A third and more sophisticated test on the beef cow suspected of having mad cow disease (search) would have helped resolve conflicting results from two initial screenings, but the U.S. refused to perform it in November.

That additional test, ordered up by the Agriculture Department's internal watchdog, ended up detecting mad cow -- a finding that was confirmed on Friday by the world's pre-eminent lab, in England.

Only 18 months ago, the department had used the Western blot test to help uncover the first American case of the brain-wasting illness in cows.

The department is pledging that, from now on, it will conduct such testing on suspicious animals.

U.S. officials in November had declared the cow free of the disease even though one of two tests -- an initial screening known as a rapid test -- indicated the presence of the disease. A more sophisticated follow-up -- immunohistochemistry (search), or IHC -- came back negative.

"They had two diametrically opposed results which begged to be resolved," said Paul W. Brown, a former scientist at the National Institutes of Health who spent his career working on mad cow-related issues.

"If you had what they had, you would immediately go to a Western blot and get a third test method and see which one of the previous two was more accurate," Brown said.

Consumer groups and scientists urged the department to perform a Western blot test and seek confirmation from the lab in Weybridge, England.

In a letter to Consumers Union last March, the department said there was no need for the British lab to confirm the results and that the Western blot test would not have given a more accurate reading.

"We are confident in the expertise of USDA's laboratory technicians in conducting BSE testing," wrote Jere Dick, an associate deputy administrator. Mad cow disease is medically known as bovine spongiform encephalopathy, or BSE.

Troubled by the conflicting test results, the department's inspector general, Phyllis Fong, ordered the Western blot test this month. By the time an aide notified Agriculture Secretary Mike Johanns, the testing was under way.

The testing was positive. The department then sent tissue samples to the British lab, which subjected the samples to all the tests again.

Johanns, amid an uproar from the cattle industry, was irked that she did so without his knowledge or consent.

"From my standpoint, I believe I was put there to operate the department and was very disappointed," he told reporters Friday morning.

By that afternoon, the verdict from Britain was in: The cow had mad cow disease.

In humans, a form of the disease -- variant Creutzfeldt-Jakob disease (search) -- has been linked to the consumption of contaminated meat. The rare but fatal disease has killed about 150 people worldwide, mostly in Britain, where there was an outbreak in the 1990s.

Johanns, who took over the department in January, said the government will use both the IHC and Western blot tests from now on when initial screening indicates an animal may have the disease.

The world's leading animal health organization recognizes both the IHC and Western blot tests.

The department's testing program, put in place after the first U.S. case of mad cow disease in 2003, calls for IHC tests when rapid screening indicates mad cow disease is present. The department uses rapid tests on about 1,000 cows a day.

The department had used Western blot tests since the 1990s to resolve conflicting results, including on the first case. But since then, the department has used the Western blot only if samples from an animal were too degraded to work for the IHC.

The current testing program "might not be the best option today," Johanns said. "Likewise, the protocol we develop as a result of this testing might not be the best option in 2007," Johanns said. "Science is ever evolving. It is not static."

Critics are saying the department should have gotten it right in the first place.

"They were afraid the truth would come out," said Carol Tucker Foreman of the Consumer Federation of America. Added Michael Hanson of the Consumers Union: "This is just foot-dragging, and these delayed reactions need to really stop."

In the IHC method, a thin slice of the brain -- thinner than Saran wrap -- is stained to highlight the mad cow protein and then examined under a microscope.

In a Western blot test, a larger portion of the brain is ground up and concentrated to pull the mad cow protein closer to one area, then treated to eliminate normal protein. It is placed in a tray of gelatin. An electric current is run through the gelatin to separate the proteins. The proteins are applied to paper and stained for scientists to examine.

The IHC test was negative when the department's lab in Ames, Iowa, performed it on tissue from the cow last November.

The lab also ran the sample through an experimental version of the IHC test. The tissue looked abnormal but, since the experimental method is not scientifically verified, the Ames lab did not report the result, Johanns said.

The IHC test was positive when the British lab ran the test last week.

Scientists said that because the tests vary from country to country, it is not unusual for results to vary as well.

"There are no two laboratories around the world that are using identical IHC methods and not a single test that you can take off the shelf," said Danny Matthews, a scientist at the British lab.