The Deadly Dinner Party: and Other Medical Detective Stories Read online

  As Riley and his team traced the meat back from McDonald’s, they found that one of the restaurant’s suppliers, a plant in Ohio, had saved some of the ground beef, ironically, as part of a quality control program. The same E. coli O157:H7 strain was found in that sample, which was from the same lot that ended up in the hamburgers the patients had eaten. This was groundbreaking science, although none of the patients from these two outbreaks developed HUS.

  This is where the work of Canadian researchers comes into the story. Dr. Mohamed Karmali, a microbiologist, was trying to put other pieces of this same puzzle together. In the summer of 1980, within a ten-day period, fourteen Toronto-area children were diagnosed with HUS—an extraordinary number of cases of such a rare disease in so short a time. Aware of the recent work by another Canadian microbiologist who had described a toxin in some E. coli that could play a role, Karmali tested these children for the presence of this toxin in E. coli, and he found it. Expanding the scope of his investigation, in 1985, he reported results of forty children with HUS and found evidence for this new toxin in most of them. He concluded, “Our findings indicate that [toxin-producing] E. coli have a close, and probably causal association with idiopathic HUS of childhood.”

  These outbreaks, especially because they were associated with such a prominent company as McDonald’s, hit the national press and were featured in large newspaper headlines. Some referred to it as “hamburger disease.” McDonald’s stock took a temporary hit, but it soon rebounded. Americans’ consciousness about developing HUS from eating hamburgers increased as well.

  The problem surfaced again in 1993 (two years after Emily’s case), once more raising HUS to national prominence when a large number of children in the Seattle area became sick after eating hamburgers from a Jack-in-the-Box restaurant. The outbreak came to light in January, when the Washington Department of Health was notified of a cluster of children with bloody diarrhea, and HUS was noticed in Seattle-area hospitals. In that epidemic, 602 patients developed bloody diarrhea or HUS, of which 477 had culture-confirmed E. coli infection.

  This outbreak was tightly clustered in time, with a peak incidence of illness between January 17 and 20; 144 patients required hospitalization, and 30 developed HUS. Three children died from the kidney disease. Investigators tracing the sources of the hamburger meat found that 73 Jack-in-the-Box restaurants were responsible; the beef had ultimately come from five slaughter plants in the United States and another one in Canada.

  In the aftermath of litigation that resulted from the epidemic, company officials admitted that they had been warned by local health departments that their hamburgers must be cooked to an internal temperature of 155 degrees Fahrenheit; however, they felt that this “overcooking” led to the meat having a tough texture. Numerous class-action and individual lawsuits were settled for many tens of millions of dollars.

  Emily’s case happened between these two large outbreaks. During this period, HUS had become one of the most common causes of intrinsic renal failure in children in the United States. For a specialist in childhood kidney disease, this is a relatively common problem, but it’s not one that most general pediatricians see with any frequency. One mark of a great doctor is the ability to know when to ask for help. Dr. Rok knew that his next move was clear: Emily would have to be transferred to a hospital in Boston where pediatric kidney specialists and dialysis facilities were available.

  “I was very worried about Emily, and I felt helpless in being able to be with her during the most frightening part of her illness,” remembers Michele, who was still having copious bloody stools and was tethered to an IV line. On the Friday after she got sick, Emily and her dad traveled by ambulance to Children’s Hospital in Boston. Her kidney function was failing. At the hospital, Emily had a catheter inserted in her bladder to measure her urine output, and she was put on intravenous fluids.

  The following day was one of the hardest for her father. “I don’t know which was worse,” Peter recalls. “Michele was still hospitalized and couldn’t be with us and Emily had to go the operating room to have a dialysis catheter placed just below her collarbone. It was all beginning to hit me.” Except for when she was in the operating room, Peter didn’t let Emily out of his sight for the next five days.

  As remarkable and unfortunate as Emily’s situation had become, it was made all the more extraordinary by events that were still unfolding back in Fall River. Dr. Rok had a second patient hospitalized at St. Anne’s. She was Margaret Carvalho, a vibrant twenty-two-month-old toddler with short light brown hair whose case seemed to be a carbon copy of Emily’s: bloody diarrhea followed by dehydration and then listless behavior. Margaret was hospitalized on Thursday, November 14. By Sunday, three days later, her kidneys were also beginning to fail.

  “So I was back on the phone to Boston,” recalls Dr. Rok, who, after not seeing a single case of HUS in his twelve years of practice, felt pretty sure that he was now dealing with his second case in as many days.

  The specialist at Children’s Hospital was Dr. Melanie Kim, director of the general renal program. On Wednesday, November 20, two more girls—sisters, and patients of another Fall River pediatrician—were transferred to Children’s Hospital with HUS. “This was very unusual,” recalls Dr. Kim, who, as a specialist at a regional referral center, generally sees about twelve cases of HUS each year, usually clustered around June and July. “So to see four cases from the same small town in November was quite striking. We notified the state Department of Public Health.” Besides the four girls from Fall River, there were already two children in Dr. Kim’s unit with HUS, one from northern New England and another from Boston.

  By this time, a key piece of information had emerged: stool cultures from three of the Fall River girls (including Emily and Margaret) turned positive for the E. coli serotype O157:H7 bacteria. It was happening again.

  Dr. Susan Lett is the director of epidemiology at the Department of Public Health in Massachusetts. “Knowing the relationship between E. coli diarrhea and HUS, we decided to investigate. I contacted the CDC in Atlanta for help.”

  Dr. Richard Besser was a pediatrician with the CDC’s Epidemiology Intelligence Service. “I had signed up with the enteric [intestinal] diseases branch because I thought I’d be on the road a lot working on various disease outbreaks. As it turned out, I had been doing desk work and fielding routine telephone calls that get directed to our branch. Being in Atlanta four months without an outbreak to investigate, I was a little stir crazy.”

  After receiving the call from Dr. Lett in Boston, Dr. Besser’s supervisor directed him to fly to Boston the next day. “I was champing at the bit,” recalls Dr. Besser. “I was very excited!” On Friday, November 22, he took the short flight to Boston. The next morning, he and Dr. Lett began the investigation. That weekend, they interviewed the families, spending several hours with each one. Dr. Besser recalls that “the families had already arrived at the tentative conclusion relating the illnesses to a frozen fish product. In fact, when we arrived at the hospital, people asked us if we were there to investigate the fish stick outbreak.”

  The epidemiologists launched a three-pronged investigation. The first task was to identify any other cases involved in the outbreak. Dr. Lett’s office sent out notices to area pediatricians, family practitioners, and medical laboratories. Simultaneously, Lett and Besser spent most of the first few days exhaustively interviewing the families to try to find the common denominator in the case. What had infected them with the E. coli0157:H7? And last, the third prong of attack, the investigators sent the bacteria isolated from each child to the lab to see if they were identical.

  This last strategy proved very helpful, because the DNA fingerprint of the bacteria from one of the children who was not from the Fall River area was different from the DNA fingerprint of the bacteria from the four local girls. “That limited the scope of our investigation,” said Dr. Besser.

  The epidemiologists spent hours trying to pinpoint a common food that all th
e patients had eaten so they might identify the source. The families all shopped at the same store, and of course there was the fish stick theory—which was quickly put to rest. Fish sticks from the store where they had been purchased, as well as some of the original fish sticks still in the families’ freezers, did not contain the bacteria. Hamburgers were another obvious possibility. After cataloguing everything the patients had eaten and drunk, the investigators interviewed neighbors who were not sick. Epidemiologists call this a case-control study, an attempt to see if there were any differences in the diets of those neighbors who did get sick compared with those who did not.

  From the case-control studies, recalls Dr. Lett, “there were no obvious high-risk foods—chicken or hamburger—and we had checked the fish sticks.” This was odd: most previous outbreaks had been linked to ground beef. The investigators were beginning to get discouraged, but using computer software to analyze the data, a most unusual and unexpected trend emerged. The patients were far more likely than their healthy neighbors to have drunk apple cider.

  While the epidemiologists were busy collecting their information, Emily Burdick was getting sicker; her kidneys had deteriorated to the point that she needed dialysis. After the dialysis catheter was placed, she developed a fever and had to be watched in the ICU for a day and a half. There she received her first dialysis treatment. This involves pumping the patient’s blood out through a catheter and into a dialysis machine— an artificial kidney—which removes the waste products. Then the purified blood is pumped back into the body. Even for an adult, this treatment is not very easy to tolerate.

  “Emily had to lie very still in an uncomfortable position for the dialysis,” remembers her father, “and the main thing she hated was when the bandage covering the catheter was removed. It hurt. We bought her a puppet—Clifford the red dog—and that was a help; Clifford occasionally got a laugh out of Emily.” Everything that happened to Emily also happened to Clifford: blood drawing, blood pressure monitoring, and bandage changes.

  To reduce the frequency of the ninety-minute dialysis treatments, Emily’s fluid intake was severely restricted, so that her body wouldn’t swell between sessions. “Anyone who has lived with a four year-old knows how little they like to brush their teeth, but Emily constantly asked me if she could brush her teeth, because she would suck up the water from the bristles of the toothbrush, she was so thirsty,” recalls Michele, who had improved enough to be discharged from the hospital and was spending all her time with Emily.

  She remembers going to dialysis with Emily and seeing outpatients coming in for their treatments, and wondering if Emily would need long-term dialysis. The families from Fall River shared the anxiety of having a seriously ill child in the hospital. “One of the children with HUS not from our area had neurological symptoms. He couldn’t walk, and when he was discharged, it was to a rehab facility. It was then that we realized that it could get a lot worse,” remembers Michele. “We also gave each other encouragement when any of the girls showed signs of improvement.”

  The advisories that had been sent out to local doctors led to the identification of eighteen cases of the E. coli diarrhea, and ultimately the tally rose to twenty-three. After apple cider became the prime suspect, Besser and Lett went back to the parents to find out where they had bought their cider. The majority remembered buying it locally at the Old Swanzey Orchard. Old Swanzey, whose main business is garden equipment and nursery supplies, is housed in a large red barn-like structure. Behind that is a smaller red shed that houses a traditional wood cider press. During the fall, the proprietor would make one or two batches of fresh cider every week.

  Apple cider is a raw, unprocessed product. The apples are ground into a pulp called pomace, which is then put into a mesh held in place by a wooden frame. The hydraulic press squeezes the pomace and the cider trickles out. It is neither filtered nor pasteurized; if you look at the bottom of a bottle of cider, you will see the sediment. Cider is a natural product that will ferment, or harden, even when refrigerated. It is different from apple juice, which is filtered and pasteurized to sterilize the product.

  The association of apple cider with disease was a surprise to the investigators because it’s a very acidic beverage—it is generally thought to be too acidic to support the growth of bacteria. But when Dr. Besser reviewed the medical literature, he found a precedent: a 1980 outbreak of HUS in Canada had been traced to a fresh pressed apple beverage. These were the same fourteen children from the Toronto area that had spurred Karmali’s work. The Hospital for Sick Children in Toronto normally sees five to ten patients with HUS per year, so to have fourteen in ten days was remarkable. Most of the children were from Pickering, a Toronto suburb twenty miles to the east. They had attended an outdoor fair where they drank apple juice. These children were very ill. One was comatose for three weeks; another was on a ventilator for several days. Six required dialysis and one infant died.

  So it was clear that apple cider could make people sick.

  The epidemiologists’ next step was to determine precisely how the bacteria had got into the cider. From the clustering of cases (no new ones had been found for three weeks after the four girls first got sick), they surmised that only one or two batches of the cider had been contaminated. The Old Swanzey Orchard is a small local operation. The owner was very cooperative with the investigation (and the establishment no longer makes the cider).

  Most of the apples used at Old Swanzey were drops—apples that had fallen off the trees naturally. The apples were neither brushed nor washed and were stored in open bins. The use of these unwashed dropped apples was forbidden by state regulation, although a later survey of cider producers found that both using drops for cider and not washing them first were fairly common practices. Cattle, the most common reservoir of E. coli, grazed in the same orchards where the drops were harvested. Deer, another animal that can carry the bacteria, were also common in these fields. The investigators cultured the cattle, the employees, the cider production equipment, and the water. They even visited deer tracking stations to culture freshly killed animals, all in an attempt to provide the decisive last link in the chain, but they never actually found the bacteria.

  Nonetheless, both Drs. Lett and Besser agree that the most likely cause was deer, or possibly cattle, which had shed the bacteria on the ground where the dropped fruit had fallen. “It’s well known that animals tend to shed the bacteria for short periods of time,” says Dr. Besser.

  Emily’s condition gradually improved. On December 6, she was discharged. Fortunately, she never developed any neurological complications and did not require long-term dialysis. By December 10, her creatinine had fallen to 1.2. In fact, all four girls from Fall River recovered.

  The doctors went on to prove that the E. coli could live in cider for much longer than anyone had previously thought. Further studies also showed that while one preservative (sodium benzoate) greatly shortened the time the bacteria survived in cider, another (potassium sorbate) actually prolonged it! The public health officials and the cider industry learned (or relearned) some valuable lessons about cider production, which will hopefully make cider even safer to drink than it already is. In their final report, the doctors recommended that consumers not drink cider made from apples that haven’t been brushed and washed. The only way to know this may be to ask the proprietor of small cider operations. The doctors’ survey of the cider industry found that larger producers were far more likely to wash and brush their apples.

  Since the Fall River outbreak, HUS and E. coli O157:H7 illnesses have become much more common, including some instances that affected many more victims. Other outbreaks have been small but interesting. In one, a local hospital in Rockford, Illinois, reported five cases of E. colidiarrhea. The common link was that the patients, all children, had been swimming in a lake in the area. Once investigators began looking, they found seven more cases. No adjacent cattle farms or sewage facilities were present, and the scientists considered the possibilities th
at waterfowl droppings or human excretion from a swimmer could have been the cause.

  In three more recent outbreaks, children were infected at petting zoos in 2004 –2005, in North Carolina (108 cases of E. coli, 15 with HUS), Florida (63 cases, 7 with HUS), and Arizona (2 cases). The children had been touching and feeding the animals or were exposed to animal manure from the wood shavings or sawdust in the pens.

  Much larger multi-state outbreaks of E. coli O157:H7 have occurred. Widespread infections of this kind result from standard practices in the meatpacking industry, where small numbers of infected cattle can contaminate huge amounts of beef. Meat that does become contaminated may end up being distributed all across the country. In 1997, Hudson Foods recalled 25 million pounds of ground beef. In 2002, Pilgrim’s Pride recalled more than 27 million pounds of poultry, and in 2006, the Topps Meat Company recalled over 21 million pounds of beef. This is only a partial list, and meat recalls are becoming increasingly common.

  The CDC estimates that 3 to 7 percent of people infected with this strain of E. coli will develop HUS, of whom about 5 percent will die. And E. coli O157:H7 infects about 73,000 Americans annually. This translates into 2,168 hospitalizations and 61 deaths. These figures seem to be increasing over time.

  Because cattle and other mammals often graze in fields that produce crops, such as lettuce and spinach, these foods have also been contaminated with E. coli and have resulted in large multi-state outbreaks. It is unlikely that we will ever eradicate the organism from our food supply. And that is why food hygiene is so important. Washing fruits and vegetables can prevent the disease. Cooking hamburger so that the internal temperature is at least 155 degrees will kill the bacteria.