Erratum

A draft version of the article Excess Mortality in Two-Year Rodent Carcinogenicity Studies by Roth et al. (Toxicologic Pathology 35:7, 1040–43 ) was inadvertently published by the former publisher without giving the authors an opportunity to review the galley proofs. The resulting errors, some quite major, are listed and corrected below.

Major Errors (alter basic recommendations or severely impact comprehension)

1. Near the end of page 1040, one of the most basic practical recommendations in the paper is given in the following sentence: Quantitatively, one reasonable rule of thumb using purely statistical considerations would be to refrain from terminating a group unless its sample size was less than 10 and some other group’s sample size was at least 5 times as large or its sample size was less than 5 and some other group’s sample size was at least 3.5 times as large. ^(*) The numbers five (in “five times as large”) and 3.5 are incorrect and should be changed to four and three, respectively.

2. The sentence quoted above is followed by the footnote symbol ^(*), which refers to the following footnote that should have been placed at the bottom of the page on which this sentence appears (page 1040): ^(*) Although this rule of thumb was included only in a fairly late revision at the request of a reviewer, an even later reviewer strongly suggested that we also provide detail to justify it. After considering the length and complexity of a full explanation, we have decided not to provide one. Briefly, the combination of at least one very small sample size and one comparatively large one leads to both power and bias issues for the analysis, regardless whether the best analysis turns out (see ensuing discussion) to be a trend test or a pairwise comparison against controls. This footnote indeed appears in the article, but in the middle of the text near the bottom of the first column on page 1041 rather than at the bottom of page 1040. In that location, there is no inkling that it is intended as a footnote, and most readers were undoubtedly left wondering why it has any relevance in the place that it appears.

3. The paragraph starting near the bottom of the first column on page 1042 begins with the following sentence: The analysis modification recommendations to provide some context, in the more common situation where excess high dose mortality is not an issue, the almost universally used procedure is to perform the analysis of high dose tumor incidence using trend testing methodology and to regard this as the primary analysis. This “sentence” is incomprehensible. The first four words are intended to be the heading of a new section, and the explicit display of this heading is important to the understanding of the basic, logical flow of the paper. The text should be formatted as follows (with only the heading in italics):

The Analysis Modification Recommendations

To provide some context, … (complete the sentence as given above)

• 4. The reader’s grasp of the logical flow of the paper was also impacted by the lack of distinction between the various intended levels of section headings. In particular, all of the section headings were supposed to be in bold print except for the two headings The Design Modification Recommendations and The Analysis Modification Recommendations (which, as mentioned above, did not appear as a heading at all). This was intended to convey the message that the above two headings represent subsections of the more global heading Increased Mortality in Only the High Dose Group. Since one of these headings appeared in the same typesetting as the rest of the headings and the other one didn’t appear as a heading at all, much of the main logical structure of the paper was lost. Along the same lines, but of less consequence to comprehension, is the fact that three of the headings (INTRODUCTION, CONCLUSIONS, and REFERENCES) appear centered and in all capital letters, unlike all of the other headings.

Mid-level Errors (impact comprehension more than trivially but less than severely)

1. The fourth and fifth paragraphs in the INTRODUCTION were intended to be just a single long paragraph. Splitting this material into two paragraphs introduces possible confusion, obscuring several important connections including the strong logical relationship between the sentence immediately preceding the split and the sentence immediately following it.

2. The next two paragraphs were also intended to be just one long paragraph; there is no convenient split point that avoids breaking up what is inherently a single, unified idea. This added “break point” might have caused readers to have to ponder things longer, but ultimately the intended meaning probably became clear anyway.

3. The first two full paragraphs on page 1042 were likewise intended to be just a single paragraph, albeit a very long one.

Minor Errors (trivial impact on comprehension)

1. The heading currently called REFERENCES was intended to be called “References Cited in FDA Guidelines Excerpt.” Though this is not of major importance, it does give the possible incorrect impression that these are the authors’ own references rather than the those of the FDA.

2. There was an assortment of purely editorial errors that seem too inconsequential to warrant individual mention. They fall into the following categories: lack of a blank space between words, a hyphen that splits a word between lines in the middle of a syllable, extra hyphens between words, and several occurrences of an Arabic numeral in the middle of a sentence instead of spelling out the number.