1Preliminary investigation Foodborne Disease Outbreak Investigation Team Training: Module 3 – Preliminary Investigation of an Outbreak

1Preliminary investigation

Foodborne Disease Outbreak Investigation Team Training:

Module 3 – Preliminary Investigation of an

Outbreak


Module Learning Objectives

At the end of this module, you will be able to

1.Describe the initial steps of an outbreak investigation including • Verify the diagnosis• Search for additional cases• Create a case definition• Generate a hypothesis about the source

2.Develop a case definition for an outbreak.

3.Generate a hypothesis about the source of an outbreak.

4.Prioritize an outbreak for further investigation.

> Learning objectives


Step 1: Verify Diagnosis

Purpose•Make sure illness properly diagnosed•Rule out laboratory and reporting errors•Determine that all cases suffer from same illness

> Verify diagnosis


Ways to Verify Diagnosis

• Obtain clinical samples for laboratory testing• Review medical records and laboratory results• Talk with health-care providers of cases• Interview (and even examine) cases • Consult a qualified microbiologist

> Verify diagnosis


Laboratory Error – Example

• Increase in intestinal amebiasis (Entamoeba histolytica infection) in Los Angeles

• 38 cases in 4 months (usual: one per month)• Investigation showed no common exposures• Diagnostic slides reexamined

White blood cellsE. histolytica

> Verify diagnosis


Step 2: Search for Additional Cases• Identified cases often “tip of the iceberg”• Might not represent all cases associated with

outbreak• Need to actively search for additional cases to

– Determine true magnitude of outbreak– Characterize outbreak accurately – Increase ability of epidemiologic studies to

link illness with true cause of outbreak

> Search for cases


Why Search for Additional Cases?

= Female = Male

80%

All cases (N=20)

First cases detected (N=10)

50%

Female

> Search for cases


Ways to Search for Additional Cases

• Contact implicated establishment to identify other ill customers (e.g., reservations, credit card receipts)

• Contact event organizer for list of attendees• Review foodborne illness complaints and notifiable

disease reports• Examine laboratory reports and hospital medical

records• Ask local health-care providers to look for cases• Ask known cases if they know of other cases• Review death certificates• Make announcements through local media

> Search for cases


Step 3: Develop a Case Definition

• Standard set of criteria used to classify ill people as being cases associated with a particular outbreak

• Criteria include

– Clinical findings (e.g., signs, symptoms, laboratory results)

– Restrictions by time, place, and person

• Different from a clinical diagnosis, used for epidemiologic purposes only

> Develop case definition


Case Definition - Example

For outbreak of salmonellosis at child care center

Clinical criteria• ≥3 loose stools in a 24-hour period OR• stool culture that yielded Salmonella Javiana

Restrictions• Time: onset from October 24-30• Place: attended/worked at Child Care Center X• Person: excludes siblings of initial case in a

family (if onset of illness after initial case)



Case Definition Criteria

• Objective – Not open to interpretation

• Discriminating – Distinguish between individuals with illness associated with outbreak and those with illness not associated with the outbreak



Class Question

“Fever”An oral temperature of 100.4 or higherOR

Which criterion is more objective (not open to interpretation)?


≥3 loose stools in a 24-hour period

A stool culture that yielded Salmonella OR

Which criterion is more discriminating (more likely to distinguish between illness associated with an out-break and illness not associated with that outbreak)?


Case Definition

• Does not include suspected source of outbreak (i.e., hypothesis you are trying to test)

• Can reflect different levels of certainty that person has disease associated with outbreak– Typical signs and symptoms only (“probable

case”)– Inclusion of laboratory testing or subtyping

(“confirmed case”)• Can change over time as more information

about the illness or outbreak is revealed



Class Question

An outbreak of gastrointestinal illness occurred following a birthday party at a private home on June 10. At the party, a meal was served including potato salad, hamburgers, and ice cream. Ten guests reported becoming ill. Ill persons reported vomiting, diarrhea (3 or more loose stools in a 24-hour period), fever, and headache.

Develop a case definition for illness associated with this outbreak. A line list of ill guests is provided at the end of this module.



# Sex Age Onset Vomiting Diarrhea* FeverHead- ache

Salmonella Positive Stool

Culture Salad BurgerIce

Cream

1 F 14 6/12 Y Y Y N Y Y Y Y

2 F 5 6/13 Y Y N N NA Y N Y

3 M 16 6/12 Y Y Y N NA Y Y Y

4 F 18 6/10 N Y Y Y N N Y Y

5 M 16 6/14 Y Y N N NA Y Y Y

6 M 15 6/11 Y N N N NA N N Y

7 M 15 6/13 Y Y N N NA Y Y Y

8 M 17 6/12 N Y N N NA Y N Y

9 F 16 6/20 N Y N N Y N Y Y

10 F 14 6/12 N N Y Y NA Y N Y

*Three or more stools in a 24-hour periodNA=culture not performed

Class Question (cont’d)



Class Question (cont’d) Looking at the line list, what symptoms are reported among ill persons (and what is their frequency)?

Which signs and symptoms would you include in the case definition for the outbreak?

Answer: 6 of 10 (60%) Vomiting 8 of 10 (80%) Diarrhea 4 of 10 (40%) Fever2 of 10 (20%) Headache

Answer: No right answer. You will want to include signs and symptoms common among cases that are indicative of the illness. Realize that some case definitions will exclude some of the ill people.




Would you include restrictions by time, place, and person in the case definition?

Answer: Restrictions might be useful such as onset of illness between June 11-14 among persons attending the party. These restrictions will exclude Case #4 and #9.

Two ill people had stool cultures positive for Salmonella. Would you include a positive stool culture in the case definition?

Answer: Probably not in this outbreak because too few cases had stool cultures.




All of the ill people ate ice cream. Would you include ice cream in the case definition?

Answer: Definitely not!! Never include the suspect source of the outbreak in your case definition.



Step 4: Generate Hypothesis

• Using available information to make an educated guess about the cause and source of an outbreak

• Purpose

– To direct immediate control measures

– To narrow focus of subsequent studies

– To determine the need to involve others in investigation

• Undertaken by entire team

> Generate hypothesis


Hypothesis

• Includes likely causative agent, people at risk, mode of transmission, vehicle, and period of interest

• Example:

“The outbreak is due to a bacterium that was spread during the first week of November by a food commonly consumed by children.”



Key Sources of Information

• Basic information about causative agent• Information on implicated facility or food• Descriptive epidemiology (i.e., describe cases by

time, place, and person)• Case interviews (“hypothesis-generating

interviews”)



• Common reservoirs of causative agent• Vehicles in past outbreaks• Growth requirements for causative agent• Incubation period (time from exposure to onset

of illness)

Basic Information about Causative Agent



• APHA Control of Communicable Diseases Manual• IAFP Procedures to Investigate Foodborne Illness

(Table B)• AMA “Diagnosis and Management of Foodborne

Illnesses: A Primer for Physicians”• FDA Bad Bug Book (online)• CDC A-Z Index (online)

References on Causative Agents



Information on Implicated Facility• Foods produced or served; production, processing,

distribution methods; past food safety problems• Help identify high risk foods, likely causative agent,

and contributing factors• Sources of information

– Online menus– Paperwork from past inspections, Hazard Analysis

and Critical Control Point risk assessment, facility plan review

– Regulatory inspector



Descriptive Epidemiology

• Simple characterization of outbreak by– Time– Place– Person

• Can provide clues about the mode of transmission and vehicle

• Comparison group usually needed to put findings in perspective



Time – Onset of Illness

Date of Onset of Illness

1

2

3

4

5

1/11 1/13 1/15 1/17

Nu

mb

er o

f C

ases = one case6

• Time or date of onset of symptoms• Relates back to likely period of exposure• Typically presented as epidemic curve (epi

curve)



Uses of Epi Curve• Grasp magnitude of outbreak• Clarify outbreak’s time course • Identify cases that are outliers • Draw inferences about pattern of spread

– Point source outbreak– Continuous common source outbreak– Propagated outbreak



Point Source Outbreak• Exposure to same source over brief time• Cases rise rapidly to a peak and fall off gradually • Majority of cases within one incubation period


Cryptosporidiosis cases by date of onset of illness, June

one incubation period = 7 days


Infections with Salmonella Enteritidis by date of onset of illness, January and February.

January February

Continuous Common Source Outbreak• Exposure to same source over prolonged time• Epidemic curve rises gradually• May plateau

one incubation period = 2-3 days



Propagated Outbreak• Spread from person to person• Series of progressively taller peaks• Peaks one incubation period apart


one incubation period = 10 days

October November December January

Measles cases by date of onset of illness, Aberdeen, South Dakota, October 15 – January 16


Class Question

Thirty seven cases of campylobacteriosis were identified among children attending a summer camp. Based on the epi curve, what is the most likely means of spread: point source,

continuous common source, or propagated? (average incubation period 3-5 days)

Answer: Continuous common source



Place

• Residence of cases (typically)• Distribution reveals clues about source of

outbreak – Over broad area commercial product with

wide distribution– Clustering locally sold product, point

source, or person-to-person spread– Concentrated areas with outliers travel to

affected area or importation of product• Typically presented as spot map



Place – Example • Outbreak of Salmonella

Typhimurium with unique PFGE pattern in San Diego

• 50 cases among Mexicans and Mexican Americans

• Spot map of households of cases



Place – Example (cont’d)


Percent Hispanic population 3-20% 21-39% 40-61% 62-97%


Place – Example • Outbreak associated with Mexican-style soft

cheese sold by street vendors; four vendors cited


Vendor #1

Case

Vendor #1 routeVendor #1 stops


Person

• Age group, sex, and other characteristics• Influence individual’s susceptibility to illness or

opportunities for disease exposure → can provide clues to source of outbreak

• Typically presented as percentage of all cases or rate among affected population



Person – Example • E. coli O104 outbreak in Germany with many

cases of hemolytic uremic syndrome (HUS)• 88% of HUS cases ≥20 years (usually 1-10%)• 71% of HUS cases female (usually around 50%)

• Think “girl food”


• Source of outbreak sprouts from fenugreek seeds from Egypt


Descriptive Epidemiology Clues• Person-to-person transmission

– Clustering in social units – Localized to one part of community– Occurrence of cases in waves

• Transmission by public drinking water – Widespread illness – Affecting both sexes and all age groups– Distribution consistent with public water system

• Transmission by food – Increased risk among certain groups – Distribution similar to distribution of foods



Class Question

Based on the following case characteristics, what mode of transmission seems likely in each outbreak?

Cases live in one community; most <3 years of age; attend same child care center; onset of cases in waves

Majority of cases live in city limits, age range 1-75 years; 52% female

Cases live in two states; high proportion 20-29 years; 65% male

Mode of Transmission

Person-to-person

Public water supply

Food (actually marijuana)



Hypothesis-Generating Interviews• Extensive exploration of illness and exposures with cases

• Purpose– To identify a common location or activity– To shorten list of foods and other exposures for study

• Much more detailed than interviews for foodborne illness complaints or pathogen-specific surveillance with hundreds of exposures examined



Hypothesis-Generating Interviews (cont’d)

• Extensive information on food and non-food exposures within incubation period of illness– Open-ended questions on exposures (e.g.,

restaurants, stores, events, food history)– Lengthy list of specific foods – Details on foods eaten (e.g., brands, where

purchased, purchase dates)– Non-food exposures

• Use of standard questionnaire

www.cdc.gov/outbreaknet/references_resources

(example form in Appendix)



Interpretation of Commonalities

If you find commonalities among cases:•Are commonalities unique to cases or a reflection of common exposures in the community?

•Comparison group needed– “Controlled” epidemiologic study– Other cases of unrelated foodborne illness

interviewed for other reasons (case-case comparisons)

– FoodNet Atlas of Exposures



FoodNet Atlas of Exposures• Population-based survey at

FoodNet sites• Exposures among respondents in

previous 7 days• Estimates of

background rate of exposure

• Available at www.cdc.gov/foodnet/studies_pages/pop.htm



Prioritization for Further Investigation

• Ideally, all outbreaks are investigated further to– Prevent others from becoming ill from the

outbreak source– Identify problematic food preparation practices

or risky foods to prevent future outbreaks– Add to our knowledge of foodborne diseases

• Given limited resources, not all outbreaks can be investigated

> Prioritization for further investigation


High Priority Investigations

• High public health impact– Life-threatening illness– Population at high risk for complications of illness

affected– Large number of affected persons

• Exposure likely to be ongoing – Suspicion of continuing source of exposure– Cases still rising– Less than 2 incubation periods since onset of

symptoms in last case • Intentional adulteration of food

> Prioritization for further investigation


Ability of Local Team to Respond?

• When to ask for help– Scale of outbreak likely to overwhelm local

resources– Nature of outbreak or response beyond the

experience of local staff– Outbreak suspected to affect multiple

counties, states, or countries

• How to ask for help – Call State Epidemiologist– Be prepared to share outbreak information

47Environmental health investigation

Group ExerciseDivide into groups by table.

1.Read the brief description of an E. coli O157:H7 outbreak that occurred in Wisconsin.

2.Answer the questions.

3.Develop a hypothesis about the source of the outbreak.

Time: 15 minutes

Be prepared to share your answers with the class.


Group Exercise (cont’d)On September 5, the WI Division of Public Health received separate reports of clusters of laboratory-confirmed E. coli O157:H7 infections in three non-contiguous counties: Manitowoc, Ozaukee, and Dane. The Manitowoc County cluster involved five ill persons, four of whom visited an animal exhibition at a county fair.

On September 7, WI state epidemiologist was called by the director of the Blood Center of Southeastern Wisconsin regarding 5 adults who received plasma exchanges during the prior 3 days to treat illnesses consistent with hemolytic uremic syndrome. Three had a lab-confirmed E. coli O157:H7 infection.


Group Exercise (cont’d)A total of 30 E. coli O157:H7 infections were reported in one week. (On average, 15 E. coli O157:H7 infections are reported in WI each month.)

Question 1: Do these cases represent an outbreak? What explanations might explain the increase? What information might help you determine if the increase represents an outbreak?

It is a cluster. Explanations include increased culturing, new lab testing, lab error, contamination of cultures, changes in reporting, data entry errors, and an outbreak. Information that might help determine if cases represent an outbreak include ruling out other reasons for an increase, shared case characteristics, and subtyping results.


Group Exercise (cont’d)Molecular subtyping of 8 of the E. coli O157:H7 isolates showed seven had PFGE patterns that were indistinguishable. Infections with this strain had only been reported sporadically in the past. The PFGE patterns from the four Manitowoc County fairgoers did not match this pattern.

Officials from the WI Division of Public Health decided that the cases of E. coli O157:H7 infection with the indistinguishable PFGE pattern represented a possible outbreak and assembled the outbreak investigation team to consider the situation.


Group Exercise (cont’d) Question 2: What are the modes of transmission for E. coli O157:H7? What vehicles have been associated with E. coli O157:H7 in past outbreaks? What is the average incubation period of E. coli O157:H7?

Modes of transmission include food, water, contact with animals, and person-to-person spread. Foods previously implicated include beef, grocery produce, and unpasteurized milk. The average incubation period is 3-4 days, range 1-10 days.


Group Exercise (cont’d)To search for additional cases of E. coli O157:H7 infection, the WI Division of Public Health staff used the WI Health Alert Network and e-mail to notify local, regional, and tribal health departments; laboratories; infection control professionals; hospitals; emergency departments; and clinics of the suspected outbreak.

For the investigation, a case was defined as a patient who was infected with the outbreak strain of E. coli O157: H7, who resided in WI and had onset of symptoms since August 1.


Group Exercise (cont’d)A total of 49 WI residents from 10 counties had illness meeting the case definition. The most frequently reported signs and symptoms among cases included diarrhea (96%), abdominal cramps (96%), bloody diarrhea (88%), fatigue (80%), watery diarrhea (63%), and chills (57%).

Among cases, illness onset dates occurred from August 20 through September 14. (Figure 1) Case ages ranged from 1 to 84 years; 38 (78%) cases were ≥ 20 years of age. Of the 49 cases, 35 (71%) were female, 24 (49%) were hospitalized, and 9 (18%) had hemolytic uremic syndrome.


Group Exercise (cont’d)Figure 1. Onset of illness among persons infected with outbreak strain of E. coli O157:H7, Wisconsin.


Group Exercise (cont’d)Question 3: Interpret the descriptive epidemiology of the outbreak. Were symptoms among patients consistent with infection with E. coli O157:H7? Does the epi curve suggest a mode of trans-mission? Was clustering of cases by selected demographic characteristics apparent?

Symptoms among cases are consistent with E. coli O157:H7 infection. The epi curve, with a gradual rise in cases and prolonged occurrence, is suggestive of a continuous common source outbreak. The distribution of cases across 10 counties with a clustering among women suggests some unique food.


Group Exercise (cont’d)Staff at local health departments and the WI Division of Public Health conducted telephone interviews of cases using a standard comprehensive enteric disease questionnaire. Of the initial 9 cases interviewed, 7 (78%) ate potatoes, 5 (56%) ate romaine lettuce, and 8 (89%) ate fresh spinach.

Question 4: Using the excerpt from the FoodNet Atlas of Exposures, are these exposures unusual?

From the FoodNet Atlas of Exposures, 77% of respondents ate potatoes in the previous week; 47%, romaine lettuce; 24%, fresh spinach. The proportion of E. coli O157:H7 cases eating fresh spinach is high.


Group Exercise (cont’d)Question 5: State your hypothesis about the source of this outbreak including causative agent, people at risk, mode of transmission, vehicle, and period of interest.

Question 6: How would you prioritize this outbreak for investigation? Why?


Quick Quiz


Quick Quiz

1. Why is it important to identify as many cases associated with an outbreak as possible?

A. Determine true magnitude of outbreak

B. Characterize outbreak accurately

C. Increase the ability of epidemiologic studies to link illness with true cause of outbreak

D. All of the above


Quick Quiz

2. A case definition might include all of the following EXCEPT:

A. Symptoms of the illness

B. Laboratory test results

C. Food that is the suspected source of the outbreak

D. Date of illness onset


Quick Quiz

3. Development of a hypothesis early in an outbreak helps direct subsequent steps of an outbreak investigation and should involve all investigation team members.

A. True

B. False


Quick Quiz

4. A cluster of E. coli O157:H7 infections is detected through PulseNet with cases from 30 states. Among cases, 66% were under 19 years of age and 71% were female. What is the most likely mode of transmission in this outbreak?

A. Person-to-person

B. Public water supply

C. Recreational water

D. Food


Quick Quiz

5. Which of the following characteristics would make an outbreak the highest priority for further investigation?

A. It has been 3 months since onset of illness for the last case.

B. Affected persons are limited to staff who attended an office potluck.

C. Listeria is the causative agent.

D. A total of 8 cases of campylobacteriosis were identified after searching for additional cases.

Documents

1Preliminary investigation Foodborne Disease Outbreak Investigation Team Training: Module 3 – Preliminary Investigation of an Outbreak