Data from two dozen industry tests that intentionally exposed people to poisons, including one involving a World War I-era chemical warfare agent, are being used by the Environmental Protection Agency in approving and denying specific pesticides.

The controversial data come from 24 human pesticide experiments submitted to the EPA by companies seeking pesticide permits. The data, provided by the EPA to congressional officials, is being studied under a policy the Bush administration adopted last November to have political appointees referee on a case-by-case basis any ethical disputes over human testing.

Aides to two California Democrats, Sen. Barbara Boxer (news, bio, voting record) and Rep. Henry Waxman (news, bio, voting record), compiled and reviewed EPA data on 22 of the cases.

"Nearly one-third of the studies reviewed were specifically designed to cause harm to the human test subjects or to put them at risk of harm," the aides concluded in a 38-page report and accompanying documents provided Wednesday to The Associated Press.

The report said scientists conducting the experiments "failed to obtain informed consent (and) dismissed adverse outcomes," adding that the tests "lacked scientific validity."

One study in 2002-2004 by University of California-San Diego researchers administered chloropicrin, a soil insecticide that during World War I was a chemical warfare agent, to 127 young adults. Trade-name products for it and mixtures of it — such as Timberfume, Tri-Con, Preplant Soil Fumigant and Pic-Chor — must carry a "danger" warning label.

Most were college students and minorities who were paid $15 an hour to be put in a chamber or have the vapor shot into their nose and eyes after signing consent forms warning they should anticipate "some irritation in the nose, throat and eyes that could be sharp enough to cause blinking and tearing."

"Because you will be participating in an experiment, we must apprise you that there may be some risks that are currently unforeseeable," the consent form read.

However, doses 120 times the hourly limit established by the Occupational Safety and Health Administration were ingested by the test subjects, according to the congressional aides' report.

Another study dosed eight people with the pesticide azinphos-methyl for 28 days, and everyone reported headaches, abdominal pain, nausea, coughing and rashes, the report said.

Boxer said the report "proves the Bush administration is encouraging dangerous pesticide testing on humans with no standards," despite the EPA's new policy.

EPA spokeswoman Eryn Witcher said Wednesday that the agency "values the importance of the scientific and ethical issues surrounding human studies and is expediting a public rulemaking process to comply with a federal court decision."

The U.S. Court of Appeals for the District of Columbia ruled in 2003 in a suit brought by the pesticide industry that the EPA cannot refuse to consider data from manufacturer-sponsored human exposure tests until it develops regulations on it.

Agency officials said last November that a new rule on human testing data would be issued by 2006, and until then each study would be looked at and accepted unless it is fundamentally unethical or has significant deficiencies.

Human tests, in the view of pesticide makers, provide more accurate results than those using animals about the risks of the products to people and the environment. The companies that use them say they follow safety guidelines set by Congress, EPA, courts and scientific groups.

The EPA for decades used industry studies gathered from human tests to help set pesticide exposure levels. Officials say they still accept the data but don't rely on it for their decision-making.

Last year, the National Academy of Sciences recommended that the EPA establish a human studies review panel to look at such studies, both before and after they're conducted.

On the Net:

EPA: http://www.epa.gov