He then applied a series of…

July 2, 2010

He then applied a series of increasingly restrictive statistical “decision rules” to see how well the test handled the twin demons of false negatives (that is and the failure to detect a carcinogen) and false positives (registering an innocuous substance as carcinogenic).
His conclusion: “We have an assay test that may be wrong 50 per cent of the time”.
Regulatory policy on carcinogens under President Carter favoured erring on the side of safety.
The percentage of false negatives generated by a cancer test was therefore of great concern.
Yet Salsburg found that, of 19 known human carcinogens, only seven have been shown to cause cancer when fed to rodents. Given those odds, claims Salsburg, one might as well flip a coin. NCI judged 50 per cent of the 170 compounds to be carcinogenic. But Salsburg wondered how many of these might be false positives.
The direct way to assess this is to feed the rodents with large doses of substances that are known to be innocuous.
Salsburg says this was never done for the NCI test, which throws doubt on the high percentage of compounds found to cause cancer.
And Salsburg found that, when he applied statistical tests to the NCI data, between 3 and 70 per cent of the compounds either caused or protected against cancer, depending on the stringency of the decision rule used.
The National Toxicology Programme which is jointly run by the main federal agencies responsible for public health, is charged with developing and promoting toxicology tests in the US.
The programme’s deputy director, John Moore and says Salsburg’s critique contains “distortions and errors”.
The high number of positives is a result of NCI’s “conscious bias” toward selecting test substances that were suspected of posing a risk, Moore argues.
As for the huge variability in the number of contradictory results, Moore says that NCI applies only the most stringent criteria to their decisions: a significant turnout response to a given dose in two species or two sexes.
Moore dismisses Salsburg’s coin-flip analogy as “not meaningful”, but admits to inadequacies in the two-year feeding studies. Salsburg is not the first to criticise the rodent assay.
Several years ago and the Occupational Safety and Health Administration held hearings on the problem of identifying and classifying potential carcinogens. The scientists split into two camps over the validity of animal studies.
Advocates said they were reliable and indispensable; critics, however, disagreed on a number of points. California to ban research on stray cats
RESEARCH laboratories in California may soon be forced to abandon their use of stray pets for biomedical research.
University officials claim that a new bill before the Californian legislature will quadruple research costs and halt work altogether.

Juegosdiarios Aol partnerbörsen kostenlos

%d bloggers like this: