The week of July 4, 1989, had been hot but quiet in upstate New York. The holiday had fallen on Tuesday and many people in the real world, as well as the central New York regional office of the New York State Department of Health located in Syracuse, had taken the whole week off. Marty Toly, the regional epidemiologist, had not. But by Friday, he was definitely looking forward to the weekend. He was a career health department employee and the local expert on communicable diseases for the 14 county health departments serving the region. In New York, counties have the primary responsibility for investigating infectious diseases that have a public health impact, and Marty oversaw their activities. He was a resource when they had a question or got into trouble. His role was somewhat of a regional lightning rod ... a last resort in the region before calling in the big guns at the main state health department office located in Albany. As a result, Marty fielded a steady stream of calls from county-level staff who investigated the reported infectious disease cases. The calls were usually from public health nurses, food sanitarians, and sometimes an epidemiologist. Questions ranged from those about amebiasis (diarrhea caused by a protozoan parasite infection) to zoonoses (infections passed from animals to humans) and everything in between. So he was not surprised just before lunchtime to get a call from the supervising sanitarian at the Onondaga County Health Department that served Syracuse and surrounding communities. But the topic was a surprise. A local doctor had just reported that a patient had tested positive for typhoid fever.
The information was that the patient was a 30-something male who had been the picture of health before falling ill several days before. In fact, he was a firefighter, a profession that requires good health and a high degree of physical fitness. However, he was now quite sick with a high fever, severe headache, and overwhelming malaise so debilitating that he was unable to get out of bed. He had been admitted to Community General Hospital in Syracuse on Monday, July 3, and had high, spiking fevers in the range of 105°F for several days. The doctors were concerned that he had sepsis, an infection in the blood that could quickly lead to dangerous drops in blood pressure and even death. Within a couple of days, blood samples sent to the laboratory for bacterial cultures were growing something. Another day went by and, through standard biochemical testing, the laboratory tentatively identified the organism as Salmonella typhi, the bacterium that causes typhoid fever. In accordance with state health regulations, the hospital was preparing to send a copy of the laboratory report to the Onondaga County Health Department, but the patient's doctor beat them to it by calling the health department.
This report surprised Marty. He knew that typhoid fever was rare now in the United States. In fact, according to New York State statistics, in each of the last 5 years only 7 to 16 cases had been reported annually in the state (outside of New York City, which keeps separate statistics). Typhoid was so infrequent that Marty thought that it was possible that it might have been a case of mistaken identity at the laboratory, and that another salmonella species was actually the culprit. All the patients with typhoid he had previously seen reported that they had recently returned from foreign travel, mostly to places in developing countries off the usual tourist-beaten path where exposure to contaminated food or water could occur. The Centers for Disease Control and Prevention estimates that more than 20 million cases of typhoid occur each year around the world. It wasn't known whether the patient had traveled outside the country recently, but the reporting doctor said that the patient had attended a conference a few weeks before. Then the sanitarian dropped a bombshell. They had just received a report of a second typhoid fever case from another hospital.
The rest of that Friday afternoon and evening would become a blur in the minds of the people who lived through it. The second typhoid case also turned out to be a young, otherwise healthy male, now very ill. Like the first patient, the second was a firefighter. But the 2 sick men didn't know each other, had never worked together, and neither had traveled outside the country.
Further questioning revealed that they did have something in common. Three weeks previously, and about 2 weeks before becoming ill, both men had attended a statewide firefighters' convention in the Catskill Mountains region of New York State. (Two weeks is a plausible incubation period—the time between exposure to an infection and the development of symptoms—for typhoid fever.) The event had been attended by thousands of firefighters and their families.
Learning of this link, Marty's mind raced. Thoughts of any possible laboratory error were now forgotten. The simplest explanation was that the 2 firefighters were exposed to typhoid at the firefighters' convention. How was that possible? Could it be coincidence? There hadn't been a typhoid outbreak in New York in Marty's several decades with the health department. If there was a typhoid outbreak in progress from exposure at this convention, it had the potential to be huge—both in terms of the number of cases and of the geographic spread.
He picked up the phone. Marty Toly needed to call his lifeline.
A Plague From Another Era
Typhoid fever has probably afflicted human populations throughout history, at least since people began to live together in larger communities. Unlike other types of salmonella infection, which often infect animals who in turn infect people, humans are the only known host for the S typhi bacteria. Typhoid bacteria are shed in the stool of infected persons and can infect other people through direct hand-to-mouth contact with ill persons or through food or water contaminated by persons with typhoid (the so-called “fecal-oral” route). Where typhoid is still present and sanitation is poor, waterborne outbreaks, particularly, can be large. Untreated, typhoid can cause a 3- to 4-week illness and 10% to 20% of infected people can die without treatment, often from intestinal perforation and sepsis. Some typhoid survivors become chronic carriers of the bacteria and continue to shed typhoid in their stool for many years. The most notorious typhoid carrier was a cook named Mary Mallon, known as “Typhoid Mary,” who was responsible for several outbreaks of typhoid in New York in the early 1900s.
The United States was not spared before the early 20th century when public health measures were instituted to control typhoid including mandatory reporting of typhoid cases by physicians to the health department, which continues to this day. Those confirmed with typhoid are prohibited from handling food until their recovery and documentation (with at least 3 successive negative stool cultures) that they are not carriers.
Typhoid began to decline precipitously in the United States in the early 1900s with the widespread use of sanitary sewer systems and chlorination of drinking water. Some cases and small outbreaks continued to occur through mid-century from typhoid carriers and international importation, but these became fewer over time. Still, even into the 1980s, the New York state health department kept a registry of known typhoid carriers, which it maintained on 3-by-5 cards, to check against when new cases without a source were reported. The registry was never updated to a computerized database, reflecting how infrequently the issue of typhoid carriers arises, and it was basically abandoned because no more cases were being reported and the known carriers had lived to old age and died. By the 1980s, typhoid had become an occasional problem; single isolated cases were an interesting oddity when they occurred, harkening back to the early days of public health, but nothing to get excited about.
Marshaling the Troops
Through years of experience dealing with communicable disease outbreaks, public health departments have fine-tuned the art of outbreak investigations. These can be complex, including many steps that need to be undertaken simultaneously and others that must proceed in a logical progression based on information gathered at each step. Many different skill sets are needed to cover all the bases. Thus, when Marty Toly's lifeline call arrived to the Bureau of Communicable Disease Control at the state health department in Albany on the afternoon of Friday, July 7, 1989, the gears of a well-oiled machine began to turn, under the direction of Dale Morse, the bureau director. Given the day and time, the ranks in the office were thin, but soon the troops began to assemble. Some key people who were on vacation were contacted. The late Friday afternoon crisis is a well-known phenomenon in public health, so emergency contact information is always available for key personnel.
The first step was to confirm the diagnosis of typhoid fever. The 2 hospital laboratories were contacted to get the details of their analysis to confirm the identification of S typhi. Arrangements were made to send samples of the bacteria to the state public health laboratory, the Wadsworth Center, for further confirmation. The information received that day and over the weekend confirmed the hospitals' diagnoses of typhoid. Wadsworth reconfirmed it a few days later.
At the same time, pertinent information gathered from interviewing the 2 ill firefighters was reviewed in detail by a bureau epidemiologist on the phone with Toly and the county staff: What dates had they attended the convention? Where did they stay? Where and what did they eat and drink? What else did they do? What other possible exposures did they have? Did they swim in untreated water such as a lake or stream? When did they start to become ill? Did they know of anyone else who was sick? The answers to these and many other questions started to hone in on the possible exposures.
The convention had been held on June 12 and 13 at the Concord Hotel, a large convention center in the Catskills, and the attendees stayed at the Concord or 1 of 5 other hotels in the area. The 2 ill firemen had stayed at the same hotel termed “Hotel A” in official correspondence per normal epidemiologic parlance but identified as the Pines Hotel in South Fallsburg, New York. The Pines and the Concord were part of a vibrant mid-twentieth century vacation area in the Catskills known as the “Borscht Belt.” More than 10 000 people had attended the convention, which was organized by the New York State Association of Fire Chiefs. Firefighters from other states and Canada may also have attended.
On a parallel track, bureau staff quickly reviewed recent surveillance reports from around the state for other typhoid “cases,” epidemiologic terminology for people with an infection of interest. This was to identify other cases that might have been reported to the counties but had not yet been investigated. A thorough search for such reports, which were sent through the mail at the time, revealed no other cases.
Good communication is important in any potentially large outbreak. That Friday afternoon, notification of the possible outbreak was sent from the bureau up the chain of command within the state health department to the level of the Commissioner of Health, Dr David Axelrod, as well as to other offices that would be involved, such as the press office, the legal office, and the food protection bureau. Health alerts were drafted to be sent to all county health departments, hospitals and clinical laboratories in New York, and to the health departments in neighboring states and the Centers for Disease Control and Prevention. These alerts were not only to inform the public health and medical communities of the possibility of a major outbreak but also to solicit reports of any recent or subsequently diagnosed typhoid cases that might be linked to the outbreak based on their travel history.
Finally, during that weekend the groundwork was laid for a multipronged investigation that would ensue on Monday morning. One prong would involve an on-site environmental inspection of the food preparation and water supply at any suspected locations where exposure might have occurred, starting with the Pines Hotel. This would involve state health department sanitarians in the food protection bureau. A second prong would involve telephone outreach to convention attendees, county health departments, hospitals, laboratories, and others to find additional ill persons (a process called “case finding”) and to begin the laborious process of an epidemiologic investigation to narrow the possible sources of disease to find the actual cause. The Wadsworth laboratory geared up to receive samples for testing from any additional patients found and from sampling done during the environmental investigation. By Sunday evening, the plan was ready to be executed.
Boots on the Ground
That Monday morning, July 10, the lead sanitarian in the state health department's local district office led his team into the lobby of the Pines Hotel and sought out the hotel manager. After absorbing the news of the probable outbreak, the hotel management quickly pledged full cooperation and assistance. That meant having all their employees fill out questionnaires about their health in general and any illnesses they had had in the past month. Employees were also asked about their roles in food preparation and whether they ate any of the food prepared on the premises. Anyone with food-handling responsibilities was interviewed in detail about his or her job, what foods he or she prepared, what ingredients he or she used, and so forth. The kitchen facilities were also inspected. Each employee who went anywhere near the kitchen was handed 3 screw-top plastic specimen containers and asked to collect and submit 3 stool samples over the next few days. These specimens would be cultured for S typhi at the Wadsworth laboratory. Providing the samples was mandatory if they wanted to continue to work in the kitchen. They were also asked to consent to having their blood drawn for Vi antibody testing. The specific disease-fighting molecules produced by the body's immune system in response to an infection, called a Vi antibody, had sometimes been helpful in identifying typhoid carriers.
In parallel with the interviewing of the hotel staff, another team member inspected the potable water system for the hotel. This was fairly straightforward because the Pines Hotel was on the municipal water system in South Fallsburg. The water was chlorinated to kill bacteria and, so far, no typhoid cases not associated with the hotel had been reported. Therefore, problems with the water coming into the hotel were not suspected. Still, the incoming waterline and the water and sewage systems in the hotel were checked for leaks and possible cross-contamination. Water samples for testing were taken at various locations.
The telephone case finding efforts began to bear fruit almost immediately. By midday Monday, 2 additional cases of typhoid fever had been reported, 1 each in Erie County (Buffalo) and Sullivan County, where the convention had taken place. The Buffalo case was also a firefighter who had attended the fire convention and had stayed at the Pines Hotel. Remarkably, the second case was a waiter who worked at the Pines. This set off immediate discussion about whether the waiter might have been the source of the outbreak. However, this was thought to be unlikely because waiters typically would not have the significant hand contact with food necessary to transmit infection (interviews with the waiter later confirmed this). And his illness had come on at the same time as the other cases, not before, so it was more likely that he was a victim of the outbreak like the other cases. In any event, the evidence was mounting that the Pines Hotel was the location where the infections had been transmitted. While the conference attendees may have eaten some meals at the Concord Hotel (the main convention venue), the waiter had not.
Still, it can be dangerous to jump to conclusions and narrow the scope of an investigation too early. Epidemiologists don't want to miss the source of an outbreak, particularly one that might still be ongoing, and getting the source of infection wrong can have legal and other ramifications. So despite the “smoking gun” of the 4 cases who had contact with the Pines Hotel, one of the first things the impromptu-phone-bank, staff who had been pressed into service from other programs throughout the state health department, did was call random names on the list of convention attendees and to ask 2 simple questions: Which hotel did you stay at and have you had a fever since attending the convention? Fever was used as a proxy for typhoid fever because at this point only 4 confirmed cases were known, it would take a while to confirm others, and the information needed to be collected quickly to be sure no other sources of typhoid were missed. Fever alone is not a very specific case definition for typhoid; typhoid has many other symptoms and most infectious diseases have fever as a symptom. But by using a very nonspecific definition, no cases of typhoid fever would be missed, and the hope was that nothing else would have caused fever in this group and muddied the waters. (Hotel guests answering the phone survey that did report fever and were still feeling ill were advised to go to their doctors immediately and tell them they may have been exposed to typhoid.)
By late Monday night, the results were in: of 200 convention attendees contacted, 25% of those who stayed at the Pines Hotel reported having a fever following their stay compared with only 2.5% of those staying at the other hotels. In other words, guests at the Pines were 10 times more likely to report fever. The chances of getting this result randomly, say by flipping a coin, were 2 in 100 000, according to the statistical tests that were done by the investigative team. So, there was a very high likelihood that the Pines was the location, and the only location, where the exposure to typhoid had occurred. The “smoking gun” was getting too hot to handle!
Opening the Floodgates
On Tuesday, July 11, 6 more confirmed cases of typhoid fever were reported to the state health department, and 3 additional cases were reported on the 12th, bringing the total to 13 cases. The ill patients came from 6 counties within New York State and 2 neighboring states, Connecticut and New Jersey. All of the cases had stayed at the Pines Hotel but not all were firefighters. Some were tourists visiting the Catskills during the peak summer season. All the known cases had stayed at the hotel on June 12 and 13, the same time as the firefighters' convention.
Tuesday was also the day that the state health department first released information to the press about the outbreak. Informing the press, and thereby the public, is an important step in any public health investigation for several reasons. First, getting the information in the press is an important way to let people who may have been exposed to the disease know what they should do. A fact sheet and a toll-free “state typhoid hotline” number were released informing people who had stayed or eaten at the Pines hotel about what action they should take (ie, go to the doctor if you're ill and say that you may have been exposed to typhoid; have your doctor call the toll-free number). The press can reach many more people more quickly than trying to call everyone who had been a guest or had eaten at the Pines Hotel for the past month, if contact information was even available to do that.
The second reason for informing the press is that people have a right to know when there is a health threat affecting their community. It is also a good way to educate the public about the often-unseen efforts of their public health departments and to educate them generally about disease prevention strategies such as handwashing. The health department dedicated a staff person in the press office to work full time just to talk to the press and answer all their questions, taking that burden off the investigative team.
Notifying the press also led to inevitable questions about whether it was safe now to stay at the hotel. It is within the power of the health commissioner to close a hotel or restaurant if there is felt to be an ongoing disease risk. On that Tuesday, the source of the outbreak wasn't known specifically, although the location of exposure, the Pines Hotel, was pretty certain. How could a decision whether to close the hotel be made? The natural inclination of epidemiologists in a situation like this is to say, “let's get some data” to help make the determination. That is what happened. Dale Morse, leading the investigative team, carved out some of the precious phone bank staff (who were by now engaged in trying to narrow the time frame of exposure to a particular day, particular meal, and particular food item in the hotel dining room to find the source) to make calls to hotel guests who had stayed at the hotel from June 14 through June 18, after the convention, to ask whether they had a fever after staying at the hotel. By Tuesday evening, the second phone survey showed that 25% of 120 guests who stayed at the hotel on June 12 or 13 reported having a fever, compared with only 2 of 230 guests who stayed there between June 14 and 18, a 29-fold lower risk for those visiting the hotel after the convention with a vanishingly small likelihood of having occurred by chance alone.
So the hotel was allowed to stay open. And as several people commented at the time, the Pines Hotel kitchen was probably the most pristine food service operation in the state at that time because of the continued presence of health department sanitarians.
Aiming for a Bull's-Eye
Through all this excitement, the epidemiology investigation team members kept their heads down and continued their work. The basic epidemiological method they employed was to categorize the telephone data they were gathering along 2 dimensions—first: who was ill and who was not ill (ie, well); and second: who had been exposed to (eaten) the meal or food item and who hadn't. There are only 4 possible combinations of answers to these 2 questions (ill-exposed, ill-not exposed, well-exposed, well-not exposed), and all possible answers fall into 1, and only 1 category. The numbers falling into each category are often displayed in a table with 2 rows (exposed and not exposed) and 2 columns (ill and well), a so-called “two-by-two table.” If the answers from any particular group surveyed for a particular illness and exposure fall primarily in the ill-exposed box and the well-unexposed box, it suggests that there is a true link between exposure and the illness. Epidemiologists say that any question about whether a disease is caused by a specific exposure can ultimately be reduced to a two-by-two table.
By Thursday, July 13, the team was ready to begin the final series of telephone surveys to determine the source of the typhoid bacteria causing the outbreak. They already knew that the source was at the Pines Hotel and that guests staying at the hotel on June 12 and 13 were more likely to become ill than guests who stayed after that time period. The next steps involved focusing the aim further to determine whether a specific meal was involved and, if so, what food item served at that meal. This sequential series of studies was jokingly referred to as the “William Tell method of epidemiology”: firing successive arrows at the target, hopefully moving from the outermost concentric circle inward toward the bull's-eye. Of course William Tell didn't have the luxury of refining his aim as he went along. He only had 1 shot!
The plan was to use the same tried-and-true telephone survey method but with some differences. Instead of using the nonspecific “fever” to define cases, it was time to become more specific. The task was to try to be sure that those people who were deemed typhoid cases for the purposes of the analysis were very likely to actually have typhoid, thus reducing the “noise” in the system and increasing the chances of finding the source. To accomplish this, a 2-part definition was used: cases were defined as persons who had been at the Pines Hotel during the implicated time period and had either a positive culture for S typhi and symptoms of disease (confirmed cases) or a fever of greater than 101°F for 3 or more days with no other identified cause (probable case). The second part of the definition was necessary because at this time, day 4 of the on-site investigation, stool culture confirmation was not yet available on all the cases.
Fortunately, the information for the determination of the implicated meal and some of the foods served at those meals had already been collected in the previous rounds of calls. Pines Hotel guests on June 12 and 13 were called and asked to recall which of the 8 meals they had eaten at the hotel starting with dinner on June 11 and going through breakfast on June 14. Next, they went through a list of food items for each meal, circling on copies of each day's menus the food items that guests recalled consuming. This was a very laborious and time-consuming process because it involved talking the respondents through the events over the 4 days of the previous month to jog their memories as to where and what they had eaten. When the numbers were crunched, the breakfast meal on June 13 stood out: 26% of guests who ate that meal met the definition of a confirmed or probable case, but none of the guests who didn't eat that meal did. The only other meal that was statistically implicated, though much less strongly, was breakfast on June 14. No other meals showed any pattern linking them to illness at all. The analysis of the foods eaten was a little less clear, perhaps because so many foods had been asked about and peoples' memories might have been stretched thin. Bill Levine, an Epidemic Intelligence Service officer who had been sent from Atlanta by Centers for Disease Control and Prevention to assist with the investigation, churned out hundreds of two-by-two tables, one for each food item. One of these many analyses suggested that, among guests who remembered drinking juice of any type at any meal, orange juice served at breakfast on June 13 was more commonly drunk by people who became ill than those who didn't. But that was an unexpected finding and the team didn't quite have faith in it. Nevertheless, the next concentric circle inward had been hit!
The final arrow in the quiver, the one intended to hit the bull's-eye, was a much shorter telephone survey honing in again on the specific foods that were served at breakfast on June 13. This time the questionnaire asked each person to answer “yes,” “no,” or “don't know” for each specific item listed on the menu. Calls were made to 28 people known from previous calls to have met the confirmed or probable case definition (called “cases”) and 40 calls were made to people who also ate breakfast on June 13 but were known not to have been ill (called “controls”). Still, these calls took time with the interviewers again helping the respondents to recall what they had eaten at one specific meal a month previously. As the afternoon of Friday, July 14, wore on, the interviewers continued with their calls, sending completed questionnaires down the hall where the data were entered into the computer. Toward the end of the day, exactly 1 week from the afternoon when the initial alarm bells had gone off, suddenly there was the answer. When the responses from persons who said “don't know” for individual food items were excluded from the analysis, only 1 food item stood out as more likely to have been consumed by the “cases” than the “controls”—only 1 item among the many served. The resulting two-by-two table showed an almost 6-fold increased risk of disease for those who ate that item. It was the orange juice.
With the answer staring them in the face on the computer screen, the members of the team collectively scratched their heads. Orange juice? This is not what they had expected. Usually bacteria need a protein source to grow. There is zero protein in orange juice. And bacteria usually don't like acidic environments. Orange juice has a pH of about 4, making it quite acidic. Was it plausible that orange juice could have been infected with S typhi? Could S typhi even survive in orange juice?
Despite these concerns, the data were the data, and they were quite clear. No other food item came close. In fact, at least one of the confirmed typhoid cases, which was not staying in the hotel but stopped by that morning to see friends, said that he had only had orange juice and nothing else. Several ill people reported only having orange juice, cereal, and coffee. So somehow these facts would have to be fit together. And the question remained, how did the orange juice get contaminated in the first place?
Conducting an investigation of a disease outbreak is a process of working to gather all the pertinent facts together and trying to tell the most plausible story consistent with those facts. In this process, usually no single fact is determinative and often not all the known facts can be neatly fit together. In the Pines Hotel typhoid fever outbreak in 1989, some things were clear. S typhi, a potentially deadly organism not usually found in the United States, caused an outbreak in hotel guests in a rural area of upstate New York. The epidemiologic data suggested that the outbreak was limited in time to 1 meal served at the hotel on 1 day and to 1 specific food item served at that meal. With the information known about orange juice as the likely vehicle carrying the typhoid bacteria that made people ill, it was now time to try to bring together the different pieces of the multifaceted investigation that had taken place and to try to make sense of the findings.
The epidemiology, food sanitation, and laboratory teams now sat down to compare notes. One question was how to establish the plausibility of orange juice as a vehicle for a typhoid fever outbreak. The biggest skeptic turned out to be Commissioner Axelrod, who had worked in the laboratory earlier in his career. Wadsworth scientists did an elegant study inoculating the S typhi outbreak strain into orange juice reconstituted from the Pines Hotel supply. The contaminated orange juice was then sampled periodically to see whether the bacteria survived. It was found that half of the bacteria survived for 6 hours but none for 24 hours. Theoretical calculations showed that as little as 1 mg (a fraction of an M&M candy) of typhoid bacteria mixed into 30 gallons of orange juice would still have had a high enough concentration that drinking 1 glassful could cause illness. Dale Morse even took a bag of M&M candies to a meeting with Dr Axelrod, who was finally convinced. The group concluded that orange juice was a plausible vehicle for the outbreak.
The health department sanitarians then focused their attention on the orange juice preparation. The orange juice was prepared from concentrate imported from Brazil (no other typhoid outbreaks related to Brazilian orange juice had been reported and so the concentrate was unlikely to be the source) and prepared with South Fallsburg town water (which tested negative for contamination). The orange juice had been prepared in a 55-gallon plastic garbage can, meaning a large volume of orange juice was present in one place at one time. A large wire whip had been used to stir the orange juice, the head of which kept falling off, so it was tied on with a piece of cloth. The kitchen staff recalled the head of the whip sometimes falling into the juice and having to be retrieved by hand. Plastic pitchers were dipped into the juice container to serve the juice and often kitchen workers did not wear gloves when carrying out this task. Finally, the orange juice can was located near the employee rest room that had no soap or towels on the day of the investigation. Foreign material had been seen floating in the can on more than 1 occasion. So there seemed to be ample opportunity for contamination of the juice, particularly if kitchen workers did not wash their hands after using the restroom.
The group next discussed the individual kitchen workers who had been involved with the preparation of the orange juice. The worker who was responsible for reconstituting the orange juice and therefore probably had the greatest opportunity for hand contact with the juice had left employment at the hotel at the end of June and could not be located. He was a migrant worker from Central America where typhoid and typhoid carriers are more common than in the United States. Unfortunately, he was not available to be questioned or tested for typhoid.
The Wadsworth laboratory was still in the process of testing the 3 stool samples submitted by each kitchen worker. About a week later, there was a revelation that 2 samples from 1 food worker were positive for the outbreak strain. This worker had not reported any symptoms of typhoid but on the second reinterview, he finally acknowledged that he had been an orange juice stirrer on occasion. Could he be a typhoid carrier? There was a great flurry of activity to reinterview him. Here was the possible source of the outbreak. However, this individual had no history of foreign travel, he was not on the health department's typhoid carrier list, and 6 additional stool cultures spread out over the next 4 months were all negative to S typhi. His Vi antibody test, a possible indicator of typhoid carrier status, was also negative (as were all the other Vi tests that were done). A detailed medical examination was done and he had no signs of gallbladder disease, which is often found in typhoid carriers. So the somewhat unsatisfactory conclusion about this individual was that he was either a victim of the outbreak, acquiring S typhi infection from drinking orange juice but not having any symptoms (the only known asymptomatic case in the outbreak) or he was a typhoid carrier, with an unknown source of his own infection, who spontaneously cleared his carrier status after causing the outbreak. Both scenarios seemed unlikely, but it appeared that one of them had to be true. The team favored the first, innocent victim scenario. That left the other option for the source of the outbreak to be the kitchen worker from Central America who reconstituted the orange juice but who could not be located for testing. The hotel manager had contacted the police, who went to the worker's New York City address to try to find him, but this may have backfired and pushed him further underground. In retrospect, sending a public health sexually transmitted disease investigator, who is skilled at finding people in the community who don't wish to be found and talking with them about sensitive information, may have been the better course. In any event, the team favored this latter explanation, as unsatisfactory and unprovable as it was. As the team had learned many times, you can't always tie these investigations up neatly in a bow.
The 1989 Pines Hotel outbreak was at the time the largest reported typhoid outbreak in the United States in almost a decade. In the end, there were 44 culture-confirmed cases of typhoid fever (including 1 hotel employee), 24 probable cases, and 1 culture-confirmed person with no symptoms.1 Twenty-one cases were hospitalized including 2 with bowel perforation. One of the cases was the child of a confirmed case. The child had not stayed at the hotel and must have acquired the infection from her parent. That was a reminder that typhoid can be transmitted 1 case at a time from one person to another, and that good handwashing and not preparing food for others when sick are always good strategies. Many additional cases were probably prevented by state and local health department interventions to remind cases and their caregivers of the person-to-person disease transmission risk and also the risks of preparing food. Known cases were not allowed to prepare food in New York until they had 2 negative stool cultures.
Based on this outbreak and others, where bare hand exposure to food was thought to be a factor, a few years later, the New York state health department adopted a new regulation requiring food workers to wear gloves when handling food that was to be served without being cooked. So the work that went into investigating this outbreak was put to good use to hopefully prevent such disease outbreaks in the future.
Although typhoid fever is now rare in the United States, this outbreak reminds public health workers, medical professionals, and members of the public that infectious diseases are never really gone and can come back if given the chance. It is the system of public health surveillance, conducted by state and local health department workers across the country, which seeks to detect and ultimately prevent infectious diseases of all types. Every day, public health professionals like Marty Toly, who is now enjoying retirement but still remembers many outbreaks over his long career, still are troubleshooting cases of rare diseases that could turn out to be harbingers of a disease outbreak like the 1989 Pines Hotel typhoid outbreak. This little-known part of the public health system is nevertheless responsible for preventing much suffering and death from infectious diseases.
Recalling the 1989 outbreak also brings into harsh focus the end of the Borscht Belt culture, which was already in decline. Both the Pines and the Concord closed in 1998. The Concord was demolished in 2014. The Pines Hotel buildings still exist but are only vandalized shells. They were not done in by the typhoid outbreak but by some of the broader forces that may also increase the risk of emerging and reemerging diseases. With the increase in international air travel and global business and tourism, diseases such as typhoid, and worse, once thought to be long gone are today just an airplane-ride away.
- What are the roles and the importance of public health surveillance in detecting outbreaks or changes in trends in infectious and noninfectious diseases?
- What are the steps in investigating a possible disease outbreak? When should these steps be initiated?
- How might the typhoid outbreak investigation described here be different if it had happened today?
- What are the risks today of the introduction of novel diseases into the United State and what steps are necessary to be prepared for them?