TUCSON, Ariz., Sept. 08, 2016 (GLOBE NEWSWIRE) -- “Science” is said to be dictating “evidence-based” policy with far-reaching effects on American life, but seriously flawed methods could lead to disastrous results, states Jane Orient, M.D., managing editor of the Journal of American Physicians and Surgeons. The fall issue scrutinizes such methods, with specific application to affordable electricity, advanced X-ray imaging, and personalized medicine.
We are doing it all wrong with regard to everything having to do with evidence, according to a book that is reviewed in the issue, Uncertainty: The Soul of Modeling, Probability & Statistics, by William Briggs. “We are suffering from a crisis of over-certainty,” Briggs states. We place far too much faith in meaningless (or misunderstood) statistical parameters. We also rely on invalidated models—e.g. global climate models. The “true and only test of model goodness,” Briggs writes, “is how well that model predicts data, never before seen or used in any way. That means traditional tricks like cross validation, boot strapping, hind- or back-casting and the like all ‘cheat’ and re-use what is already known as if it were unknown; they repackage the old as new.”
Briggs debunks many logical fallacies, with special attention to what he calls “the epidemiologist’s fallacy” because it is so common and so harmful. It occurs whenever an epidemiologist says “X causes Y” but never measures X. The most egregious example that he mentions is the assertion that a tiny increase in small particulates called PM2.5s (dust) causes enormous numbers of deaths. This is being used, Dr. Orient notes, for the U.S. Environmental Protection Agency’s Clean Power Rule, through which the coal industry, which fuels 40 percent of our electricity-generating capacity, is being destroyed.
Epidemiological studies that instill fear of radiation from CT scans are analyzed by Dr. Bobby Scott. He states that there is no evidence for excess cancer from low-dose radiation (less than 100 mGy) from diagnostic imaging. Claims of predicted risks are in fact “phantom” risks linked to forced application of the linear-no-threshold model and extrapolating to low doses, he points out. His analytic method and tables to correct for sampling variability and bias are broadly applicable to epidemiologic studies of other environmental exposures, states Dr. Orient.
“Evidence-based medicine” (EBM), which is largely applied epidemiology, supports the “guidelines” that increasingly dictate medical treatment. Its flaws are so serious that even its former ardent proponents are beginning to recognize that EBM is in crisis, writes Dr. Hermann W. Børg.
“Policy is increasingly justified by politicized, outcome-based science,” states Dr. Orient. “Citizens need to learn to scrutinize the methods.”
The Journal of American Physicians and Surgeons is published by the Association of American Physicians and Surgeons (AAPS), a national organization representing physicians in all specialties since 1943.