Hazard analyses during outbreak investigations must focus on identifying the source and mode of contamination of the implemented food, the situations that allowed these contaminants to survive heat processing or other potentially lethal processes and the situations that permitted or promoted the pathogen to propagate to quantities or to produce toxins sufficient to cause illnesses in the population involved. These are based on factors that have been identified as contributing to the causation of previous outbreaks and applied research data that has been obtained during investigations, hazard analyses and challenge studies. Practical observational, measurement and sampling techniques to detect these situations and activities at the places where foods were mishandled or mistreated are given. This information is used to prevent further occurrences in the place of mishandling and other places that process or prepare the same foods in a similar manner and to incorporate into surveillance data the factors that contribute to foodborne outbreaks. Such a surveillance database ought to be used to guide inspections and hazard analyses, promulgate food regulations, train public health personnel and food workers and educate the public.

Essential in risk analysis and situations that require crisis management are effective foodborne disease investigations. To be effective, these investigations need to be more than statistical analyses of interview data and results of laboratory test although these procedures are part of epidemiological investigations. Greater emphasis is needed in the on-site investigations where foods are eaten, prepared, processed, harvested and produced. These on-site investigations are essentially hazard analyses that attempt to identify the source and mode of contamination of the etiologic agent, the way it survived any processing or conditioning that was done, and the means by which it propagated or concentrated to the extent that produced the dosage sufficient to cause illness in the population exposed. The task of the investigator is to prove or refute the validity of these hypotheses. This phase of the investigation should be active rather than passive.

The following activities ought to be done:
• Anticipate from initial or past data where to focus attention

• Prove that the illness is caused by the pathogen under consideration

• Confirm the vehicle and eliminate other possible vehicles

• Identify, if possible, the mode of contamination; this includes consideration that the raw food or an ingredient was contaminated and the possibility that an ill person, carrier or colonized person contaminated the food

• Evaluate and explain time-temperature exposures during cooking or effects of other potentially lethal processes to determine the likelihood of survival or destruction of the pathogen

• Evaluate and explain time-temperature abuse that may have led to growth and multiplication of the pathogen during holding and storage

• Ascertain the initial source of the pathogen, if this is not obvious from epidemiologic and etiologic information on the agent

• Confirm findings

Anticipate the Problem Source
During the investigation at the place of presumed mishandling, focus attention on contamination, survival and/or propagation of the etiologic agents under suspicion. There had to be contamination by a microbial or a chemical agent. This agent, however, may not have contaminated the food at the site being investigated, the food or an ingredient may have been contaminated at a previous link of the food chain. If a microbial agent was involved, it had to survive the process (e.g., cooking/heat processing) or the contamination had to have occurred after the lethal process. Pathogenic bacteria, unless highly virulent or affecting highly susceptible persons, must propagate to attain populations or to produce a quantity of toxin that overwhelms the susceptibility/resistance threshold of all those who became ill. Some parasites have to develop into an infectious stage. Therefore, it may be necessary to investigate food handling or processing activities at sites of production, processing and/or preparation to determine one or more of these situations. Be guided by data on factors that have contributed to previous food- borne disease outbreaks.[1-6]

Whether the hypotheses is based on epidemiological reasoning or statistical associations or results of laboratory testing, on-site investigations are necessary to prove or refute them. With this understanding, attitude and appropriate equipment and supplies, an investigator is ready to do an on-site investigation.[7]

Prove the Illness is Caused by the Targeted Pathogen
To prove that a particular pathogen caused the illness under investigation, obtain stool, blood or other appropriate specimens from those who are ill with a characteristic syndrome and incubation period typical of the disease under consideration. These specimens must be analyzing for the etiologic agent. This is essential to confirm diagnosis; otherwise the investigation will result in a classification of “Unknown Etiology.” Further testing to determine marker strains (e.g., seravars, phage patterns) is essential to trace the source of a pathogen. Laboratory services are essential for this phase of the investigation.

Confirm the Vehicle and Eliminate Possible Vehicles
A vehicle can be associated with the illness either by comparing attack rates of those who ate specified foods with those who did not eat those foods or by comparing eating histories of cases with controls. This is a basic epidemiologic procedure. It calls for first making a case definition. It is advisable to start with a broad case definition (e.g., everyone with diarrhea which may be defined as three or more stools during a 24-hour duration) and if applicable, proceeds to a more limited case definition (e.g., persons from which Escherichia coli O157:H7 has been isolated from a stool specimen).

Compare attack rate data of each. If they indicate the same vehicle, this should provide confidence in the results. If they differ, more information about the food history will be required. Additional statistical calculations can be done to evaluate the strength of the association of disease and eating the vehicle (e.g., relative risk or odds ratio) and the probability that the differences in rates occurred by chance (e.g., chi square or Fisher’s exact probability tests). Avoid stopping the investigation based on these calculations. This is too often done. A field investigation must confirm the possibilities of contamination, survival and/or propagation at the place of preparation, processing or elsewhere before an investigation can be considered complete. To implicate a particular food as a vehicle, all other possible vehicles and reservoirs should be eliminated from consideration by statistical, laboratory and/or field observations, measurements and testing. Although possible, it is unlikely that multiple vehicles are involved. They would each have to be contaminated from the same source, the contaminants survive the different processes, and probably multiply to, and if toxigenic produce toxins, sufficient quantities to cause illness in all those who ate that food item. Therefore, there must be a history of all the foods being contaminated from a common source and most likely, time-temperature abuse for all the foods for them all to be vehicles.

Another way to implicate a vehicle is to collect and test samples of prepared foods that were served at the event or meal under investigation for the pathogens under consideration. Samples of either finished products or foods after a particular handling operation or process are selected, depending on the observed circumstances and the type of information sought. Sampling and laboratory testing is necessary to prove that the food under investigation was a vehicle, and the results may suggest the source and/or mode of contamination. Unfortunately, foods of concern may have been consumed or discarded. If any are available, however, select the foods to be sampled based upon attack rate data and other calculations, on-site observations and measurement and information gathered from staff at the facility being investigated. Also, based on information gathered at the site and upon data about factors that have contributed to previous outbreaks involving the food and etiologic agent, select the portion of the food to be sampled. Contamination may have been on specific sites rather than throughout a food item. For example:
• If the item was rolled or pushed across a contaminated surface, such as a table or cutting block, large portions of the surface but not the interior may be contaminated

• If a food was handled by a worker, only a portion of the surface may have been contaminated

• If a fly or other insect walked over a limited area of the surface, only a small random zigzag line may be contaminated

• If the food has been sliced, the first few slices would be contaminated more heavily than the next slices, and the last few slices may not be contaminated

• If a fork or thermometer was contaminated, the internal portions at the puncture site and along the puncture hole may be the only portions contaminated

• If solid-food mixtures have been stirred or otherwise mixed, contaminants are spread through the mass, but not always homogeneously

• If liquids are mixed, there is a greater likelihood that the contamination is homogeneously distributed in the solution

• Additionally, if multiple dishes or containers of a food are prepared, each may not be contaminated and if contaminated not to the same extent and their subsequent processing and storage may have differed.

These examples are not the only possibilities of specific site contamination, but they show a need to select carefully the items to be sampled and the region in which to sample, such as the surface, the interior or a particular unit.

The oxidation-reduction condition around and within the food is created by the food mass, processing and packaging or storage container. This situation dictates which groups of bacteria (aerobic, anaerobic, microaerobic or facultative) can grow or grow well. The ambient (i.e., refrigerator, room, outdoor, warmer) storage temperature (e.g., <0°C/<32°F, 0-10°C/32-50°F, 21-50°C/70°-120°F, >55°C/>l30°F) and the thickness of the mass greatly influence microbial growth. Furthermore, the depth of the food in storage containers during cooling significantly influences bacterial growth and regions where multiplication is likely to occur. As the depth increases, the likelihood of bacterial growth increases, resulting in larger populations in the center of the mass, if contaminates are present there. The manner by which the food was served (e.g., cold, room temperature, warm or hot) further influences the quantity of pathogens present. Also, the manner by which the food was stored during the interval between serving and eating influences further growth of the pathogen and competing bacteria. This can result in either increased or diminished populations of the etiologic agent than that present at the time of eating the implicated food. Pathogens can even be eliminated in leftovers due to overgrowth of competitive microbes between serving and sampling.

A lack of understanding these situations is the main reason that samples of foods give negative results or the impression that pathogenic bacteria are not present, and therefore the etiologic agent is falsely designated as a virus. The etiologic agent can be among a host of others that were not recovered because of an inappropriate sample, the choice of the wrong site of an appropriate food, the environmental conditions of the food before and after its serving, the agents tested, and/or the laboratory methods used. All these situations and conditions must be considered when selecting the item and site to sample, and these matters must be considered when interpreting the results of the analysis.[8]

Identify the Mode of Contamination
Sources of Campylobacter jejuni, Escherichia coli, Salmonella, Vibrio parahaemolyticus and Yersinia enterocolitica are usually incoming raw foods of animal origin. If these foods are eaten raw, if raw ingredients are incorporated in unheated foods, or if the foods are lightly heated for culinary purposes or insufficiently cooked, the raw food is a likely source. If the bacterium is a spore former (e.g., Bacillus cereus, Clostridium perfringens, Clostridium botulinum) then cooked foods are the likely vehicle because the spores, which are commonly on raw soil- grown foods and on animal hair and skin, survive typical cooking and pasteurizing but not retorting. Another soil-borne pathogen is Listeria monocytogenes. If it is found or data suggests that the product was contaminated before its entrance into the place under investigation, redirect the investigation to an earlier processing or production phase (e.g., traceback).

Additionally, cross-contamination can occur from contaminated raw products to any other food. This can occur from direct contact with the contaminated food or from liquids dripping, seeping or flowing from the contaminated one to the other. Cross-contamination commonly occurs when a food worker handles a contaminated raw food and then handles (without intervening effective hand washing) ready-to-eat (RTE) food or an ingredient that will not be subsequently heated. Cross-contamination also can occur when a contaminated raw product is processed or prepared on equipment surfaces (e.g., cutting board, slicer, grinder, mixer, table) or with utensils (e.g., knife) that are used subsequently for RTE food (or an ingredient that will not be subsequently heated) without intervening washing and sanitizing the piece of equipment or utensil. Furthermore, cleaning cloths, sponges and other cleaning aids can spread pathogens from surfaces soiled by raw foods to surfaces used for foods that are not subsequently heated. If these items remain damp between uses, the contaminating bacteria can multiply. These matters need to be evaluated by observation of practices, by interviewing kitchen personnel about practices, and by swabbing with cotton-tipped swabs or sponges the surfaces of equipment or other items that may have contributed to the spread of the pathogen. The swabs and sponges need to be tested for likely pathogens and the isolates further identified by definitive typing.

If it is likely that the product became contaminated after a potentially lethal process, take steps to determine the source and subsequent mode of contamination. This is done by interviewing food workers and supervisors who were involved with its preparation and processing and others who may know about these situations; by observing operations, but realizing that food workers will be on their best behavior while being watched, and managers may have cautioned them before hand to use modified practices; and by collecting samples of the epidemiologically implicated food(s) or other foods that may have been involved previously as vehicles or as a source of cross-contamination. From this information, construct flow diagrams for all foods for which there is concern about being vehicles. Insert into each rectangle, representing an operation, a comment or symbol that indicates the type of contamination confirmed or suspected.[7]

Except for persons infected with Salmonella typhi, Shigella, Staphylococcus aureus, Streptococcus pyogenes, hepatitis A and E viruses, Norwalk-like virus and enteric viruses, carriers have not played a major role as the source of contamination. Enteric parasites (e.g., Cyclospora cayetanensis and Cryptosporidium parvum) are shed in the feces of infected persons. Vibrio cholerae is also discharged in human feces of infected persons, but carriers are seldom associated as the source of contamination. Nevertheless, do not ignore this possibility. If enteric illnesses are under investigation, collect stool specimens or rectal swabs from food workers who handled the epidemiologically implemented foods and test them for the pathogens being considered as etiologic agents.

Human beings commonly carry S. aureus in their noses and on their skin. If a staphylococcal food poisoning is under suspicion, collect nasal swabs (or skin lesion specimens, if such lesions are present) from those who handled the implicated food.

Epidemiologic data show that the a common event involving contamination of foods by workers is that they handled with bare hands cooked foods or foods that are not subjected to subsequent heating.[1,2] Investigate this situation whenever human host-adapted pathogens or staphylococci are under consideration as etiologic agents.

As part of the decision making about whether to collect specimens from food workers, interview all who handled the foods under investigation to identify whether they were ill with diarrhea or had other signs and symptoms of an enteric illness or had skin lesions either before or when preparing the food or meal in question. This is done by questioning the workers and supervisors and checking absentee records. Because a carrier may remain infected with salmonellae, shigellae and other pathogenic enteric bacteria for several weeks, the history of diarrhea should extend back two to three months. Also, question these persons about contact with other persons such as family members who have recently been ill with diarrhea or pets that may have been infected. Find out whether the workers had bare-hand contact with RTE foods or ingredients that were not heated, and get an impression of their typical hand-washing and personal hygiene practices. Also, determine whether the food workers, ill or not, ate the epidemiologically implicated food or handled raw foods of animal origin. If a worker had eaten the implicated food and became ill during the same interval as the other cases, it is very likely that this person was a victim and not the source of the pathogen.

Specimens are particularly warranted when the food worker had diarrhea before onset of the outbreak (or a close contact had), when there is suspicion of poor personal hygiene practices, and when there was a likelihood of bare- hand contact of the epidemiologically implicated food, particularly if it was not subsequently cooked. Therefore, evaluate whether an infected food worker was the source of the pathogen or a victim of the outbreak from eating the implicated food.

Evaluate and Explain Time-Temperature Exposures During Heating
Whenever a cooked food is contaminated with vegetative forms of foodborne pathogens, these bacteria must have survived the time-temperature or other potentially lethal exposures or contamination occurred after cooking. Evaluate time-temperature exposures for all possible vehicles, such as those implicated by attack rates, those suggested on bases of on-site observations, and foods that are likely to be initially contaminated with pathogens (such as poultry, eggs, meat, seafoods, rice, and beans) for which there is a history of being such vehicles. Inadequate cooking is commonly a contributing factor identified in outbreaks.[2,6]

Methods to determine survival or destruction of the pathogen(s) in question are to observe potentially lethal processes and to test the products and extrinsic environmental conditions. Initially, interview cooks and others who have knowledge of the procedures used for all heating processes (i.e., cooking, reheating). Ask about equipment used, its operation, temperature settings, cooking times, and any product temperature measurements. Better yet, measure temperatures throughout the heating process and continue afterwards until the products cool to below 55°C/130°F the next time the implicated food is prepared. Measure the geometric center of poultry, large cuts of meat and other food masses. Also, measure multiple locations for ground meat and for products such as scrambled eggs that are frequently rotated or flipped during heating. Measure regions near food surfaces following microwave heating. These regions, as well, may be appropriate for other foods (e.g., roast beef) being heated by other means. If feasible, use thermocouples to measure temperatures at specific locations and record process times continuously or periodically. Graph the data and put in guidelines to predict likely destruction of vegetative cells of pathogens. For example, 55°/l30°F for a two-hour exposure or 74°C/165°F for a few-second exposure will be necessary for destruction of these cells in moist foods. Intermediate times are required for temperature intervals within this range. Higher temperatures (100°C/212°F and 121°C/250°F) guidelines are needed to evaluate effects on spores. Measure time of exposure within lethal temperature ranges for pathogens of concern to evaluate likely survival or destruction. More specific analyses can be made by evaluating the findings according to published D values for specific foods at specific temperatures, or use z values to evaluate possible destruction of the pathogens of concern at other temperatures.[9]

Measurements of pH and/or water activity of the foods under investigation may provide additional information that is useful in evaluating the lethality of time-temperature exposures. In general, destruction is accelerated progressively as the pH of a product decreases below and increases above optimal growth temperatures. If a food with a pH value below 4 is epidemiologically implicated, seek an explanation of the means of survival. For example, type acid, short exposure time to the heat, improper mixing and high-acid tolerance of a strain are possibilities. Considerable more time-temperature exposures are necessary for dry, high-sugared or otherwise low water activity products than necessary for the same or other foods that are moist. Insert comments or symbols relating to survival into the flow diagrams of suspected vehicles that was initiated when gathering information about the process.

Samples can be collected before and after processing and tested for pathogens in question or quantity of mesophilic organisms. When interpreting results be aware of the location of the site in the food that the sample was collected, the measured time-temperature exposures, the environment in which the food was located at the time of sampling (and previously if determined), the statistical limitations based on the number of samples collected and the limitations of laboratory tests.[10]

Challenge studies can provide information on survival or destruction potentials during time-temperature exposures to simulate processing situations.[11] Such studies should be done and interpreted under supervision of a competent food microbiologist.

Evaluate Time-Temperature Exposures During Holding and Storage
There must be sufficient populations of foodborne pathogens in the vehicle to cause illness in all the cases associated with the outbreak. This usually involves proliferation of pathogenic bacteria at some stage of the food chain. Concerning eggs, this may occur during room or outdoor storage, and the potential for this is increased after breaking while pooled eggs or batters are held in large containers in refrigerators or while held at room temperature waiting preparation or holding before preparation. In an ice cream mix, bacterial growth can occur during the interval between the time when the mix was prepared and when it became frozen if the duration is sufficiently long. Bacterial multiplication in or on cooked meat, poultry and other cooked foods usually occurs after cooking either when the food was held at room or warm, outdoor temperature or while the food is stored in large containers in refrigerators, particularly if covered with tight-fitting lids or when the containers are stacked. It also may occur during prolonged warm holding of cooked foods if the temperatures are below the maximum growth temperature of the pathogens of concern. In rice, beans and moist-pasta, bacterial growth occurs after cooking either during overnight refrigeration, when held at room or outdoor temperatures for several hours or when held for several hours in warming devices. These foods cool very slowly. Bacterial growth in cold- stored foods occurs during storage over long durations. Slow cooling and room- temperature holding of foods are the most frequent contributory factors in outbreaks of foodborne diseases.[1-2,4] Inadequate hot holding is also a frequent contributory factor.

Food storage and holding need to be evaluated by interviewing kitchen personnel about routine practices and any changes in these practices when the meals under investigation were prepared; by observing practices during the on-site investigation; and/or by measuring temperatures of the products of concern through the entire time of their holding and storage, during processing and between preparation until serving the next time that they are prepared. Focus attention on food temperatures, not cooling unit air temperatures.

Time-temperature measurements of foods are made during operations and under conditions that simulate likely holding of the foods when the suspected vehicle was processed or prepared. A thermocouple probe should be inserted so that the sensor is at the geometric center of the food mass and other probes located at other sites of concern that are unique for the food or process. The data are plotted on graph paper with guidelines at maximal, optimal and minimal temperatures for the pathogens of primary concern and/or temperature ranges within which pathogenic bacteria multiply very rapidly (e.g., 30-45°C/85- 115°F), rapidly (e.g., 20-50°C/70-120°F), or slowly (e.g., 0-20°C/32-70°F). Be aware that a few foodborne pathogens can multiply at commonly used or required cold-storage temperatures (e.g., 5°C/41°F) if the duration is sufficiently long. Times within these ranges are interpreted in reference to potential bacterial growth. Insert comments or symbols about the potential for bacterial growth or propagation of parasites onto the flow diagrams.

As with other phases of the investigation, other activities can be done. For example, samples can be collected from foods after a holding interval and tested for the presence and quantity of pathogens. This can provide supporting information that the food was a vehicle. Tests for certain indicator organisms and/or mesophilic populations can show or suggest possibilities of bacterial growth. Measurements of pH and water activity also can give information that is helpful in evaluating the potential for bacterial growth. For example, it is unlikely that pathogenic bacteria grew at pH values below 4 or at water activity values below 0.85; only a few of them would grow below 0.92. If such foods are considered to be vehicles, give biological plausible explanations. Challenge tests, directed and interpreted by a competent food microbiologist, can be done to demonstrated growth of pathogens in the foods under investigation during simulated holding at various time-temperature exposures.[11]

Ascertain the Initial Source of the Pathogens
Determination of the source and mode of contamination or means of survival or propagation may include a traceback of the product from a foodservice establishment to a processor and/or distributors back to the farms or ranches where animals were raised or products grown or harvested. Observations and measurements should be made at all sites where mishandling to the extent that contributed to the outbreak is suspected. At processing plants, storage facilities and transportation vehicles, the procedures are as stated above. Evaluate water sources and well or spring construction and collect water samples (if water is suspected as a source of contamination) at rural processing plants and on farms.10 Fecal droppings, litter samples, drags over litter, or environmental samples can be collected to attempt to identify the marker stain and associate it with the heard or flock. Evaluate personal hygiene practices and facilities at these locations. Success in these endeavors will involve cooperation and action (and official agreements and documentation) by either state or federal departments of agriculture or the U.S. Food and Drug Administration (FDA) and the food industry.

Confirm Findings
If on-site investigative data are inconclusive, state the outcome as just a hypothesis or formulate other hypotheses that better explain the situation and test them by further investigation. Review tables of factors that influence growth and survival9 and primary and secondary factors that have contributed to outbreaks of the disease under investigation to aid in interpreting these data.[1-7]    

During the investigation, identification of operations that have contributed to previous outbreaks do not prove that they occurred before the outbreak, but it shows conditions that could have allowed contamination, survival and/or propagation. Contrarily, the lack of their identification during the investigation does not indicate that these or other malpractices did not occur. Simulation studies can give predictive results.

The statistically and/or laboratory confirmed vehicle should be further proven by observations and measurements that demonstrated the biological plausibility of contamination, survival and/or proliferation, as applicable, of the etiologic agent while disproving the role of other suspected or possible vehicles. Often more questions are raised than are answered during investigations. If the conclusions are not plausible under prevailing scientific knowledge, research should be initiated to resolve the differences. This is a way that new knowledge is generated.

All aspects of this approach may not be possible in all outbreak investigations for a variety of reasons. Nevertheless, all available evidence ought to be collected to implicate a food and to clear others of their possible involvement and to identify contributory factors. Despite barriers to complete investigations, the cause, contributory factors and source of etiologic agent should not be Just a matter of opinion.[12]

Training
To attain improved investigations of outbreaks and sporadic cases of food- bo