n December 23, 2003, the brain-wasting disease called Mad Cow hit home in the United States. As word of the death of a single cow spread, it tested the resolve of the federal officials who made the announcement and put beef industry leaders on edge. Suddenly, Mad Cow--a cousin of sheep scrapie that enters the gut and creeps up the spinal cord into the brain--was inside U.S. borders, heightening consumers' fears and threatening beef sales and trade. National attention turned to a 2001 analysis by the Harvard Center for Risk Analysis (HCRA) for the federal government, which asked: What would happen if Mad Cow were to invade the American herd?
HCRA Executive Director George Gray spent his Christmas break fielding phone calls from the news media. Patiently he repeated HCRA's findings: The risk of spread in the herd was low, and in humans, "minute."One cow--a Canadian import, it turned out--was not cause for panic. Days later, Gray would also meet on Capitol Hill with Ulrich Kihm, the Swiss member of an international committee on Mad Cow, to allay his concerns by explaining HCRA's study.
Despite the low risk, the United States Department of Agriculture (USDA) proposed to cut the Mad Cow hazard further, taking cues from HCRA's report. Clearly, more testing, more monitoring, and tighter controls on potentially infectious bovine tissues would be needed to calm a wary public.
New policies guided by HCRA's risk analysis, added to government regulations enacted in previous years, have helped corral the Mad Cow threat in the U.S. Yet questions remain: How much regulation is enough, and at what cost to industry and consumers? The case of Mad Cow shows the critical role of science in decision making--and its limits for assuaging unease.
CALL FOR ACTION
It was in 1998 that USDA officials called HCRA's offices regarding Mad Cow, or bovine spongiform enceph-alopathy (BSE). Assuming the disease managed to enter the U.S., how might it impact the herd, 100 million strong? Would an outbreak lead to an epidemic, as it had in Great Britain, where it had emerged in the mid-1980s and nearly destroyed the beef industry?
BSE was now on the wane there, thanks chiefly to a halt in the modern practice of feeding processed, protein-rich cattle tissues, unfit for humans, back to cattle. But the damage was done: Nearly 180,000 cattle had been diagnosed with the disease. Some 50 people had died from a novel human form of BSE identified in 1996, variant Creutzfeldt-Jakob disease (v CJD). And trust in government health and agricultural agencies was badly eroded.
The U.S. began enacting regulations to guard against BSE in 1989. Cattle imports from Britain and other BSE-afflicted nations were banned. In 1997, the recycling of cattle tissues into cattle feed was halted. Were these steps enough?
"We knew we had the aptitude to quantify the threat--its probability and magnitude," remembers Gray, "but we knew nothing of the biology of this risk." To assist, he recruited Silvia Kreindel, a veterinarian and epidemiologist who studied livestock diseases, and Joshua Cohen, an applied mathematician who could make sense of data and use computers to simulate a BSE outbreak. An MIT graduate student, Keith Duggar, would do the programming for Cohen.
The team's learning curve would prove steep: Almost nothing quantitative on BSE had ever been published. "I'd hoped to go to a library," says Cohen now, laughing at his naiveté.
"People had useful data--in their desk drawers," Gray explains. To understand the U.S. agricultural system, the key to any meaningful risk analysis, he says, "We needed a crash course."
HUNTING FOR DATA
HCRA's team began by tapping experts for anything and everything they knew about BSE. The three traversed three continents, conducting 150-plus interviews with veterinarians, infectious disease specialists, cattle industry representatives, zoologists, faculty at universities from Tuskegee to Oxford, public health practitioners, and government officials.
But to estimate the chance that BSE proteins, called prions, might find their way into the beef supply and spread through the herd, the researchers had to see how cattle were raised and processed. So off they went--to dairy farms, cattle markets, slaughterhouses, meatpacking plants, cattle-feed manufacturers, and rendering facilities, where inedible materials were turned into tallow and bone meal. Also on the itinerary was the National Veterinary Services Laboratory, where cattle brains were tested for BSE.
As Kreindel explains, BSE is similar to scrapie, an infection that has sickened sheep for over 200 years. Unlike scrapie, however, BSE is "probably manmade," she says, thought to have originated in the late 1970s or 80s from the British practice of feeding scrap sheep protein to cattle. (The U.S. had begun feeding cattle proteins to cattle in the late 1960s.) Bizarre though it may seem, it makes biological and economic sense to recycle animal proteins that humans can't use, Kreindel says. Using these nutrients to fortify animal feed not only boosts beef and milk production, but eliminates an immense waste-disposal challenge.
Certain questions were pivotal to HCRA's risk assessment. Given that the infectivity in animals increased over time, how did the ages at which beef and dairy cattle were slaughtered impact BSE's potential to spread? Which tissues did people eat? How likely were inspectors to miss a sick animal? How well was the feed industry following U.S. Food and Drug Administration (FDA) rules?
At each step in meat production, HCRA's team estimated the odds that infectious tissue might contaminate the system. They studied processes by which spinal cords were stripped from carcasses, for example, and noted that one stun gun used to kill animals humanely fired too forcefully, propelling brain matter, along with prions, into the circulatory system.
Data collection was "intense," says Cohen. There were moments, too, that he laughs at now--like the day he blanched at the sight of raw cow brain, then stopped eating red meat for almost two years.
IN SEARCH OF R0
Thirty-six months later, in the spring of 2001, the HCRA team was ready to test a computer model that could simulate the state of the U.S. bovine agriculture system over several decades. Hypothetically introducing from 1 to 500 infected animals, what course would BSE take? Running the program at night on desktop machines, they crunched the numbers. And held their breath.
was almost disappointing. Nothing happened," Cohen says now. But
"nothing" was a good thing: Even if 500 diseased animals entered
the country, the model predicted that BSE would affect at most a few hundred
more cattle, then die out within about 20 years. The key to measuring
the disease's potential to spread lay in a fundamental epidemiologic parameter
called R0. If R0 were greater than 1, each cow would infect more than
one other on average, thus propagating disease. If R0 were less than 1,
the infection would peter out.
And the human risk? "Slim to none," says Gray. While the data were insufficient to calculate R0 in humans, the HCRA team extrapolated from Britain's experience, where the ratio of sick cattle to sick people was roughly 5,000 to one.
Using a measure of infectivity known as a cattle oral ID-50, the dose that could be expected to make 50 percent of cattle that ingest it sick, HCRA figured that the number of cattle oral ID-50s in human food in the U.S. would be small, given existing regulations--at most 170 units over 20 years. By contrast, several million units had entered Britain's food supply, affecting approximately 100 people. Clearly, Mad Cow did not jump to humans readily.
BSE IN PERSPECTIVE
HCRA's risk analysis was reconfirmed on the heels of the first infected cow's discovery. Available online at www.hcra.harvard.edu/publications.html, the findings have undergone extensive peer review.
What more could, or should, be done to protect the U.S.? Since December, the USDA has tightened controls, taking added precautionary measures, including some identified by HCRA's study. The FDA has stepped up enforcement of the ban on using cattle parts in feed. Inspections of slaughterhouses, rendering plants, and feed mills have been "beefed up." High-risk tissues like brains and spinal cords have been banned from human food, as have all lame (nonambulatory or "downer"), potentially diseased cattle. Also proposed is a ban on recycling high-risk materials in pet and animal feed, a costly measure considering the problem of disposing of this waste.
Still, Americans feel a twinge of unease whenever BSE rears its head. In February of 2004, an opinion survey by HSPH's Robert Blendon, professor of Health Policy and Management, found that 41 percent had "some" or "very little" confidence in the nation's beef inspection system. Although humans in the U.S. had yet to be affected, 27 percent of Americans thought they had--an indication that public perceptions of BSE risk do not match reality.
Likewise, perceptions of HCRA's analysis aren't always accurate. "We've been criticized for downplaying Mad Cow's chances of entering the U.S., when in fact our model assumed it would," Gray says. "What we show is that BSE won't spread significantly, as it has in other countries."
Nonetheless, since December, Japan, South Korea, Mexico, and other nations have banned American beef imports. To satisfy these customers, some U.S. meat companies are lobbying for the right to test all cattle. U.S. regulators resist, arguing that universal testing would push up meat costs with no gain in safety, while consumers might shun untested beef. Moreover, the test is generally not accurate in cattle under 30 months of age.
To eat U.S. beef, or not to eat? The answer depends on people's tolerance for uncertainty. In a sign that Americans still crave beef, Wendy's, Outback Steakhouse, and other restaurant chains in 2003 reported a hike in beef sales. This, despite the first U.S. death in June of 2004 of a woman who had developed v CJD after coming to the states from England.
Meanwhile, HCRA continues to fight BSE fears with facts. George Gray met with the Japanese as part of a U.S. team seeking compromise on the BSE issue. He has also begun thinking about extending HCRA's risk analysis to the U.S.'s North American trading partners, Canada and Mexico.
Can the BSE risk be brought to zero? At what cost? There will always be tension among the public, government, and an industry reluctant to act against ill-defined threats for no clear gain. Says Gray, "The question for the public health system is: Where do we draw the line?"
Karin Kiewra is Editor of the Harvard Public Health Review and Associate Director of Development Communications at the Harvard School of Public Health