HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:
Guest Access | Sign In via User Name/Password

This Article Full Text (PDF) Free Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Article Archive Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Barnett, S. D. Search for Related Content PubMed PubMed Citation Articles by Barnett, S. D.

Increasing Statistical Accuracy

Inova Heart Institute at Fairfax Hospital Falls Church, VA

Correspondence to: Scott D. Barnett, PhD, Manager of Epidemiology and Biostatistics, Inova Heart Institute at Fairfax Hospital, 3300 Gallows Rd, Falls Church, VA 22042; scott.barnett{at}inova.com

To the Editor:

I was pleased to read the article by Moss and colleagues in the March 2003 issue.¹ As a professional epidemiologist and biostatistician, I can only welcome any press that may increase the need for the services of my profession. I agree that too many readers take the presented methodology and reported statistics on good faith. In fact, a certain amount of faith on the part of the reader that the investigators have performed an exhaustive and thorough statistical analysis is inherent.

This letter is directed to the question of how investigators can increase either their own individual statistical knowledge or can increase the accuracy of statistical reporting. The answer is to increase one’s own knowledge or to solicit the help of a trained analyst. Self-study on the part of the physician is certainly an option, and no shortage of statistical texts exists, but what physician has the time? Formal academic training is an option, but, again, the issue of time, and more importantly, the topics discussed by Moss et al, are not readily covered in depth until the third or fourth class in a given course series. Furthermore, faculty probably would not be inclined to let individuals skip ahead.

The solution may be simple. Contact an epidemiology or biostatistics faculty member at a local academic center. For physicians already at an existing center, this should be no problem. I trained at the University of South Florida where the School of Public Health is less than 500 yards from the School of Medicine, the School of Nursing, and the local Veterans Affairs hospital. Biostatistical academicians still must "publish or perish." Graduate students need experience and, more importantly, curriculum vitae exposure. For physician researchers at nonacademic centers, this issue is bit more complicated. Funding may be an issue, but I think physicians would be surprised at how far an offer of authorship would go.

References

1. Moss, M, Wellman, A, Cotsonis, GA (2003) An appraisal of multivariate logistic models in the pulmonary and critical care literature. Chest 123,923-928[Abstract/Free Full Text]

This Article Full Text (PDF) Free Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Article Archive Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Barnett, S. D. Search for Related Content PubMed PubMed Citation Articles by Barnett, S. D.