WASHINGTON -- Officials at the Environmental Protection Agency who are supposed to evaluate scientific studies warning of potential health dangers now say the agency has overlooked such warnings on at least 10 chemicals.
One of those warnings concluded that metam sodium, the pesticide spilled last month in the Sacramento River after a train derailed in California, caused severe spinal and brain birth defects in laboratory animals.
The EPA said yesterday that it had received a study from the manufacturer that the chemical could cause birth defects in rats in 1987 but had never reviewed it.
For three weeks after the spill, California environmental officials told nearby residents that the chemical did not pose a serious health risk.
But after reviewing reports from a state file on chemical companies' trade secrets, the state officials found two studies describing the risk to lab animals.
On Aug. 6, they warned pregnant women in Dunsmuir, Calif., the town closest to the spill, to be tested to determine whether their babies had been harmed.
One study was sent to the agency in 1987, an agency official said yesterday. Another study had been sent to federal environmental authorities in July 1990.
But Linda J. Fisher, the top EPA official responsible for pesticide regulation, said that her agency had overlooked the studies until she ordered a review after the state warning was issued. The review found that warnings on 10 to 20 chemicals had been similarly overlooked.
Under a provision of the federal pesticide control law, the EPA is supposed to review such studies as they come in and if appropriate, issue new regulations that could range from banning to new labeling requirements.
In an interview this week, Douglas D. Campt, the director of the Office of Pesticide Programs, acknowledged that the early warning system promised by the agency had broken down.
In the last three years, according to the agency's review, manufacturers had sent the EPA 154 notices of adverse health or environmental effects for 75 different pesticides.
Complete reviews of studies indicating adverse effects had not been conducted for 10 to 20 chemicals, a number that is changing as the agency proceeds with its investigation.
So far, the most important notices of adverse effects that were missed by the Office of Pesticide Programs involved metam sodium, a potent poison applied by farmers to strip soil of microbes, insects and weeds before planting.
On July 14, 19,000 gallons of metam sodium was spilled after a derailment in northern California, killing thousands of fish and all aquatic life in 45 miles of the Sacramento River.
When it mixes with water, metam sodium changes chemically into methyl-isothiocyanate, the poisonous byproduct that makes it such a potent pesticide.
In the days after the spill, Dr. Richard J. Jackson, of the the California EPA, said residents in Dunsmuir exposed to fumes from the spill were in no danger.
To make certain, Dr. Jackson directed one of his staff members to investigate thoroughly the health studies submitted to the state by the chemical's five manufacturers.
The staff member discovered two studies in the state's confidential files that showed pregnant rats and rabbits exposed to high levels of the byproduct had offspring with abnormalities of the spinal cord and brain.
On August 6, Dr. Jackson reversed himself and announced that there could be a long-term effect. Dr. Jackson also called a top assistant to Ms. Fisher at the EPA and told him about the data.
The EPA said yesterday that it had received the rat study in 1987 and the rabbit study in July 1990 and did not read either of them until after Dr. Jackson's call.