1. "...details on how well the mortality information used compared even with the ecological analysis by Pope is not readily available."
On pages 18 and 19 of our paper, a thorough discussion of our concerns with the Pope mortality data was presented. We clearly stated that he refused to release his data to us so that we could replot his data as well as cross check our analysis. We explained exactly what data we used and its source, i.e., CDC Wonder. Your comment is absolutely false.
2. "Jones et al are in error regarding the standard promulgated by EPA. The July 18, 1997 Federal Register clearly states that the standard is 15 ug/m3 annual mean, not a median."
Nowhere in our paper did we make such a statement. Our entire discussion relates to EPA's mistaken interpretation of the Pope ecological plot as discussed on pages 5-8. EPA assumed that there was an apparent threshold at the median concentration of 15 ug/m3 which is equivalent to a mean concentration of at least 18 ug/m3 with which EPA agreed. Our whole discussion relates to how EPA's mortality estimates became flawed in light of the sophomoric mistake -as well as the Agency's error in calculating death benefits below an air quality benchmark which it clearly stated had an inherent margin of safety. Do you have a definition for margin of safety different from the rest of the scientific community? Your accusation is naive and totally false.
3. ..."it is important to note the general agreement between Pope et al (1995) with respect to mortality and indicators of particulate pollution in a second recent prospective cohort study (Dockery et al, 1993)..."
Dockerty et al did not find an association between PM 2.5 and mortality among non-smokers. See our discussion on page 18. Your inference about "general agreement" is not true unless you believe black is white.
4. "Furthermore, Jones own estimates submitted in comments to to EPA suggest that ... a substantial fraction of the estimated numbers of lives prolonged would come (at) concentrations exceeding 15 ug/m3 annual mean."
Where is the citation? No such comments were submitted to EPA. My handwritten calculation methodology was shared and confirmed by John Bachman of EPA. Those results were never formally submitted to EPA as my professional opinion. The CSE paper represents my best estimate of the upper bound mortality estimate Pope's questionable results. Again you have made a false statement.
It seems to me that you have a professional responsibility to think about comments made about your report and provide ethical responses. Let me point out two areas where you didn't think through our comments.
1. You make the point that we focused too much on Pope et al's simple ecological analysis and not on the core prospective cohort results. Is it not true that all of the cohort data would have to fall on the same discreet PM 2.5 levels for each city, i.e., 53 discreet median PM 2.5 concentrafions? Wouldn't you expect that the removal of the other variables cited by Pope that the slope through these data would in all likelihood be flatter than the all cause ecological mortality plot? This is what Pope found. Our independent analysis found no ecological association for disease related deaths and hence questioned whether Pope's cohort results were valid. Is the use of PM 2.5 data which overlaps the upper range of the cohort death window, i.e., 87-90 PM 2.5 data versus 82-89 mortality data really unreasonable given the lifetime exposure concentration uncertainty? (Since the nationwide and regional PM 2.5 trend data is available such data could have been adjusted to the 1980 time frame.) Doesn't the added data, as well as the corrected Ashland data point in Pope's ecological plot (Figure 1 in the Jones et al addendum) raise any questions in your mind? Your report made no attempt to caveat the use of the suspect Pope work. Why didn't you challenge Pope's own concern which we quoted?
2. How would you explain scientifically the absence of an obvious threshold in observed concentrafion response relationships which you believe to be "a hallmark of numerous epidemiology studies." I suggest you might want to review the critical review paper by Vedal, "Ambient Parficles and Health: Lines that Divide," JAWMA, Vol. 47, May 1997. He explains quite well the interpretation problems in such studies.
I have also noted your slight regarding our monograph: "has not been published in a refereed journal." The vast majority of your report references were not published in refereed journals. Further, you have no personal authorship citations in the report which would reflect your expertise in conducting such risk studies although you are cited as the principal author. How is it that a sociologist becomes the lead investigator of a complex air quality risk study? Did your funding come from EPA's competitive extemal research program? I doubt it.
How sloppily and hastily the Davis response was put together is evidenced in the many typographical errors, omissions and misquotations (at least seven) presented in the response. For example, you cite Pope's memo to John Bachman (Ref. 2) but fail to mention my documented rebuttal transmitted to Bachman on May 8, 1997 (attached). Were you not made aware of this by EPA?
You have mislead the Lancet editorial staff by making false and in my opinion unethical statements. Unfortunately, the Lancet staff has no familiarity with the subject, accepted your comments at face value, and denied publishing our commentary. It was also obvious that they have no interest in truly refereeing your work. The peer reviewers chosen by Lancet to review your report clearly lacked the expertise to critically review your politically motivated rather than scientifically credible effort. I request that you provide copies of this letter to your co-authors or provide me with their addresses so that they will be informed of your unprofessional response to The Lancet.
cc:
Richard Horton, The Lancet
Stephanie Clark, The Lancet
Joel Bucher, CSE
Michael Gough, CATO
John Bachman, EPA
Sincerely,
/s/
Kay H. Jones, Ph.D.
Comments on this posting?
Click here to post a public comment on the Trash Talk Bulletin Board.
Click here to send a private comment to the Junkman.
Copyright © 1998 Steven J. Milloy. All rights reserved on original material. Material copyrighted by others is used either with permission or under a claim of "fair use." Site developed and hosted by WestLake Solutions, Inc.
Material presented on this home page constitutes opinion of Steven J. Milloy.