CLARIFICATION

CLARIFICATION; Contrary to a notice in yesterday's Magazine, staff writer Peter Perl will not be fielding questions about food safety today at www.washingtonpost.com/liveonline. He will be doing so at noon Tuesday. (Published 01/17/2000)

With loud blasts of steam, the Ball Park franks came shooting out of the peeling machines like bullets from a machine gun. The relentless wave of pinkish meat cascaded onto a fast-moving slotted conveyor strip that the workers called "the piano belt." Their fingers flew from slot to slot, picking out occasional discolored or malformed hot dogs, as if they were playing the piano.

When everything was working perfectly, the 1,600 men and women at the giant Bil Mar Foods meatpacking plant in western Michigan cranked out 1 million pounds of frankfurters and Sara Lee premium deli meats a day, seven days a week. National brands such as Ball Park and Hygrade are produced here, helping make Sara Lee Corp. the largest purveyor of packaged meats in the country and Bil Mar its largest producer.

But everything was not working perfectly at Bil Mar.

Looming overhead, above the piano belt, was a giant refrigeration unit that kept the plant cold. Bil Mar was a modern, clean facility, but the hot dog area was the oldest part of the plant, and its cooling unit looked like an enormous old car radiator. It had been patched and repaired, yet it continued to leak and cause condensation, a breeding ground for bacteria.

The moisture trouble had gotten so severe that inspectors from the U.S. Department of Agriculture took the rare step of shutting down the plant for several days in November 1997. Management repaired and replaced faulty equipment, reengineered production lines and assigned additional sanitation duties to remedy the problem.

Now it was back. "Fluid was dripping onto exposed franks," a USDA inspector reported on the morning of March 12, 1998, "and these franks were being conveyed down the line to the packing machine." A second inspector cited Bil Mar that same day because he found beaded droplets of condensation falling from overhead pipes and the ceiling, landing in the cooked food.

So over the July 4th weekend, during a rare production shutdown, construction workers with chain saws came to tear out the old refrigeration unit. They cut it into pieces small enough so they could haul it out of the plant. Replacing the old unit, plant officials told USDA, should go a long way toward curing the problem.

Indeed, the condensation eased for a while. But another problem emerged: bacteria. The company took weekly swab samples from environmental surfaces such as slicers, conveyors and packing equipment that came in contact with food. During the following six weeks, 11 of 12 samples tested positive for bacteria, compared with only 3 of 12 in the preceding time period. The readings remained high for two more months -- until November 1998, when Bil Mar stopped testing altogether.

The plant's management stepped up its sanitation efforts. Workers put iodine tablets into floor drains, took apart machinery and sprayed surfaces with citric acid and other disinfectants, used hand sanitizers and foot dips on their boots to prevent them from tracking bacteria near the food.

Meanwhile, the condensation persisted. By the end of 1998, USDA inspectors had written up Bil Mar for at least 45 "Noncompliance Records" that specifically cited condensation or water dripping throughout the plant. Each citation carried a standard USDA warning: "Further failure to implement effective corrective and preventive actions may result in additional regulatory and administrative actions."

No such action was taken by USDA -- not until it was too late.

Nobody paid much attention, outside family and friends, to the death of Helen Bodnar on October 19, 1998, in Memphis. Health officials did not know enough then to attach much importance to the demise of a 74-year-old woman, a wife and mother of four who loved to eat Ball Park franks.

Nor was any particular notice taken of the death of Gloria Andrzejewski, a 56-year-old church secretary who died November 22 in Detroit.

Or of Christopher Ntukogu, age 31, a teen counselor in Columbus, Ohio, who ate a lot of deli sandwiches and died on December 24.

Or of Irene Wood, a 75-year-old woman who lived alone in Upstate New York and died on Christmas Day.

Or of the case of Lisa Lee, a pregnant 27-year-old video store manager from Columbus, who survived a severe flu-like illness, but watched helplessly as it killed her premature twins last January.

By the time it was over, the food poisoning outbreak that originated at Bil Mar had swept across 22 states, killing at least 21 people and seriously sickening at least 100 others from Tucson to Baltimore, according to the Centers for Disease Control and Prevention. It was the most lethal case of food-borne illness in the United States in 15 years.

The culprit was Listeria monocytogenes, a rare bacteria that is far less common than its better-known biological cousins -- salmonella and E. coli -- but far more deadly.

When the listeria outbreak was finally pinpointed by the CDC, Sara Lee officials on December 22, 1998, voluntarily recalled an estimated 35 million pounds of Bil Mar's hot dogs and deli products. It was possibly the largest meat recall in history -- and one of the quietest.

The Bil Mar case attracted relatively little public attention partly because the news media were riveted on the impeachment drama of President Clinton and on the U.S. bombing of Iraq, while the public was additionally preoccupied with the holiday season.

But it was also because the Agriculture Department -- in contradiction of its own policies -- failed to issue a press release informing the public of the danger, even though the CDC already had reported four deaths. Instead, USDA relied on an announcement issued by Bil Mar that did not mention the full scope of the recall or the dire nature of the illness, let alone the deaths.

As a result, many consumers continued eating Bil Mar products and the death toll climbed -- to 14 by January 20, 1999, nearly a month after the recall. That day, after the first lawsuits had been filed against the company on behalf of the victims, Sara Lee took out full-page advertisements in newspapers across the country to further publicize the recall, although these announcements, too, made no mention of the deaths. Finally, on January 28, USDA issued a press release, more than a month after federal officials had first confirmed that eating the contaminated meat could be fatal.

Because of the Bil Mar case, the listeria threat is now considered so serious that USDA and the CDC issued a warning last May that consumers whose health is vulnerable -- including pregnant women, the elderly, cancer patients, diabetics and those with AIDS -- should refrain from eating hot dogs or deli meats unless they fully recook them. Yet this warning, which USDA is not requiring on product labels, has also received relatively little public attention.

Food-borne diseases in America cause an estimated 76 million illnesses annually, resulting in 325,000 hospitalizations and about 5,000 deaths, according to a new study by the CDC. It is an epidemic that costs upwards of $30 billion in medical expenses and lost productivity, by government estimates. The U.S. food safety system is currently undergoing its biggest changes in a century, as USDA, which oversees meat and poultry, and the Food and Drug Administration, which regulates most other foods, phase in new science-backed inspection systems that for the first time require testing for the most common microbes, salmonella and E. coli. Advocates and critics alike agree the new system is generating promising results, particularly in the nation's slaughterhouses.

But the Bil Mar outbreak illustrates some crucial weaknesses in USDA's oversight of another key sector: food-processing plants.

A politically powerful industry, a compliant Congress and a timid federal bureaucracy have fashioned a system that still has glaring problems:

Lack of Regulatory Power: Government can order the recall of products like unsafe cars, toys and insecticides -- but not food. Health officials must rely on persuading companies to voluntarily recall tainted meat. Agriculture Secretary Dan Glickman has sought mandatory recall authority in Congress, but has failed partly because of the lobbying clout of the food industry.

Bureaucratic Conflicts: Although one set of government officials at the CDC pinpointed Bil Mar as the source of the outbreak, another set at USDA refused to request a recall. Agriculture officials, worried about the legal implications of wrongly accusing Sara Lee, did not believe the CDC had sufficient evidence to merit a recall. While USDA continued waiting for more than a week for definitive laboratory results, company officials were prodded to act by an angry Minnesota state health official and by the CDC.

Long Delays: While USDA now requires microbial testing in slaughterhouses, it has not yet mandated similar testing for food-processing plants. Even after the 21 deaths, the department has left to companies' discretion the decision on whether to test, while it researches whether to enact a mandatory rule, a process that will take up to three years.

Lack of Penalties: USDA's new inspection system, like the old, contains no system of penalties for repeated violations. USDA has sought the power to issue fines and civil penalties from Congress, which has refused. Its only enforcement tool is to shut down plants, an option it rarely uses.

The U.S. attorney's office in Grand Rapids, Mich., has convened a grand jury to investigate the Bil Mar deaths, and USDA's Office of Inspector General is also investigating. Thomas Billy, administrator of USDA's Food Safety and Inspection Service, said he could not discuss the specifics of the case but defended the department's actions, saying, "We do not have the regulatory tools" to have moved any more swiftly or decisively.

Meat industry officials said that while listeria is a serious threat, the Bil Mar outbreak is an aberration. "This went through the industry like a lightning bolt," said Dane Bernard, a vice president of the National Food Processors Association. "People realized there was a critical need to reemphasize old lessons" about the dangers of bacteria. But Bernard and other spokesmen said they do not believe Bil Mar signals a need for any fundamental changes in government oversight.

Sara Lee officials declined to cooperate with this article. "We were completely forthcoming" with government investigators, said company spokeswoman Theresa Herlevsen. "We don't conclusively know the reason for the outbreak." She noted that since the Bil Mar incident there have been more than 40 recalls nationwide for listeria. "We think it is unfortunate that we continue to be the poster child for listeria," she said. "More than 90 percent of our customers have come back, which is a tribute to the strength of our brand and the trust in our company."

Herlevsen's comments suggest one of the paradoxes of the food safety issue. Companies have powerful economic incentives to avoid the legal actions and stigma that invariably accompany a food poisoning outbreak. The science of bacteria detection is more refined than ever, and consumers, with access to reams of instant information, are arguably more knowledgeable than ever. Yet at the same time, the system has never seemed more vulnerable. With more food being imported from around the world, faster production lines and wider mass distribution, critics contend government agencies are overwhelmed because of outdated practices and insufficient tools.

The recent food safety reforms were sparked in large part by the public outcry following the 1993 Jack in the Box case, in which an outbreak of E. coli 0157:H7 at West Coast restaurants of the fast-food chain killed four children and sickened hundreds. Food safety advocate Caroline Smith DeWaal of the Washington-based Center for Science in the Public Interest said the 21 deaths in the Bil Mar case should be a wake-up call that further improvements are needed. "It is clear in this outbreak," she said, "that these failures probably cost lives."

In his cramped basement cube at CDC headquarters in Atlanta, a slender, bespectacled 41-year-old physician and epidemiologist named Paul Mead was completing work on a new survey on the nationwide impact of his medical specialty -- food-borne diseases -- when he received the first official notification: four listeriosis cases in rural southern Tennessee. They appeared to be only a local cluster of illnesses, requiring routine follow-up laboratory tests.

That was on October 20, 1998. Over the next few weeks came similar reports. Five cases from Ohio. Nine from New York. One from Connecticut. There was no way to tell immediately whether these were related, and Mead, a soft-spoken, methodical scientist, was not a man to jump to conclusions. The CDC would await bacterial lab cultures from victims' blood samples. Meanwhile, the CDC faxed questionnaires to the four states to get detailed food histories from the victims. By mid-November, as more cases trickled in, Mead found himself slowly moving toward what he called his "red-alert mind-set."

Mead knew enough about this particular disease to be very afraid. Listeria monocytogenes is among the deadliest and most elusive of pathogens. The species is found throughout nature, in animals, fish, plants, water and soil. It is psychrophilic, or cold-loving -- it is not only invulnerable to refrigeration but thrives in it. It can endure in the body more than eight weeks before sickening a victim, so that it is unusually difficult to determine from a patient's dietary history just which food was the outbreak source.

Mead also knew that listeria kills much more efficiently than most pathogens. Most healthy people's white blood cells will engulf and eliminate a listeria colony. But among the roughly 50 million Americans with some form of immune problem, the battle turns the other way: Listeria will invade and kill the white blood cells, permeate the blood, then find and attack the weakest links in the body's armor. It often lodges in the brain, causing meningitis; it also can find its way into the placenta of pregnant women. If caught soon enough, listeriosis is treatable with antibiotics, but it kills at least 20 percent of its victims.

Mead knew from his own disease survey that listeria sickens about 2,500 people a year, killing about 500. He recalled just how dangerous the disease can get: In 1985 in Southern California, poor sanitation in the production of Mexican-style soft cheese sickened several hundred predominantly Hispanic consumers and killed at least 48.

But this outbreak, if his fear was correct, would not be confined to one region. With modern methods of mass distribution and refrigeration, listeria monocytogenes could spread for weeks over a broad geographic area, possibly the whole country. Years ago, meats spoiled fairly quickly and were discarded long before listeria could reach lethal levels. But improved sanitation and preservatives have created products with much longer shelf life, so listeria colonies often have 50 days or longer to reproduce and grow.

On November 20 when lab results came in, Mead and his CDC colleagues convened a conference call with health officials from the four initial outbreak states. Using the most sophisticated scientific technique on the bacteria cultures grown from the blood samples, the CDC had determined that 18 victims had been attacked by the identical strain of listeria. The lab analysis, yielding a genetic fingerprint similar to DNA sampling, had pinpointed something the CDC would never otherwise have been able to determine so quickly: All the victims had eaten something from a common source.

But what was it? The CDC sounded a nationwide alarm via fax and e-mail to 50 state health directors and epidemiologists, asking them to review all recent medical surveillance data for any increase in listeria. Mead also urged the four outbreak states to redouble their efforts at getting complete food histories so they could pinpoint the source of the outbreak strain, which the CDC named "Type E."

By December, the outbreak had spread to Massachusetts, Michigan, West Virginia and Oregon. Mead sent another e-mail to state epidemiologists, telling them the latest finding: A preliminary statistical analysis of victims' food histories showed a strong association between those with Type E and those who remembered eating hot dogs.

As December progressed, with at least four listeriosis victims known dead, the accompanying data piled up in Mead's 9-by-12-foot cinder block work space. Up to 20 CDC scientists, lab technicians and statisticians were working overtime. By December 14, Mead and his colleagues were able to compare the food histories of 20 victims of Type E and 21 patients with other, unrelated strains. Certain brand names appeared statistically linked to Type E, including Ball Park, Bryan and Kahn -- all made by Sara Lee. The CDC also obtained its first sample of leftover hot dogs, from the refrigerator of a suburban Cleveland woman who contracted listeriosis and passed it to her newborn. They were Ball Park brand, and they came from a package that bore the stamped USDA establishment number code: P-261.

Bil Mar.

Mead remembered then that the previous month in New York, the rare Type E strain had unexpectedly shown up in a leftover package of sliced chicken breast. He hadn't paid much attention because he'd been focused on hot dogs, not deli meat. But now the New York finding suddenly loomed larger: The outbreak strain had been found in a Sara Lee deli product that bore a USDA stamped code.

P-261.

Knowing that at least four people were dead and 37 seriously ill, Paul Mead dispatched a CDC investigator to Bil Mar on December 15. That same day, Mead and his boss sent a memo to USDA headquarters, informing the agency that, based on preliminary findings, the CDC believed the outbreak was coming from Bil Mar.

Six of 10 victims who could remember brand names said they had eaten Sara Lee hot dogs and another recalled eating Sara Lee deli meat, the CDC said. In addition, both meat samples from the Ohio and New York victims had originated at the Bil Mar plant. "We conclude that both hot dogs and deli meats produced by this company are likely to be contaminated with the outbreak strain," the CDC said. It requested from USDA information about product distribution and "anticipated regulatory action."

But USDA was in no rush to take action. The department's long-standing policy was not to even request a recall until it found the suspected adulterant in an unopened package of meat. Otherwise the contamination might conceivably have originated outside the meat plant, long after production, agency lawyers had said, and the department could be held legally liable for wrongly blaming the plant.

Mead knew the policy demanding unopened samples and assumed, incorrectly, that it was a legal requirement. On December 15, he also sent out an urgent e-mail to state epidemiologists seeking still more ammunition to help persuade Sara Lee. "It is not clear that the company will feel that the data is strong enough to support a recall at this time," Mead wrote. "WE DESPERATELY NEED ADDITIONAL PRODUCT INFORMATION: brand names, establishment numbers, dates of production for both hot dogs AND deli meats. We ask you to please redouble your efforts."

Meanwhile, at the plant in Michigan, the CDC's on-site investigator, Eileen Dunne, a 33-year-old physician, went to work. She was new to the CDC and had never been in a meatpacking facility. From the start, USDA officials made clear to Dunne that, occasional problems notwithstanding, they considered Bil Mar a well-run and modern operation with a solid record. The plant continued production, but Sara Lee temporarily postponed shipping products, pending the outcome of the investigation.

Federal investigators, donning down-filled coats, white smocks, boots and hairnets, scoured the entire plant for two days, scrutinizing the production process and taking swab samples from numerous surfaces. Preliminary results turned up negative, but the CDC was more concerned with what conditions had been like five months earlier, when the first known packages of tainted meat were produced. Mead asked Dunne to question Bil Mar management about any unusual events, particularly in early summer; they were both intrigued to learn of the July dismantling of the hot dog cooling unit.

Mead believed the refrigeration unit would have been a likely harbor for large colonies of listeria, and that cutting it apart could have unleashed bacterial dust. That would explain how the outbreak strain could have grown in both hot dogs and deli meats. Company officials, however, insisted that they had taken precautions to cover all the equipment and carefully clean the area.

Then Dunne learned from Bil Mar technicians about the swab testing. The tests give a quick indication of the presence of the listeria species of bacteria. In swabs that test positive, up to 40 percent of the bacteria colony can turn out to be the most virulent form, listeria monocytogenes, although a lengthier test is necessary to detect it.

Dunne faxed to the CDC several hundred pages of swab-testing data from throughout the huge plant. She also obtained from Bil Mar unopened samples of hot dogs and deli meat produced months earlier, and shipped them, packed in dry ice, to Atlanta. But the test for listeria monocytogenes would take about a week. Meanwhile, the list of victims continued growing.

Sara Lee, a company with $20 billion in annual revenue and a good reputation, was facing a corporate nightmare.

In recent years, there had been four recalls at Bil Mar involving shipments of meat that were undercooked, spoiled or contaminated by foreign objects. The recalls had generally gone smoothly with little negative publicity.

But now a company whose wholesome image was captured in its advertising slogan "Nobody Doesn't Like Sara Lee" was confronting the prospect of being identified as the source of a fatal nationwide outbreak that seemed to be spreading with each passing day. The recall, if it came, would be massive in size and potentially devastating in publicity. Yet the company was getting no clear guidance: CDC scientists would only say they were investigating and would not discuss the question of a recall. USDA regulators were assisting the CDC, but had no information of their own and no guidance to Sara Lee about what to do.

Suddenly, the decision took on greater urgency. Someone leaked the December 15 CDC memo to the Associated Press. On December 17 the news service reported that federal officials were investigating four deaths and dozens of illnesses, and that Sara Lee was awaiting results from government inspectors before deciding whether to initiate a recall. The story ran in newspapers around the country.

The next day, Sara Lee sent two lawyers to the Washington office of USDA's Thomas Billy, offering cooperation and seeking the department's guidance. But Billy and his deputy, Kaye Wachsmuth, proffered little help. In a tense but cordial hour-long meeting, USDA officials said they did not yet have the evidence to seek a recall and were awaiting further lab results.

Then toward the end of the meeting came a moment that seemed to sum up USDA's paralysis. Billy, a former biochemical engineer who sometimes speaks in a professorial tone, suggested that instead of a recall, Sara Lee might consider a "market withdrawal." This is a little-used option in which products are taken back from distribution points, but without any public announcement. Sara Lee officials were thoroughly puzzled by what seemed like an offhand suggestion in dealing with a fatal multi-state outbreak. Neither side pursued it any further.

The company sought help elsewhere. It called Michael Osterholm, then the Minnesota state epidemiologist, whose reputation as a bio-detective was based on high-profile investigations. His department was credited with tracking salmonella-tainted ice cream that sickened tens of thousands of people in 1994 and uncovering the link between toxic-shock syndrome and tampon use. A bearded and buoyant man, he was an outspoken health advocate with diverse interests, among them trying to swim the English Channel and serving as an adviser to King Hussein of Jordan on biological terrorism.

When a Bil Mar supervisor had called to seek his opinion on December 17, Osterholm already knew the details of the case from other state epidemiologists and believed Sara Lee should immediately start a recall. He agreed to take part in a series of conference calls over the next few days with Sara Lee executives, scientists and lawyers.

Osterholm was convinced that waiting would mean more illnesses and deaths. "I told them they had a serious problem and should do a recall, and they said, `Nobody in the federal government says we should,' and it was true, they were getting mixed signals," he recalled in an interview.

"USDA blew this one," Osterholm said. "They were not convinced it was true." With people dying, he said, "you cannot sit there and wait two weeks for the perfect data, and that is what USDA and CDC were doing."

Osterholm held several impassioned conversations with Mead and others at the CDC, urging them to be more definitive and forceful with Sara Lee. But Mead told him he had just reviewed the latest data in a conference call with several state epidemiologists who were not yet convinced they'd proved the source. Osterholm chided him, "Come on, Paul, you know this is it! You gotta cut through the [expletive]! You gotta make sure Sara Lee understands!"

While Mead held back, Osterholm went forward alone. He called Sara Lee President C. Steven McMillan, who had cut short a Caribbean vacation to return to Chicago headquarters on Sunday, December 20, because of Bil Mar. A recall would cost Sara Lee $75 million, at minimum, and company executives were torn between wanting to do the right thing and not wanting to jump to a conclusion that might needlessly damage Sara Lee's reputation and market share.

But Osterholm was convinced it was time to jump. "I said to Sara Lee that this was like driving down a highway and seeing a herd of deer," he recalled. "You can put on the brakes now and you'll still hit them. Or you can hit them at full speed and the deer will come through your windshield."

Over the pre-Christmas weekend, Mead retreated to his home with hundreds of pages of Bil Mar's swab tests, spread them out across his living room carpet and carefully catalogued the results. It was only then that he saw the pattern emerge clearly: a dramatic spike in bacteria readings at several places in the plant coinciding with the chain saw dismantling of the refrigeration unit.

That same weekend, the CDC lab came up with the results from the opened package of Bil Mar hot dogs from Ohio: Type E.

Final lab results on the crucial unopened hot dog samples would take several more days. But Mead believed it was time to give Sara Lee the official word. On Monday, December 21, he sent the company a three-page report of the CDC's findings.

The CDC recommended a recall of all hot dogs and deli meats produced since July 4. It also recommended Bil Mar resume environmental swab testing to prevent future illnesses.

Outlining the mass of circumstantial evidence, the CDC said it believed contaminated food was still sitting in supermarkets and refrigerators across the country. "By the time definitive proof is available," the report warned, "more illnesses and deaths will likely occur."

The next day, Sara Lee announced it was recalling all meats produced at Bil Mar since July 4.

The December 22 press release under the Bil Mar letterhead was decidedly understated. The company said it was voluntarily recalling "specific production lots of hot dogs and other packaged meat products . . . The CDC has indicated it is studying whether some of these products might contain the Listeria bacteria." The announcement made no mention of any illnesses or deaths, and described the recall as a "precautionary measure." It gave no symptoms of listeriosis, except to say that it "affects primarily the elderly, pregnant women, newborns and adults with weakened immune systems."

USDA, even at this late date, still wasn't buying the recall. Unsigned memos sent to headquarters by the USDA team at Bil Mar questioned both the CDC's findings and Sara Lee's action. A December 21 memo stated it was "difficult to imagine" that listeria monocytogenes could have contaminated both hot dogs and deli meat because "listeriosis has never been documented to be transmitted in airborne particles." (That assertion is incorrect, according to disease experts.) On the day the recall was announced, USDA received a daily update from its staff at Bil Mar with this concluding assessment: "In the absence of further laboratory data implicating the plant, this recall action, which is the perogative [sic] of the plant, may be premature."

Once a recall begins, USDA guidelines specify that press releases are to be issued whenever the information will help consumers avoid products that can cause illness. In this case, though, officials decided not to issue one.

"This was rather unprecedented," said Wachsmuth, because the recall had been initiated without USDA involvement. "We did not have that link to say that it was adulterated, and I guess that was part of the problem because that is the kind of press release that we put out. We say, `It is adulterated.' " Since USDA did not believe it could state that, it stated nothing.

Without any government announcement, and without mention of deaths or illnesses in the Bil Mar press release, the news media paid relatively little attention. The Chicago Tribune and USA Today carried brief front-page stories the next day. Other newspapers ran short wire service stories on inside pages. Televised news items on the recall were rare.

Over the next few days, Osterholm, who found recalled meat in his own refrigerator, grew incensed. "There was almost no publicity, and a recall depends on publicity," he said. In Upstate New York, for example, Osterholm learned, a Meals on Wheels program had continued to serve sandwiches of recalled Sara Lee deli meats to its elderly patrons because nobody had heard the news.

An irate Osterholm repeatedly called USDA and Sara Lee during the week between Christmas and New Year's, imploring them to put the word out more strongly. When he reached McMillan, he recalled, "Steve got upset with me. He said he had put the company through enough. He said, `We already had enough bad publicity.' "

Osterholm did not fare any better with USDA. He said it took several days to get his calls returned. When he finally reached Catherine Woteki, the USDA undersecretary for food safety, she declined to issue an announcement, saying she believed there had been adequate public notice. He said Woteki was not even aware that laboratory tests had finally confirmed the presence of the outbreak strain of listeria monocytogenes inside unopened packages from Bil Mar. That result had been confirmed at a CDC lab on Christmas Day; USDA didn't learn about it for another week.

News of the recall never reached Park Row Drive in Columbus, Ohio. There, almost every night, Lisa Lee, in the fifth month of her first pregnancy, would have a midnight snack of cold cuts with her fiance, Jonathan Southworth. They had met when they both worked at a Blockbuster video. Now, he was working long hours at two jobs and they were saving for a minivan because they were expecting twins. They bought whatever deli meats were on sale -- ham, turkey ham, turkey breast -- and didn't pay much attention to brand names.

It was an uneventful pregnancy until January 16, when Lisa felt feverish and dreadfully ill. She went to the emergency room of Riverside Methodist Hospital, where she ended up spending a fitful night. Doctors told her it appeared to be the flu and sent her home the next day. Four days later, her fever spiked to 104 and she was back in the hospital, struggling to save her babies.

She underwent more than eight hours of harrowing treatment, first trying to stop the labor and prevent a miscarriage, and then undergoing a lengthy induced labor. The boy was named Andrew Presston, the girl Alia Marie. They were delivered eight minutes apart, on January 20, both dead.

That same day, Sara Lee ran advertisements in some 80 newspapers around the country, calling attention to the fact that the recall was still in effect. The company acted after Kenneth Moll, a Chicago lawyer, filed a class-action lawsuit on behalf of victims, held a press conference to publicize the recall, and accused the company of negligence.

Lee and Southworth had continued eating deli meats for several weeks after the date of the recall. "If we knew it as soon as Sara Lee knew it, we might have had a chance," she said in an interview. "I want people to be more informed. I wish I would've been watching the news reports. I wish they would have publicized it better. I wish my doctor had been better informed."

Working two jobs, Southworth said, he rarely watched television or read newspapers, but frequently did the grocery shopping. "They should have a sign up saying, `We sold a product that could make you sick,' but no grocery store wants to admit that," he said. "If there was a notice posted, it might have helped us."

Lee and Southworth learned the full story only after the babies' deaths. A biopsy of the placenta showed the presence of listeria, and one of their friends visiting the hospital then remembered hearing about the recall. The couple said they were bewildered that neither they nor anyone who treated them in the medical system had had any knowledge of the danger.

After the babies were cremated, further testing on Lee's placenta showed that it carried the outbreak strain. The couple recently retained a lawyer and plan to sue Sara Lee.

"When this listeria thing happened, I felt really bad for them. We were always very careful to make sure it didn't happen," said Joseph Keller, a 53-year-old veteran of Bil Mar. "We took added precautions. We had a strong sense of pride." Keller, a supervisor in the hot dog plant, quit in early 1998 after 14 years because he was burned out from working 60 hours and seven days a week, he said. "I was making good money, but it was no life."

Keller, like a dozen other former workers, managers and USDA inspectors interviewed for this story, described the Bil Mar plant as a high-stress, high-turnover environment in which working conditions were difficult and the push for increased production was relentless. The work force, earning on average about $10 an hour, was predominantly recent immigrants from Southeast Asia and Mexico, most of them initially hired as temporaries, and the quality of supervision by management was spotty, according to interviews and company documents.

When Keller started at Bil Mar, the family of brothers Bill and Marvin DeWitt still ran the company they had founded in 1938. But he said the atmosphere changed dramatically after Sara Lee bought it in 1987. "It was a good company, but the switch from the DeWitts to Sara Lee, it lost the family feeling. It became just `How much money can you make for me?' . . . Nothing was ever good enough. Upper management just kept pushing and pushing, and you were only as good as your last production numbers."

Bil Mar previously ran five or six days a week, with two eight-hour production shifts, plus a third shift devoted solely to sanitation. Under Sara Lee, Bil Mar not only moved to a seven-day schedule, but lengthened production shifts to 10 hours, at the expense of cleaning time, according to Keller. "There was real head-butting between speed and cleanliness. We would run up to 20 hours a day, which left four hours to clean. They put added pressure on sanitation. I don't think it caused the problem, but it didn't help."

Listeria arose as an issue occasionally, when a particular department received positive bacteria readings. "They told people we had to sanitize more, and then they would push you for about a week, week and a half, and then everything would go back to normal," said Rick Lyttle, who worked 18 years on Bil Mar's production line. His wife, Vicky, also an 18-year employee, said managers reported in mid-1998 finding listeria in floor drains, but their warning did not appear dire. "I didn't take it as a real serious matter . . . There was concern when meat fell on the floor. But the possibility that it could kill someone? That was never mentioned."

Sanitation and other demands were also complicated by the language barriers. "I was running a line with seven people who could not speak or understand English," Vicky Lyttle said. To improve communication, Keller said, some company directives were printed in at least five languages, and the company offered language lessons to employees and supervisors.

But supervision was not always effective, as Bil Mar acknowledged in a memo to USDA on July 28, 1998. The department had criticized Bil Mar's sanitation, noting that repeat citations had been written for poor conditions in areas in the hot dog plant that "appeared to have been neglected for a significant period." In response, Bil Mar Vice President Mary Delrue wrote that the deficiencies were caused by "ineffective execution." She said five of the plant's top managers, including the sanitation and quality assurance directors, would change their work hours to take on increased responsibility.

Several former employees said supervisors would periodically attempt to sneak problems past both USDA and Bil Mar's own quality-assurance staff. Rick Lyttle and Paul Cherry, a former supervisor, said it was not uncommon to have meat falsely redated so it appeared fresher than it actually was, and to otherwise falsify records so that cooking and cooling temperatures appeared satisfactory, even if they were not.

Rick Lyttle said meat would sometimes be misplaced for several days in huge refrigerators and he would be instructed by a supervisor to change dates. Jerry Cannon, a retired USDA inspector, said Bil Mar had "significant" problems and that whistleblowers periodically alleged that the company had tried to hide unsanitary practices. But Cannon, who has volunteered to work with lawyers suing Bil Mar, said USDA over the years was only able to prove minor violations.

Bil Mar, in a February 13, 1998, internal memo to its top managers, acknowledged problems with the "integrity of company documention records" that are examined by USDA, and pledged to correct the matter. Inspectors had cited Bil Mar for failing to properly record the correct time of cooking and cooling temperatures, which are supposed to be carefully monitored for bacterial control purposes.

Delrue, in a memo to 14 supervisors, warned: "Falsifying company records is a very serious allegation which could ultimately lead to criminal prosecution . . . Bil Mar Foods does not condone and will not tolerate incomplete or inaccurate company records. Any future occurrences of this nature will result in immediate termination."

Delrue urged managers to try to reduce the number of citations Bil Mar was receiving and she concluded by saying: "We are very proud of the fact that we are the leading producer of meat products labeled under the Sara Lee brand name. We must do our utmost to protect this brand along with all of the products we produce by insuring they are manufactured under the highest quality and food safety standards possible. Quality Products made by Quality People!"

"To have people die on your watch and have a number of miscarriages, it is terrible. It is awful."

Catherine Woteki, arguably the nation's top food safety official, is sitting in her office in USDA's vast complex on Independence Avenue in Washington. It's December 1999 and USDA's undersecretary for food safety is pondering for a moment the meaning and consequences of Bil Mar. The department, she believes, did the best it could under existing law and regulations; the new system now in place is making significant progress. Her agency is building its public health capability and adapting to changes in technology that will allow enforcers to move far more quickly against future outbreaks. Still, for all of that, she acknowledges something went badly wrong in the Bil Mar case.

"In hindsight, it would've been the right decision to have requested a recall," she says. "With hindsight, yes, we should have put out a press release."

In the aftermath of the outbreak, various interested groups took familiar paths. Plaintiffs' lawyers filed multimillion-dollar lawsuits against Sara Lee. Consumer advocates called for stricter regulations. Industry spokesmen said that while they could never totally eliminate a problem as ubiquitous as listeria, they would redouble their efforts in research and sanitation. And USDA took no decisive action, but said it would reevaluate what to do.

One way to prevent future Bil Mars is for food processors to conduct regular tests for bacteria. Industry, government and public health advocates all agree on that much. But they differ over whether testing should be mandatory, and whether USDA or the industry should oversee the tests and decide how to act on the results.

The shock of the outbreak prompted a series of public and private meetings. At one such session last July, Carol Tucker Foreman of the Consumer Federation of America and Caroline Smith DeWaal of the Center for Science in the Public Interest prodded USDA's Thomas Billy to require mandatory testing.

Billy told them he was not ready to make that proposal, even on a temporary emergency basis. He said the department did not yet have enough research data to satisfy the Office of Management and Budget, which demands a detailed cost-benefit analysis for any new rule affecting businesses.

DeWaal was incredulous. "I said, `You mean that 100 people sick and 20 dead is not enough?' and he said, `No.' "

Billy, in an interview, said the federal rule-making process takes two to three years and that in the meantime he believed he could get quicker results by asking companies to voluntarily reassess their food safety plans to address listeria.

USDA last May sent all meat companies guidelines that recommend -- but do not require -- listeria swab-testing at the plants and, if the swab tests are positive, testing of the meat products themselves. The department also told companies that they must reassess safety plans to prevent listeria, but they are not required to make any changes unless they wish to.

USDA's own inspectors denounced this approach as "much too vague, confusing and contradictory," according to the National Joint Council of Food Inspection Locals. Foreman, a former assistant secretary of agriculture, called the department's response to listeria "the worst gobbledygook I have seen come out of this department" in recent years. "They explained it to me two or three times and I still don't understand it."

A coalition of consumer organizations has petitioned USDA to require listeria testing and "not await another food-safety disaster before acting." Said DeWaal, "More people are going to die, and we'll need another outbreak to tell us there's a problem."

A former meat industry executive agreed that the department needs to mandate environmental testing. "The USDA needs to be not so afraid of requirements," said James Marsden, retired president of the American Meat Institute Foundation, who is now a professor of meat science at Kansas State University.

To prevent listeria, Marsden said, the meat industry will have to remodel its post-cooking processes to create a germ-free environment, similar to the rigorous practices of pharmaceutical and milk producers. "It sounds extreme, but it is attainable," he said. Officials of the American Meat Institute and the National Food Processors Association said industry-funded researchers are already vigorously pursuing those and other solutions. Sara Lee said Bil Mar is now using similar technology.

The only government testing for listeria is a spot check that USDA has conducted since 1989 on a tiny percentage of cooked meats. Prompted partly by the 1985 fatal outbreak in Southern California, USDA enacted a "zero tolerance" standard for listeria monocytogenes. Nationwide, the agency takes about 3,500 samples of processed meats annually, which equates to two or three samples per plant per year. Since 1989, listeria monocytogenes has been found in up to 5 percent of certain types of cold cuts, which are recalled and destroyed. This spot-sampling program affects only a single work shift's production.

The "zero tolerance" standard and the fear of lawsuits have scared off some companies from doing voluntary listeria swab tests or product tests, according to industry officials, in what some call a "Don't look, don't find" policy. "Lawyers are telling people not to test your food and don't test your environment because you don't want results lying around" that could be used against you in court, said Dane Bernard of the National Food Processors Association.

At Bil Mar, employees told CDC and USDA investigators that the company stopped its swab testing in November 1998 because of a corporate decision by Sara Lee, according to CDC documents. It is not known what kind of testing, if any, has been done since the plant began reopening production lines last March.

The CDC's Paul Mead believes the Bil Mar case could be a watershed in controlling listeria because many companies were shocked that a large, reputable firm like Sara Lee could have such a disaster. "This outbreak is a tragedy, but it is lessened a little bit if it leads to new improvements," he said. "It's a tragic way to get there, but I hope there is a silver lining."

Nancy Donley is not so optimistic. She founded a consumer group called Safe Tables Our Priority, or STOP, after her only child, Alex, a 6-year-old, died in 1993 from eating a hamburger at Jack in the Box. In the Bil Mar case, Donley said, the firm, as a repeat violator, should have been heavily fined for sanitary infractions. "They were allowed to continue business as usual with nothing but a slap on the wrist, and it is about time that companies feel this in the pocketbook. Unless we change that, the public will continue to bear the burden of this."

Sen. Tom Harkin of Iowa, the ranking Democrat on the Agriculture Committee, said the Bil Mar case showed that USDA is not being aggressive enough in pursuing the listeria issue, and that its lack of enforcement powers "borders on the ridiculous." Harkin has submitted for the past two years a Safer Meat and Poultry Act to give the department stronger powers for recalls and civil penalties. But his legislation has gone nowhere.

The meat industry poured more than $41 million into political campaigns between 1990 and 1998, much of it aimed at members of the congressional committees that oversee agriculture, according to the Center for Public Integrity. Harkin said he hopes the Bil Mar case will improve his chances when he resubmits his bill in the next session, but he is not particularly optimistic. "The meat lobby," he said, "has a very powerful influence here."

Sara Lee is facing at least 10 lawsuits arising from the Bil Mar outbreak, which ceased shortly after the recall. The corporation says Bil Mar's new practices make it "the safest meat plant in the United States." Sara Lee has sought to rebuild consumer confidence, revamping an advertising campaign featuring basketball legend Michael Jordan and also publishing newspaper ads with coupons for free Sara Lee deli meats. Wall Street analysts said they did not think the outbreak would have lasting impact on the company.

Back at Bil Mar, the firm has resumed producing deli meats and has completely rebuilt its hot dog operation, preparing to resume production. Six months after the outbreak, USDA notified Bil Mar that based on its record of accumulated noncompliance reports -- it had 287 total citations in 1998 -- the plant's performance for food safety was considered inadequate and the department was moving to suspend federal inspection, thereby shutting it down.

But USDA rules allow it to suspend the suspension. Last June, Bil Mar was allowed to submit a new plan to correct its problems. So far, USDA has taken no action.

Peter Perl is a Magazine staff writer. He will be fielding questions and comments about this article Monday at 1 p.m. at www.washingtonpost.com/liveonline.

Anatomy of an Outbreak

NOVEMBER 1997

Bil Mar Foods plant in Zeeland, Mich., is shut down for several days by USDA because of condensation dripping in meat-processing areas.

JANUARY-JUNE 1998

USDA cites Bil Mar repeatedly for continued condensation troubles, including fluids dripping from pipes and ceilings directly onto cooked meats.

JULY 4TH WEEKEND

Workers with chain saws cut apart refrigeration unit in hot dog plant. Afterward, bacteria tests show elevated levels.

JULY 26-28

USDA cites poor sanitation, saying some areas "appeared to have been neglected for a significant period of time." Bil Mar pledges improvement.

OCTOBER 20

CDC receives notice from Tennessee of four listeriosis cases and begins investigation. More cases are reported from

Ohio, New York and Connecticut.

NOVEMBER 20

Lab results show 18 victims were attacked by single strain of listeria from common food source.

DECEMBER 3

CDC statistical analysis of victims' food histories shows strong link between listeriosis and eating hot dogs.

DECEMBER 14

CDC identifies three Sara Lee brands -- Ball Park, Bryan

and Kahn -- as likely sources of outbreak.

DECEMBER 15

With four victims known dead, CDC informs USDA and Sara Lee of preliminary finding that Bil Mar is source of outbreak. Bil Mar continues production, but stops shipping.

DECEMBER 18

Following news reports, Sara Lee meets with USDA officials, who do not recommend recall.

DECEMBER 21

CDC recommends recall to Sara Lee, warning if company waits for further lab results "more illnesses and deaths will likely occur."

DECEMBER 22

Sara Lee announces recall of all meat produced at Bil Mar since July 4.

DECEMBER 25

Lab tests of unopened package of Bil Mar hot dogs confirm outbreak strain.

1999

JANUARY 20

With death toll at 14, Sara Lee takes out newspaper ads reminding consumers of the recall.

JANUARY 28

USDA issues first press release, adding several brand names to list of tainted meats.

MARCH 3

Bil Mar resumes deli meat production.

MAY 25

USDA warns that 50 million consumers with immunity problems should not eat hot dogs or deli meats unless they fully recook them.