In the referenced article, the author attempts to demonstrate that a significant increase in Staphylococcus aureus infection resulted from nearby land application of Class B biosolids.Conducting my own Chi-square analyses of the data given in Table 2, I found that 48% of the sum of squares was due to the Robesonia,PA site. Three people out of three interviewed had Staphylococcus aureus infection. However, Dr Lewis states that eight additional individuals lived within 50 meters of the site and "they could not recall any infection". These individuals were strangely left out. When they are added to the Table 2 data, the Chi-square results are not significant. His entire arguement, therefore is based on accepting a false hypothesis.
The problem with your less than adequate peer review is that Dr Lewis has publically used this article to attacks land application of biosolids. He recently stated in a public meeting that I attended, that his arguement had merit because this article had undergone peer review. Now, I have published research in refereed scientific journals and have reviewed manuscripts for publication while a professor at the University of Georgia. This article should never have been published.
Competing interests
None declared
Author response to Charles Pehl
David Lewis, U.S. Environmental Protection Agency
26 December 2002
Our paper did not include a statistical analysis of S. aureus infections; therefore, it is not possible that Mr. Pehl corrected such an analysis by including omitted data. Table 2 only describes the age, sex, and occupancy of individuals who lived in or regularly visited the single house in which the S. aureus outbreak occurred. Obviously, individuals who lived in the vicinity but did not regularly visit this house could not be included in such a table. The fact that these neighbors were not infected supported our conclusion that the S. aureus infections could have developed, in part, due to cross-infections among family members and close relatives dwelling in or regularly visiting this particular house. Case studies, as the paper states, are not statistical comparisons of exposed and unexposed groups. That type of analysis is best accomplished with dose-response curves (e.g., Figure 2 of our paper).
David L. Lewis, Ph.D.
Research Microbiologist
U.S. Environmental Protection Agency
National Exposure Research Laboratory
Ecosystems Research Division
Athens, GA 30605
Competing interests
No competing interests
Response to Author's response
Charles Edward Pehl, K-3 Resources, Inc
18 February 2003
The author states that the paper included no statistical analysis of S. aureus. However in the methods section he states: " Using chi-squared analysis, proportions of individuals reporting a particular symptom at each site were compared with the combined reports for that symptom from site to site". This is the analysis to which I referred. But I agree with Dr. Lewis that any statistical analysis of his data is seriously flawed since,as he stated,"This survey, which did not include an unexposed control group, was primarily intended to describe and document self-reported illnesses". In other words this entire study is based on anecdotal data and hearsay evidence. He fails to establish any connection between land application and perceived infections due to S. aureus, an ubiquitous, opportunistic pathogen.
My original comment still stands. This paper is based on seriously flawed research and should never have been published.
Response to D.L. Lewis et al. vol 2 article 11
20 December 2002
In the referenced article, the author attempts to demonstrate that a significant increase in Staphylococcus aureus infection resulted from nearby land application of Class B biosolids.Conducting my own Chi-square analyses of the data given in Table 2, I found that 48% of the sum of squares was due to the Robesonia,PA site. Three people out of three interviewed had Staphylococcus aureus infection. However, Dr Lewis states that eight additional individuals lived within 50 meters of the site and "they could not recall any infection". These individuals were strangely left out. When they are added to the Table 2 data, the Chi-square results are not significant. His entire arguement, therefore is based on accepting a false hypothesis.
The problem with your less than adequate peer review is that Dr Lewis has publically used this article to attacks land application of biosolids. He recently stated in a public meeting that I attended, that his arguement had merit because this article had undergone peer review. Now, I have published research in refereed scientific journals and have reviewed manuscripts for publication while a professor at the University of Georgia. This article should never have been published.
Competing interests
None declared
Author response to Charles Pehl
26 December 2002
Our paper did not include a statistical analysis of S. aureus infections; therefore, it is not possible that Mr. Pehl corrected such an analysis by including omitted data. Table 2 only describes the age, sex, and occupancy of individuals who lived in or regularly visited the single house in which the S. aureus outbreak occurred. Obviously, individuals who lived in the vicinity but did not regularly visit this house could not be included in such a table. The fact that these neighbors were not infected supported our conclusion that the S. aureus infections could have developed, in part, due to cross-infections among family members and close relatives dwelling in or regularly visiting this particular house. Case studies, as the paper states, are not statistical comparisons of exposed and unexposed groups. That type of analysis is best accomplished with dose-response curves (e.g., Figure 2 of our paper).
David L. Lewis, Ph.D.
Research Microbiologist
U.S. Environmental Protection Agency
National Exposure Research Laboratory
Ecosystems Research Division
Athens, GA 30605
Competing interests
No competing interests
Response to Author's response
18 February 2003
The author states that the paper included no statistical analysis of S. aureus. However in the methods section he states: " Using chi-squared analysis, proportions of individuals reporting a particular symptom at each site were compared with the combined reports for that symptom from site to site". This is the analysis to which I referred. But I agree with Dr. Lewis that any statistical analysis of his data is seriously flawed since,as he stated,"This survey, which did not include an unexposed control group, was primarily intended to describe and document self-reported illnesses". In other words this entire study is based on anecdotal data and hearsay evidence. He fails to establish any connection between land application and perceived infections due to S. aureus, an ubiquitous, opportunistic pathogen.
My original comment still stands. This paper is based on seriously flawed research and should never have been published.
Charles E. Pehl,PhD,MS (Texas A&M)BS(USNA)
Certified Professional Soil Scientist
Certified Nutrient Management Specialist
Certified Teacher of Mathematics
Captain,US Naval Reserve(Retired)
Compliance Director K-3 Resources, Inc.
Alvin, Texas 77512
Competing interests
None declared