May 3, 2004 The Oregonian by JIM BARNETT

Two years before she confirmed the first U.S. case of mad cow disease, Agriculture Secretary Ann Veneman assured concerned consumers that her agency had halted the infection at the nation's borders.

Evidence, she told news reporters Nov. 30, 2001, came from a study by the Harvard Center for Risk Analysis showing the risk of mad cow, or BSE, entering the country was "extremely low." Her assertions became the foundation of major USDA decisions dealing with mad cow.

But the agency-sponsored study did not assess efforts to keep the disease out of the United States as Veneman said, The Oregonian has found. Researchers with the Harvard center, part of the university's School of Public Health, had spent two years trying to measure the risk that mad cow might enter the United States before concluding they could not.

"It was a hard thing to do; it was a hard thing to think about, how it might get in," George Gray, lead author and executive director of the Harvard risk center, said in an interview. "So we just said, 'Let's assume it gets in.' "

In making her assertions, Veneman turned attention away from prevention of a potential health threat, critics say, while banking on odds that if mad cow disease were to infect U.S. herds, its spread could be halted in a few years.

Despite the discovery of an infected Holstein in Washington state last December, Veneman's statement still guides U.S. policy. The Harvard study was Exhibit A for an agency plan to restart cattle imports from Canada after mad cow was discovered there last May.

President Bush met Friday with Canadian Prime Minister Paul Martin and pledged his support for reopening the border "as soon as possible."

"It's in our nations' interest that live beef be moving back and forth," Bush told reporters on the White House lawn. "It's also in our interest to make sure that we make decisions based upon sound science."

Two questions For years, mad cow was mainly an issue of animal health. But it exploded into a human health crisis in 1996 when doctors linked dozens of deaths in the United Kingdom to the consumption of tainted meat.

At the time, U.S. agriculture officials assumed agency policies had kept mad cow out of the domestic herd, agency records show. But the news from overseas made them realize they knew little about the disease and its threat to U.S. citizens.

So agency officials commissioned the Harvard study in 1998 to focus on two questions: How might the disease enter the United States? And what happens if it gets in?

Catherine Wotecki, under secretary of food safety in the Clinton administration, said agency officials regarded the questions as equally important, because the risk of entry and the risk of spread might require different responses.

"Knowing what I know about these types of diseases, you need to have a twofold level of assurance in place," said Wotecki, now dean of the College of Agriculture at Iowa State University.

Gray launched his study with plans to address the first question by assessing "tens or even hundreds of pathways" by which the infection might enter the United States, according to agency notes from a public meeting Sept. 28, 1998.

As Gray's team searched for answers, the researchers found a world of unknowns. The United States had banned imports of cattle and byproducts from the United Kingdom and Europe by 1997. But foreign governments had no data tracking illegal shipments, so the researchers could not measure whether U.S. defenses were adequate.

"Harvard didn't assign risk to the pathways in," said Linda Detwiler, a senior agency veterinarian who advised Veneman until August 2003. "They just assumed it came in. And we were very cognizant that they did that."

With the blessing of agency officials, the Harvard researchers shifted focus to the second question. They studied the U.S. "feed ban," which prohibits use of rendered cattle protein in feed, and they developed a computer model to mimic its impact on herd health.

Although the model did not measure the risk of entry as Veneman indicated, it did produce a breakthrough conclusion: Even if mad cow entered the country undetected, the feed ban would halt spread of the disease and eliminate it within 20 years.

But where science was uncertain, Veneman said: "I am pleased to report that using complex mathematical models, the study found that the risk of BSE entering this country is extremely low."

Two warnings As Veneman unveiled the Harvard study for reporters, the General Accounting Office was conducting its own investigation of mad cow. But the GAO, the investigative arm of Congress, took a different approach.

The GAO looked for cracks in the regulatory wall -- and found plenty. Its January 2002 report contradicted Veneman, concluding in part that "federal actions do not sufficiently ensure that all BSE-infected animals or products are kept out."

Veneman shot back, saying the GAO "fails to appropriately recognize" findings from the Harvard study. Again, she offered mischaracterization of those findings.

"The Harvard Risk Analysis showed that the risk of BSE occurring in the United States is extremely low and that early government protection systems have been largely responsible for keeping BSE out of the United States," Veneman said on Feb. 26, 2002.

In October 2002, Veneman received a second warning -- this time from scientists that her agency had hired specifically to review Harvard's work. The scientists seized on what appeared to be an offhand observation in the Harvard study -- that the United States imported millions of cattle each year from Canada and Mexico. But as the scientists pointed out, U.S. trade regulations gave agency officials no way to guarantee that the imported cattle were disease-free.

"Maybe they don't pose any risk, but what if they had been fed contaminated starter ratios as calves in Mexico?" the scientists' report said. "Even if they would not live until patent clinical stages, they would introduce infectivity into the system."

The report would prove prescient. Ultimately, Veneman's agency would reveal that the first U.S. case, discovered in Mabton, Wash., in December 2003 was a Holstein raised on tainted feed in Canada and imported legally into the United States.

Two views Veneman declined requests to be interviewed for this article. Lisa Ferguson, senior staff veterinarian who responded on Veneman's behalf, said agency officials never intended to mislead the public about the Harvard study's conclusions.

"I don't think we've ever said Harvard quantified the risk that BSE would enter the United States," Ferguson said. "I believe our statements are supported by what is in the document."

Critics disagreed.

Veneman and her deputies repeated her mischaracterization of the Harvard study's conclusion with such frequency and conviction over a two-year period that many of their most knowledgeable critics assumed it was true.

Carol Tucker Foreman of the Consumer Federation of America acknowledged she was among them. After reviewing Veneman's statements and Harvard's conclusions recently, Foreman said she felt duped.

"Whether intentional or not, the statements are misleading because the study did not find what she clearly said it had found," said Foreman, an assistant secretary of agriculture under President Carter.

Sen. Dick Durbin, the Illinois senator who requested the GAO study, said he was concerned that Veneman's statements obscured the health risks posed by mad cow. As a result, the health of the U.S. beef industry also was jeopardized, he said.

"If we don't deal with food safety forthrightly, honestly and comprehensively, it not only endangers American consumers, but it's going to be destructive with our global position in terms of trade," Durbin said.

Two options But Veneman's version of the Harvard study continues to shape the USDA's response to mad cow.

On May 20, 2003 -- six months before the first U.S. mad cow case was found -- Canadian authorities revealed they had found a case in Alberta. Veneman was forced to halt U.S. imports of Canadian cattle, costing U.S. beef processors an estimated $455 million a year.

Food-safety advocates urged Veneman to safeguard U.S. consumers by adopting new meat-processing rules -- rules shown by Harvard's computer model to reduce the risk of spreading mad cow disease into the food supply.

Veneman took a different tack. On Oct. 31, 2003, she proposed reopening the border to Canadian cattle. Her agency cited an update of the Harvard study that said consumers would be safe, but again failed to assess risk that mad cow might enter from Canada.

"The continued protection of the U.S. food supply is our top priority," Veneman said. "This proposal reflects a thorough review of the scientific evidence, which shows the risk to public health to be extremely low."

Fifty-three days later, on Dec. 23, Veneman announced mad cow was in the country.

Jim Barnett: 503-294-7604; jim.barnett@newhouse.com