The U.S. Department of Agriculture's records of mad cow screenings, conducted on 35,000 animals between 2001 to 2003, also reveal no animals were tested for the past two years at Vern's Moses Lake Meats, the Washington slaughterhouse where the mad cow case was first detected.
In addition, no mad cow tests were conducted during the two-year period at any of the six federally registered slaughterhouses in Washington state. This includes Washington's biggest slaughterhouse, Washington Beef in Toppenish -- the 17th largest in the country, which slaughters 290,000 head per year -- and two facilities in Pasco that belong to Tyson, the largest beef slaughtering company in the United States.
In 2002, nearly every test conducted in Washington was on animals from Midway Meats in Centralia, the packing plant where Vern's Moses sent the infected cow carcass. The meat was distributed to several states where some people apparently consumed it, raising concerns about the possibility of contracting the human equivalent of mad cow, an always fatal, brain-wasting condition known as variant Creutzfeldt-Jakob disease.
The USDA said the meat posed little risk to consumers because the most infectious parts -- the brain and spinal cord -- had been removed.
The testing records, obtained by UPI under the Freedom of Information Act, which the USDA delayed releasing for six months, also show a number of other gaps in the agency's national surveillance strategy for mad cow disease, including:
-- Tests were conducted at fewer than 100 of the 700 plants known to slaughter cattle.
-- Some of the biggest slaughterhouses were not tested at all.
-- Cows from the top four beef producing states, which account for nearly 70 percent of all cattle slaughtered each year in the United States, only accounted for 11 percent of all the animals screened.
-- Though dairy cattle are considered the most likely to develop mad cow, some of the top dairy slaughtering plants were sampled only a few times or not at all.
-- The test tally for 2003 includes more than 1,000 animals ages 24 months or less, which would not test positive for the disease on the test used by the USDA even if they were infected. Many of these animals displayed signs that could indicate mad cow disease, including being downers or unable to stand, and symptoms suggesting a possible brain disorder.
"I can't believe that," Felicia Nestor, food safety program director of the whistleblower organization the Government Accountability Project, in Washington, D.C., said of the USDA's lack of testing in Washington.
Nestor questioned why the USDA would not implement more testing after the finding of a case of mad cow in Alberta, Canada, in May of 2003, in a close border state such as Washington. The records show after May and through July, however, no commercial cows in Washington state were tested.
"It's right near Alberta ... and everybody knows a lot of cattle cross over the border from Canada into the United States," Nestor told UPI. Approximately 1.7 million Canadian cattle entered the United States in 2002.
GAP has followed the mad cow surveillance program closely for several years and Thursday is to release statements from current USDA inspectors, who said the surveillance system is not administered uniformly across the country. In some cases, the inspectors said, the plant personnel -- not USDA veterinarians -- are in charge of selecting which animals go for testing.
"In the interest of transparency, USDA needs to answer the obvious questions raised by these findings," Nestor said.
USDA spokesman Jim Rogers told UPI some states, such as Washington, may not get tested during some periods of the year because the agency's system is based on sampling from eight regions of the country rather than each state.
Asked if the agency tries to sample from all slaughterhouses, Rogers said, "Not necessarily." Some plants do not take downer cattle so the USDA will not conduct much, if any, testing at these facilities because the agency wants to target the high-risk animals, he said.
In addition, Rogers said, the samples taken each year by the USDA are adequate to detect mad cow if it is present at the rate of one-in-a-million animals.
"As long as they take the required number of samples, they're OK," he said.
Nestor argued the failure to screen any animals for a two-year period at Vern's in Moses Lake, Wash., where a Holstein cow tested positive for mad cow on Dec. 22, raises questions about the ability of the mad cow surveillance program to focus on cows most vulnerable to the disease.
Vern's Moses Lake is known for slaughtering older and injured dairy cows, which are considered the cattle most at risk of developing mad cow disease. The cow that tested positive was a 6.5 year old dairy cow.
Many of the top dairy slaughtering plants around the country either do not appear in the testing records at all or are listed only a couple of times.
Dairy cattle often are given feed supplemented with animal protein to enable them to produce the vast quantities of milk required in today's mass dairy operations. Ranchers in the United Kingdom incorporated cattle tissue into their cow feed because it was a cheap source of protein. This is thought to have contributed to the spread of the mad cow epidemic that hit the country in the 1980s because some of the cattle turned into feed were infected.
Although that practice has been banned in the United States, the ban did not go into effect until 1997 and several feed firms have been and are still in violation. Because cows infected with mad cow disease can take as long as six years before they show symptoms, this raises the possibility that animals infected before or after the feed ban were processed at slaughterhouses such as Vern's Moses Lake, yet were never tested or detected by USDA's surveillance program.
The top four dairy slaughtering states -- Wisconsin, California, Pennsylvania and Minnesota -- slaughter 1.7-million dairy cows each year, or about 67 percent of the 2.6 million butchered annually in the United States. Yet tests were conducted on only 11,794 animals from these states, or less than one-half of 1 percent of the more than 3.4-million dairy cows processed in these states over the two-year period. The actual percentage may be even lower because, presumably, not all of the animals tested in those states were dairy cows.
In Wisconsin, the state that slaughters the most dairy cattle, the bulk of the testing focused on just two firms, with few coming from the four slaughterhouses in the state that process dairy cattle. Only three animals over the course of two years came from Packerland Packing in Green Bay, the fifth largest slaughterhouse in the country, which slaughters some 260,000 dairy cow each year.
Nestor questioned the rationale behind USDA's apparent strategy of ignoring the large beef companies and targeting efforts at smaller plants.
"It's really significant that they're focusing all of their attention on the very smallest plants," she said. "It's almost like the USDA wants to protect the big plants from a finding because the implications would be too scary. If they find a case at a small plant, the USDA can then say it's an isolated problem" and infected meat wasn't distributed all over the country or internationally as might happen with a larger plant, she said.
The findings "don't inspire confidence," said Michael Hansen, senior research associate with the watchdog group Consumers Union in Yonkers, N.Y., who has monitored the government's mad cow strategy closely for a number of years. "It really makes one wonder how useful the whole program is," Hansen said.
Former USDA veterinarian inspectors told UPI the problems with mad cow surveillance may go back more than a decade.
Lester Friedlander, who worked for the agency from 1985 to 1995, said in the early 1990s, when he was the USDA's chief veterinarian inspector at Taylor Foods in Wyalusing, Pa., the plant with largest number of downer cows at the time -- about 25 to 30 a day -- the USDA never asked him for a single brain to test for mad cow.
"From 1991 until the day I left on Aug. 28, 1992, they never asked me for one brain," Friedlander said, noting the facility should have been key in a national mad cow surveillance program because cows from several different states were processed there.
Tom Damura, who spent 12 years with the USDA, said for almost an entire year -- from Feb. 2000 to Dec. 2000 -- no animal was tested for mad cow at an Iowa Beef Packers facility in Amarillo, Texas, where he was the agency's veterinary medical officer.
In addition, due to staffing shortages he was often unable to perform inspection on the animals before they were slaughtered to see if they displayed the staggering and swaying that can indicate mad cow disease. That job fell to an employee of the plant.
The USDA inspectors contacted by the Government Accountability Project will offer similar accounts on Thursday. Some work at the largest slaughtering plants in the country and say they have seen hundreds of downer cows, yet only one or two are ever tested.
The mad cow testing records for 2002 and 2003 appear to support their allegations.
Only about 1,500 animals combined were tested at the top 10 slaughterhouses, which slaughtered nearly 60 million of the 70-million animals killed in the last two years.
In contrast, more than 5,900 animals -- or nearly four times as many as at the larger plants -- were tested from just six slaughtering plants classified by the USDA as small or very small, including one firm with fewer than 10 employees that was sampled 2,011 times over the two-year period.
Of the top four beef producing states -- Nebraska, Kansas, Texas and Colorado -- which slaughter some 24-million cattle or nearly 70 percent of the total annual U.S. slaughter, only 3,694 animals or about 11 percent of all animals tested in the last two years originated from these areas.
Scant testing was done at the top five slaughter companies -- Tyson, Excel, Swift, Farmland National Beef and Smithfield -- which combined slaughter more than 100,000 cows per day, accounting for 78 percent of the U.S. beef industry and $97.3 billion in annual sales.
Only two cows were tested at Excel's facilities in Colorado in 2002, and in 2003, only two were tested from its Nebraska facility.
Only one animal was tested from a Tyson facility in Illinois in 2003 and this was a 24-month-old cow with signs of an unknown brain disorder that probably was too young to test because animals generally do not show up on the test used by the USDA until they are 30 months of age or older. But recent cases in Japan -- which uses a different kind of test than the USDA -- indicate cows as young as 24 months and even younger can carry the disease.
The USDA maintains it tested approximately 20,000 animals in fiscal year 2003, which ended Sept. 30, but the records suggest that may be an inflated number. Through July, the agency had tested 15,139 animals, but this included more than 1,000 animals age 24 months or younger.
Randall Levings, chief pathologist at the USDA's National Veterinary Services Laboratories in Ames, Iowa, which conducts all the agency's mad cow tests, told UPI those animals are not included in the yearly totals because they are considered too young.
Rogers offered a different perspective, however, saying the total test tally should reflect "all the testing done regardless of the age of the animal."
Excluding those animals would drop the total tested to approximately 14,100 through July, meaning the agency would have needed to test approximately 3,000 animals each of the last two months of the fiscal year to have achieved the 20,000 total. Even if the younger animals are left in, it still means the lab would have needed to run 2,500 tests each of the last two months.
Either number would be a Herculean feat, because the most cattle the lab had ever tested in a one month period during fiscal year 2003 was approximately 1,900 in October, 2002.
Even if the lab had been able to pull off such a large number of tests over the two month period, the validity of the results might be questionable since the test it uses, something called an immunohistochemistry test, has to be meticulously prepared to ensure accuracy and is both time and labor-intensive.
Markus Moser, a molecular biologist and CEO of the Swiss firm Prionics, which manufactures rapid tests for detecting mad cow disease, said getting valid results on an IHC test will limit the speed at which they can be processed. The test can be processed faster but doing could sacrifice accuracy, he added.
"It's like reading a book sentence by sentence vs. just browsing through the pages," Moser said.
Arthur Davis, chief of the biopathology lab at NVSL, told UPI the staff focused on mad cow testing -- about six technicians and eight pathologists -- could process up to 200-300 IHC tests per day, and perhaps more if necessary.
The USDA withheld the results for the tests conducted in 2003 in the documents it provided to UPI -- which covered the period Jan. 1 through July 24 -- but officials said all were negative for mad cow. On Monday, the agency's Freedom of Information Office said it would provide the test results, probably by Tuesday.
As of late Wednesday, UPI still had not been given the results.
Steve Mitchell is UPI's Medical Correspondent. E-mail firstname.lastname@example.org