The testing problems mean it may never be known with certainty whether these animals were infected with the deadly disease. Department officials said these animals were not included in the agency's final tally of mad cow tests, but the records, obtained by UPI under the Freedom of Information Act, indicate at least some of them were counted.
The USDA Inspector General, in a report made public last month, rebuked the agency for a slew of problems in its mad cow surveillance plan that could have reduced the chances of finding a positive case in U.S. herds. The only confirmed case of the disease in the United States is a cow that tested positive last December in Washington state.
Consumer groups say the untested cows are evidence of further problems with the USDA's surveillance plan.
"This adds to the clear documentation in the Inspector General's report that the program is in shambles, from its design to its implementation to its record keeping," Dr. Peter Lurie, deputy director of Public Citizen's Health Research Group in Washington told UPI.
Lurie testified about problems with the USDA's surveillance plan at a congressional hearing last month.
In total, 486 animals in 2002 and 2003 went untested or the wrong portion of their brain was collected, according to the USDA documents. The mad cow pathogen accumulates in a specific region of the brain called the obex.
More than 200 of the animals were not tested at all. Almost all of the rest had the wrong part of their brain sent to the USDA's National Veterinary Services Laboratories in Ames, Iowa, that conducts mad cow tests, with a few unable to be tested because the sample was unsuitable.
"Somebody must be asleep at the switch if they can get this kind of data in their database and not launch some kind of investigation," Lurie said. "It suggests inappropriate collection of samples and failure to test even those samples that were collected, on top of the failure to test categories of animals USDA told the American public it was testing."
USDA spokeswoman Julie Quick told UPI, "None of those (untested animals or ones with the wrong portion of the brain collected) were counted in official sampling."
In addition, Quick said, the USDA has safeguards in place to prevent possibly infected meat products from reaching consumers.
The concern is humans can contract a fatal brain disorder called variant Creutzfeldt Jakob disease from eating beef products contaminated with the mad cow agent.
The Inspector General's report faulted the agency for failing to install screening procedures to ensure it gets access to the cows most likely to be infected with the deadly disease. This includes those that test negative for rabies and those with signs of a disorder of the brain or central nervous system, also known as CNS cows.
The USDA's testing records indicate even when the agency obtained samples from these animals, many went untested or the wrong region of the brain was collected.
In 2002, four CNS cows and 11 rabies-negative cows were not tested or the wrong part of the brain was collected. In 2003, the testing problems appeared to increase, occurring with 10 CNS cows and 38 rabies-negative animals.
Nearly all of these animals were older than 30 months -- many were 6 to 8 years old -- which puts them in an age range most likely to test positive if they were infected. The USDA's current surveillance plan focuses almost exclusively on animals in this age range because they are old enough for the slow-incubating disease to have reached detectable levels if they are infected.
In addition to the CNS and rabies negative cows, the 486 untested and incorrectly sampled animals included other risky groups, including 310 downers -- those unable to stand -- 79 dead, eight sick and five condemned. About 182 of the animals were 30 months of age or older, with many of them as old as 8 years.
Quick said the explanation for the untested animals is "most likely the sample was autolyzed (deteriorated) and we weren't able to run a sample, so we wouldn't count that."
The type of mad cow test used by the USDA does not work well with autolyzed samples, but another type of test called Western blot, which is commonly used in Europe and Japan, can yield accurate results with degraded tissue samples.
In the cases where the wrong brain region was collected, the agency ran a test "in the event that it might have contained some of the correct tissue," Quick said. These samples were "fully expected" to be negative because the sample was not from the correct part of the brain and thus they were not included in the official tally, she said.
However, it would have been necessary to include some of the untested animals in order to arrive at the USDA's final tally of 19,990 animals tested in fiscal year 2002, as stated in a Jan. 15, 2003, news release.
The USDA data obtained by UPI consists of 20,132 mad cow tests conducted in 2002. Subtracting the 209 animals not tested or with the wrong brain region collected that year drops the total to 19,923, or 67 less than the total given by USDA Secretary Ann Veneman. A comparable analysis is not possible for 2003 because UPI only has records for the first 10 months of that year.
Michael Hansen, a biologist and senior research associate with Consumers Union, the watchdog group in Yonkers, N.Y., called the USDA's testing problems with the high-risk CNS and rabies-negative cows "outrageous" and said "it raises some serious questions" about the agency's surveillance program.
"We'll have to wait and see in the next couple of months what the final Inspector General's report says and whether some of the problems with sampling are ever going to get cleared up," Hansen told UPI.
Although UPI initially received the testing records last January, the USDA refused to release a key that would help decipher the meaning of several obscure codes and acronyms. After several more months of requests to obtain the key from both the Freedom of Information Office and the agency's Animal Plant and Health Inspection Service, the USDA finally released the key in late May, but insisted it was not legally required to do so.
Several of the 486 animal tests showed peculiar results. Nineteen rabies-negative animals from various unnamed facilities in New York all went untested on the same day, June 13, 2003. "Other" was listed under the column for results of a histopathology screening -- a test that cannot diagnose mad cow disease conclusively, but can indicate abnormalities consistent with the disease.
The USDA has not explained what "other" means. The key provided to UPI says "other" means "other."
It was unusual, in the data provided to UPI, to have any results listed under the histopathology, as this column is blank for most of the 35,000 animals contained in the records obtained by UPI.
"See comments" was listed in the histopathology column for another rabies-negative animal that went untested, but the USDA did not include any comments in the records.
On Feb. 24, 2003, a rabies-negative cow from an unnamed Kansas facility could not be tested accurately because the wrong portion of its brain had been collected. The undefined description "inflammatory" was listed under histopathology.
On average, 22 animals per month went untested or the wrong portion of the brain was collected, but some months were worse than others. In July, 2002, these problems occurred with 106 cows.
Certain dates and companies appear repeatedly. On one day, July 25, 2002, 48 downers from Batlar Enterprises in Wisconsin were not tested. Earlier that month, 24 downers from two plants in California were not tested.
A couple of weeks later, on Aug. 5, 2002, 17 downers from Midway Meats in Washington state -- which processed meat from the infected cow in December -- went untested. The next month, on Sept. 5, 2002, 20 dead animals from ETI in Georgia were not tested.
"When you look at all that, then how do you expect the American consumer to have any confidence in this, or our trading partners?" Lester Friedlander, a former USDA veterinarian, told UPI.
More than 60 countries have closed their borders to U.S. beef due to the December mad cow case.
"Where are the Senate and congressional Agriculture committees? How come they don't say anything?" Friedlander asked.
He also questioned how the untested animals and the repeated collection of the wrong brain region went unnoticed at the USDA.
"Somebody (there) should've caught on, but instead it has to be somebody from UPI under the Freedom of Information Act," Friedlander said. "How did it get by so many people?"
Steve Mitchell is UPI's Medical Correspondent. E-mail firstname.lastname@example.org