Manual of Naval Preventive Medicine
Chapter 1: Food Service Sanitation
X: Foodborne Illnesses

1-73. Investigating Foodborne Illness.

Department of the Navy
Bureau of Medicine and Surgery

1. General. In the event of a suspected foodborne outbreak, prompt action must be taken to: identify cases associated with the outbreak, determine the agent causing the outbreak; identify implicated food or beverage items; and determine the factor or combination of factors which permitted the outbreak to occur. Early identification of the agent allows specific treatment of patients. Additional cases can be prevented by halting service or sale of an implicated food item. Future outbreaks can be prevented by modifying or correcting procedures for acquiring, processing and handling the implicated food. Assistance with the investigation may be obtained from the nearest NAVENPVNTMEDU or from the Occupational Health/Preventive Medicine Department at a Naval hospital or clinic by telephone or message request. Procedures to Investigate Foodborne Illness, a publication of the International Association of Milk, Food and Environmental Sanitarians, Inc., P.O. Box 702, Ames, Iowa 50010, provides excellent guidelines for conducting an investigation.

2. Investigation Procedures. An investigation is composed of several parts, which often must be accomplished promptly and simultaneously by the person or persons conducting the investigation. Ideally, procedures, materials, personnel and responsibilities for initiating and conducting an investigation should have been developed in advance.

a. Verify the diagnosis. When suspected cases of foodborne illness are reported, the first steps involve verifying the diagnosis of a foodborne disease outbreak.

(1) Complete case history questionnaires.

(a) A case history questionnaire must be completed for each ill person. Table 1-8 provides an example of this type of questionnaire.

(b) A questionnaire should also be completed for any person who has not been ill, but who may have been exposed to the suspect food item, meal, or facility. These "controls" can include family members, roommates, coworkers, shipmates, and any others at risk of illness who remained well. Comparisons of ill and well persons (e.g., food-specific attack rates) are used to analyze factors contributing to the outbreak.

(c) Case history questionnaires collect information about: the person (name, rate/rank, social security number, residential address or work/berthing as assignments, duty station, age, race, sex, and telephone number); their illness, if any (specific symptoms and specific times at which symptoms developed), and food history (when, where and what was eaten, as precisely as possible). The time at which food was eaten and symptoms started must be recorded precisely, e.g., "0100" or "1245." Responsible persons should interview and complete a questionnaire for each person.

Table 1-8. Case History Questionnaire

(2) Obtain clinical specimens. To confirm a diagnosis for most foodborne diseases, laboratory analyses must isolate or identify the etiologic agents in specimens from ill persons. Ideally, specimens should be collected while the patient is ill or when the initial interview is conducted. Specimens collected after the patient recovers may still be useful. If the patient has diarrhea, obtain a stool specimen or rectal swab. If the person is vomiting, collect vomitus. Blood specimens are used to detect antibodies, e.g., hepatitis A, or isolate pathogens, e.g., listeriosis or typhoid fever. Blood and/or urine specimens may also be useful in confirming diagnosis of chemical food poisoning. Contact the laboratory officer at the nearest medical treatment facility or NAVENPVNTMEDU for guidance on collecting, storing, and shipping samples for analysis. If the demand for laboratory analyses exceeds the capability of the MTF laboratory, contact the nearest NAVENPVNTMEDU. The units maintain a public health laboratory capability to conduct many analyses of clinical specimens from an outbreak investigation or can assist in arranging for appropriate laboratory analyses.

(3) Collect food samples and/or containers. If food items are leftover from a suspect meal, or if a commercial product is suspected, collect and preserve samples for laboratory analysis. Remaining stocks of suspect food should not be used until the investigation is complete. Use aseptic techniques and containers to collect samples; seal and label each container. Collect a sample of each item weighing 1/2 to 1 pound or measuring 1/2 to 1 pint, if less is available collect all of it. Samples of perishable foods should be chilled and held below 40 degrees F (4 degrees C), but should not be frozen. Commercial foods in containers (e.g., jars or cans) should be kept in those containers. Empty containers of suspect commercial products should also be collected and preserved. Contact the nearest NAVENPVNTMEDU for additional guidance on collecting, storing and shipping samples for analysis. NAVENPVNTMEDU laboratories can analyze food samples or can assist in arranging for appropriate laboratory analyses.

b. Develop a case definition. A case definition allows exposed persons to be classified as either cases or non cases. A case is usually defined by symptoms, e.g., a person who was at risk and developed diarrhea (3 or more loose watery stools within a 24-hour period), and a time frame. Use the data collected during the initial phase of the investigation to establish the definition. A case definition may be specific, e.g., diarrhea and fever (temperature greater than 100.5 degrees F), or more general, e.g., diarrhea, nausea or vomiting with or without fever. Cases can be categorized further as confirmed or suspected. A confirmed case meets the case definition and has laboratory evidence of infection (e.g., diarrhea and laboratory isolation of a pathogenic bacteria), while a suspected case meets the case definition but laboratory confirmation is lacking or incomplete (e.g., diarrhea only).

c. Make epidemiologic associations.

(1) Although the investigation is not complete, a preliminary assessment of available data helps to confirm that an outbreak has occurred. The investigator needs to decide if two or more persons experienced a similar illness and that the cases are associated by time (e.g., onset within a few hours or days of each other), place (e.g., eating at the same establishment or event) and/or person (e.g., eating same foods).

(2) If there is an outbreak, develop a hypothesis about the type of illness, possible vehicles of transmission and means by which the vehicle was contaminated. Hypotheses are possible explanations for the outbreak; more investigation and/or more data may be necessary to prove or disprove their role in the outbreak. Table 1-12 lists potential foodborne illnesses, common foods involved in their transmission, common means of contamination, and incubation periods. Decisions on additional investigative efforts (case and control finding, laboratory analyses, etc.) and their priority should be guided by the resulting information's value in providing or disproving the current hypotheses.

(3) If available evidence supports a hypothesis on the etiology or method of contamination, it may be appropriate to take or recommend precautionary actions at this time.

d. Provide information. Keep everyone with a "need to know" informed of the progress and findings to the investigation. Who "needs to know" varies with the outbreak but may include: appropriate line commanders; the commanding officer, preventive medicine staff and/or laboratory officer of the supporting MTF; appropriate public affairs officers (PAO); and local health department representatives. If the situation requires informing the public, work with a PAO to provide objective factual information about the outbreak and clear recommendations on actions that the public should take.

e. Expand the investigation. Often the initial investigation may have identified a pathogen. The investigator may have a plausible hypothesis for the vehicle and its method of contamination. The food service manager may have implemented the recommendations to prevent further illness. It is often tempting to conclude the investigation at this point. Such superficial investigations may underestimate the true number of cases, miss the true method of contamination, and fail to alter potentially hazardous food handling procedures. At this point it is important to find and interview additional persons (both ill and well) at risk. Complete food history questionnaires on both ill and well and obtain clinical specimens from ill persons. It may be appropriate to seek assistance, either consultative support or on-site support, from the nearest NAVENPVNTMEDU.

f. Investigate food handling procedures. The investigation must inquire into the source and method of preparation of each item of food or drink served at a suspected meal. Although a standard inspection may be conducted, an investigation focusing on high risk foods and their handling may be more productive. A flow chart documenting the individual steps from delivery, through preparation, to service of highly suspect items may be helpful. Talk with the food service officer(s), manager(s), shift supervisors and the watch captains. Collect menus, recipes, and lists of personnel with their assignments. Separately interview food service personnel involved in handling the suspect item(s). Food service personnel should have a physical examination and specimens should be collected (e.g., stool sample or rectal swab), if appropriate.

g. Analyze the data. The organization and summary of data collected from ill and well persons who ate or drank the suspect item or meal help to classify the illness, identify involved groups, and identify a possible vehicle for transmission.

(1) Plot an epidemic curve. Prepare a graph of the distribution of cases (ill persons) by the time of onset of their symptoms (figure 1-6.) The period of time covered by the outbreak determines the unit of time used on the graph. For staphylococcal food poisoning, use a scale of hours; for a possible salmonella outbreak, use 6 or 12 hour periods; and for hepatitis A, use days. A common source outbreak graph will show a sharp peak when many cases developed their symptoms followed by a gradual tapering off of cases. Figure 1-6 displays data for a common source outbreak of staphylococcal food poisoning. An outbreak with person-to-person spread (e.g., shigellosis) will show a slower rise to a less distinct peak or may have no dominate peak.

(2) Identify the common symptoms and signs. Symptoms are felt by a person, while signs are noted by an observer. Use data from ill persons to prepare a chart showing the percentage of cases with specific symptoms (e.g., nausea or headache) and signs (e.g., fever). The predominate signs and symptoms, whether enteric, neurologic or generalized, help limit the list of possible agents that caused the outbreak.

(3) Calculate incubation periods and determine a median incubation period.

(a) The interval between ingestion of the suspect food and the appearance of an initial symptom or sign of illness is the incubation period. Knowledge of the median incubation period further limits the list of possible causative agents for the outbreak. The median is used because it is not affected by exceptionally long or short incubation periods, as is the mean (average) value.

Figure 1-6. Example of an Epidemic Curve Graph of Cases by Time of Symptom Onset.

(b) Calculate the interval for each case, and determine the range of incubation periods by identifying the shortest and longest incubation period. Calculate the median )ncubation period. (Make a list of the individual incubation periods from shortest to longest. The middle value on the list, or the average of the two middle values if there is an even number of cases, is the median incubation period.)

(c) Table 1-9 displays data on symptom onset and incubation period for a common source outbreak of staphylococcal food poisoning. Table 1-10 shows the incubation periods grouped by two-hour intervals. Both the median incubation period (3.5 hours) and the large number of cases with illness onset between 2 and 4 hours after eating the suspect food are consistent with staphylococcal food poisoning.

(4) Calculate attack rates.

 

(a) Attack rates, the percentage of ill persons, may be food or meal-specific. For either type of attack rate to be meaningful, the investigator must have food and/or meal histories on both ill and well persons who were at risk of eating the suspect food or meal. Food-specific attack rates help pinpoint a suspect food within a meal, and can support observations and conclusions on food handling that contributed to the outbreak. Meal-specific attack rates are appropriate when an investigation has not pinpointed a particular meal; the results may help focus further investigative efforts.

(b) To calculate the rates, divide the number of persons who become ill after they ate a particular food or meal by the total number of persons (both cases and controls) who ate that food or meal, and multiply the results by 100. Do the same for the persons who did not eat that particular food or meal. A highly suspect food or meal will have the highest attack rate for those who ate that food or meal, and the lowest attack rate for those who did not eat that food or meal. The difference between the two rates provides an easy method of comparing different meals or different foods. The statistical significance of the difference in attack rates may be tested using tests such as Fisher's exact test or the chi squared test.

(c) When investigating a disease with a long incubation period (e.g., hepatitis A), attack rates based on food preference rather than actual consumption may be necessary. A person's food preferences may be determined by asking if, when given a choice, they always or usually eat certain foods (e.g., raw oysters), purchase particular brand items, or dine at a particular restaurant.

(d) Table 1-11 is an example of a food specific attack rate analysis. Persons who reported that they ate potato salad have a high rate of illness. The difference in attack rates is greatest for potato salad, which implicates this food item as the vehicle in the outbreak. Not all people who reported eating potato salad became ill. Some people may not accurately remember what they ate or did not eat. The inoculum of infectious agent can vary because of the size of the portion or focal areas of contamination within a food. Occasionally, one tray or container of an item may be contaminated, while the food in another tray is not contaminated. There is also individual variation in susceptibility to infection.

Table 1-9. Example of Incubation Periods, Onset and Meal Times by Patient for a Staphylococcal Food Poisoning Outbreak

Patient (number)	Ate meal (time)	Became ill (time)	Incubation period hours
8	1300	1345	0.75
20	1130	1300	1.50
2	1130	1330	2.00
12	1130	1345	2.25
21	1200	1415	2.25
13	1130	1415	2.75
9	1130	1430	3.00
10	1145	1445	3.00
7	1130	1430	3.00
4	1130	1445	3.25
5	1130	1500	3.50
14	1200	1530	3.50
16	1130	1515	3.75
22	1230	1615	3.75
23	1200	1600	4.00
3	1130	1545	4.25
11	1230	1715	4.75
15	1200	1730	5.50
18	1300	1845	5.75
1	1200	2000	8.00
6	1300	2115	8.25
17	1130	2230	11.00
19	1130	0030	13.00
Total (23 cases)			102.75

Incubation Period:
Range: 0.75 hours (shortest) to 1300 hours (longest)
Median: 3.5 hours
Mean: 4.5 hours (102.75 divided by 23)

h. Use investigative data for prevention. Preventing further illnesses is the primary purpose of a foodborne illness investigation. During or immediately after completing the investigation, recommend and/or implement measures to prevent further illness. Those measures can include removing suspect items from use or distribution and suspending or modifying hazardous food handling practices. When your conclusions and recommendations for preventive measures are based on well-documented data and findings from a thorough investigation, it is often easier to obtain support for implementing the preventive measures.

Table 1-10. Example of Incubation Periods Grouped by Two Hour Intervals for a Staphyloccal Food Poisoning Outbreak

Incubation Period	Number of Cases
First 2 hours	2
2nd-3rd hours	12
4th-5th hours	5
6th-7th hours	0
8th-9th hours	2
10th-11th hours	1
12th-13th	1

Table 1-11. Example of Food-specific Attack Rates for an Outbreak Investigation

Item of Food	Total	Persons exposed (ate food)No. ill	% ill	Total	Persons not exposed (did not eat food) No. ill	% ill	Difference in % ill
Potato Salad	246	192	78.0	58	4	6.9	71.1
Tomatoes	253	127	50.2	51	19	37.3	12.9
Ice Cream	201	98	48.8	103	48	46.6	2.2
Beans	258	129	50.0	46	17	37.0	13.0
Ham	230	108	47.0	74	38	51.4	-4.4
Crab cake	235	124	52.8	69	22	31.9	20.9

June 1991

Table 1-12. Table of Incubation Times

Name	Etiologic Agent	Foods usually involved	Incubation Time (Onset of Symptoms)
Foodborne Intoxication
Staphylococcus Food Poisoning	Several Enterotoxins of Staphylococcus aureus	Pastries, custards, salads dressings, meat, meat products	30 minutes to 7 hours, usually 2 to 4 hours
Botulism	Toxins procuded by Clostridium botulinum types A, B, and E, rarely F and G	Home canned vegetables, fruits, meats, baked potatoes, potpies, usually low acid or alkaline.	12 to 36 hours, sometimes several days
Clostridium perfringens Food Poisoning	Toxins elaborated by Type A and C strains of Clostridium perfringens	Meats, stews, meat pies, gravies made of beef, turkey, or chicken	6 to 24 hours, usually 10 to 12 hours
Vibrio parahaemolyticus Food Poisoning	Enterotoxins and hemolysis (responsible for a hemolytic reaction, the "kenagwa phenomenon") of Vibrio parahaemolyticus	Seafood, any food cross-contaminated with raw seafood, food rinsed with contaminated sea water.	4 to 96 hours, usually 12 to 24 hours
Bacillus cereus Food Poisoning	Two enterotoxins of Bacillus cereus, one heat stable causing vomiting and one heat labile causing diarrhea	Rice (such as fried rice), vegetables and meat dishes	1 to 6 hours where vomiting is the symptom, 6 to 16 hours where diarrhea is the predominant symptom
Foodborne Infections
Viral Hepatitis A (Infectious Hepatitis)	Hepatitis A virus	Contaminated water and food, including milk, sliced menats, salads, and raw or undercooked mollusks	15 to 50 days, average 28 to 30 days
Epidemic Viarl Gastroenteropathy (Norwalk type disease)	Norwalk virus, adenoviruses, astroviruses, coronavirus, and others	Clams, oysters, cockles green salads, pastry, and frostings	24 to 48 hours, range 10 to 51 hours
Salmonellosis	Numerous serotypes of Salmonella, e.g. S. Enteritidia, S. typhimurium	Raw (especially cracked) eggs, egg products, raw milk and products, meat and meat products, poultry, pet turtles and chicks	6 to 72 hours, usually 12 to 36 hours
Typhoid Fever	Salmonella typhi	Food or water contaminated by feces or urine of a patient or carrier. Shell-fish from sewage contaminated water. Flies can infect foods.	Usually 1 to 3 weeks

i. Submit a Disease Alert Report. Any foodborne disease outbreak must be reported following the guidelines of the current instruction on Disease Alert Reports. A reportable foodborne disease outbreak is any incident in which two or more persons experience a similar illness after ingestion of a common food or beverage, and epidemiologic analysis implicates the food or beverage as the source of the illness. There are two exceptions: one case of botulism or one case of chemical food poisoning constitutes an outbreak.

Approved for public release; Distribution is unlimited.

The listing of any non-Federal product in this CD is not an endorsement of the product itself, but simply an acknowledgement of the source. 

Operational Medicine 2001

Health Care in Military Settings

This web version is provided by The Brookside Associates Medical Education Division.  It contains original contents from the official US Navy NAVMED P-5139, but has been reformatted for web access and includes advertising and links that were not present in the original version. This web version has not been approved by the Department of the Navy or the Department of Defense. The presence of any advertising on these pages does not constitute an endorsement of that product or service by either the US Department of Defense or the Brookside Associates. The Brookside Associates is a private organization, not affiliated with the United States Department of Defense.

Contact Us  ·  ·  Other Brookside Products