SEARO – CSR Training on Outbreak Investigation

Outbreak Investigation

Best Practice/MethodsPractical Reference Points

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

2

Learning Objectives

• At the end of the session, the participants will be able to:– List down the objectives of outbreak – Describe steps in outbreak investigation– Explain the importance of conducting timely

outbreak investigation

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

3

Detecting an outbreak

DetectionRoutine surveillanceClinical/laboratoryRumor verificationGeneral publicMedia

Is this anOutbreak?

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

4

Is it an outbreak?

• More than expected cases– Illustrates the importance of surveillance and timely

analysis• Clustered in time, place or person

– Pattern recognition• Concern from a HCW, school or media

– Rumour verification– Encourage participation in the system

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

5

Why investigate outbreaks?

• Stop the epidemic (outbreak)• Increase our knowledge • Prevent new episodes• Evaluate the surveillance system• Establish a surveillance system• Learn field epidemiology by

doing

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

6

Specific objectives of an investigation

Identify:• Causal agent• Mode of transmission• Source• Carrier• Population at risk• Exposure causing disease (risk factors)

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

7

Real time vs. retrospective investigation

• Epidemic in existence for several days, weeks, months.

• Based on the memory of the people

• Data already collected– To be or not to be used

It is never to late, but it can also be more difficult

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

8

Preparation• Collect preliminary information

– Available data– Consult experts (microbiologist, veterinarian,

entomologist etc)– Check search engines e.g., PUBmed– Search from both formal and informal surveillance

system (event based and indicator based)– Prepare a short memo

• Inform the concerned• Get authorization, travel itinerary• Investigation committee

– Multidisciplinary– Assign person in charge– Define tasks

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

9

Example: Community epidemic due to S. Typhimurium, Jura, May-June 1997

Context• Alert: PH medical officer• 80 cases of salmonellosis in 5 weeks• Salmonella Typhimurium• Clustered in the South department of Jura• No connection (a priori) among cases• Pressure of media, of politicians• Local Department of Public Health, Veterinary

Services, Centre National Reference, National Institute of Public Health

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

10

Steps in Outbreak Investigation

• Descriptive steps1) Is it an outbreak?2) Confirm the diagnosis:

– Which diseases are we talking about?

3) Define, find, count the cases4) Analyse data:

– When? (Time)– Where? (Place) – Who? (Person)

the sequence is not important !

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

11

• Analyse5) Generate hypotheses6) Test the hypotheses7) Compare each hypothesis with facts 8) Conduct complementary studies

• Synthesis and action9) Write a report, communicate findings10) Control measure and prevention

Steps in Outbreak Investigation the sequence is not important !

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

12

1. Confirm the outbreak

• Outbreak– n° observed cases > n° expected cases

• Expected cases? – Surveillance data– Clinicians, hospital registers– Hospital investigation, lab, doctors, schools..

• Be careful of artefacts!– Seasonal variation: (diarrhoea)– Notification variation: (new surveillance system

in place)– Diagnostic variation: (new technique)– Diagnostic mistake: (“false epidemic")

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

13

Number of Legionella cases per week, France January 1996 – August 1997

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

14

2. Confirm diagnosis

• Laboratory – serology– isolation, serotype, lysotype, etc.– toxic agent

• Meet the doctors• See the patients• Visit the laboratories

It is not necessary to confirm all the cases

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

15

Define a case

• A case definition is a standard set of criteria for deciding whether an individual should be classified as having the health condition of interest

• Includes:– clinical criteria – restrictions by time, place and person

(epidemiological link) – laboratory findings (generally)

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

16

• Suspected cases– clinical case

definition– enough for

immediate action

• Confirmed cases– stool samples– laboratory– few cases (10-

20)

Do not wait for laboratory results tostart treatment and control activities!

3.1. Define a case

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

17

Example: Case Definition Outbreak of S. Typhimurium, Jura, May-June 1997

Confirmed Probable

Clinical Diarrhoea (> 2 liquid stools/day)or Fever > 38°C ( + one day)

Diarrhoea (> 2 liquid stools/day)or Fever > 38°C ( + one day)

AND

Place, Person

resident in Jura or neighbourhood

resident in Jura or neighbourhood

Time Since 12 May 1997 Since 12 May 1997

Biological Identification of S. Typhimurium

None, but contact with confirmed case

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

18

3.2. Find and count the cases• Information sources

– All possible sources (NGOs, local leaders, etc)– Hospitals, health centres, laboratories, doctors,

nurses– schools, camps, settlements– Radio, door to door – « snow ball »– Laboratory

• How many? No strictly all • Collected information

– Demographics– clinical and biological– eventual exposure

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

19

4. Descriptive Epidemiology

• Time• Place• Persons

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

20

4. Data description: TIME epidemic curve

• Case distribution over time (according to the date - hour, week - of onset of signs)

• Onset, peak, importance, time, end of epidemic

• Abnormal cases• Allow to make hypothesis:

– incubation period, pathogen responsible

– source, mode of transmission– time of exposure

• Epidemic evolution

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

21

Example: Epidemic curve of Cases due to S.Typhimurium by week of onset of symptoms of isolated bacteria,

Jura, May- June 1997.Nombre de cas

30

1 cas

25

20

15

10

5

14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Avril Mai Juin Juillet

Semaines d’isolement ou de début des symptômes

One Case

April May June July

Number of Cases

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

22

4. Data description: PLACE

• Residence• Place of exposure

– work, food places, journeys, tour• Maps (points, attack rate)

Identify areas at risk, Identify population at risk

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

23

4. Data description: PERSON

• Distribution of cases by age, sex, profession, etc (Numerator) – ex: 50 women, 100 men

• Distribution of variables in the population from where cases are coming (Denominator)

– ex: 1500 women, 1000 men• Compute attack rate

– ex: women 50/1500 , men 100/1000

=> Identification of sub-group(s) at risk

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

24

Example: Data Description - PersonInfection by S.Typhimurium

Attack Rate by age group, Jura, May-June 1997

Age group (years)

Number of Cases

<1 2

1 – 5 36

6 – 14 22

15 – 64 29

> 65 9

Total 98

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

25

Example: Data Description - Person Infection by S.Typhimurium,

Attack Rate by age group, Jura, may-June 1997

Age Group (years)

Number of Cases

Population Attack Rate/ 100,000

<1 2 522 383

1 – 5 36 16,014 225

6 – 14 22 30,385 72

15 – 64 29 157,989 18

> 65 9 41,948 22

Total 98 246,858 40

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

26

5. Formulating hypotheses

Starting from:– Descriptive information (TPP)– Knowledge of the disease– Exploratory study on some cases

Explaining:– Causal agent– Source– Way of transmission– Carrier

DIFFICULT !!!!

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

27

Example: Formulating HypothesisInfection of S.Typhimurium

• Descriptive data:– Agent: S. Typhimurium lysotype 12 atypical – Time, epidemic curve: persistent common

source– Place: cases clustered in the south of Jura– Persons:

• Attack rate higher among children• All ages affected• Muslim among the cases

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

28

Example: Formulating HypothesisInfection S.Typhimurium

May – June 1997

Descriptive information Hypothesis

S. Typhimurium Meat (cow), salami, poultry, milk products, etc

South of Jura Regional products, local distribution

Children more affected Consumed products (also) by some children

Muslim among cases Pork less probable

Good weather condition Barbecue, poultry

Documentation Epidemic of roasted poultry described

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

29

5. Formulating HypothesisExploratory Survey

• Formulate hypothesis• Interview some cases:

– Open questionnaire and complete• Common exposure?

Example Jura:• Big questionnaire, inclusion of

regional products (cheese)• 17 cases interviewed

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

30

Example: Results of Exploratory Survey Exposure of cases to specific food

Jura, May - June 1997

Food No. of people who ate

No. of respondents

% of cases exposed

Chipolatas 6 15 40Cooked chicken 58 17 29Raw chicken 7 16 44Minced beef 7 17 41Pork 9 17 53Veal 8 17 47Cheese (comte) 13 17 77Cheese A (Fromage a)

14 16 88

Rochfort (Bleu de Gex)

6 10 60

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

31

6. Testing the hypothesis• Objectives

– Specific exposure: the carrier and the source– Factors facilitating or protective

• host, agent, environment

• Survey to identify aetiology:

– Cohort• uses attack rates• best in a small, well-defined population

– Case-control• odds ratio quantifies the relationship b/w exposure and

disease

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

32

6. Testing the hypothesisCase-control

• Compare– Proportion of exposed among cases– Proportion of exposed among the non cases (controls)

• Compute Odds Ratio (OR) and Confidence Interval (CI) at 95%

• Select controls– “Not sick”– Susceptible (e.g., not immunised)– Coming from the same population of cases– The same chances of being exposed

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

33

Exposure of Cases and Controls to specific foodJura, May - June 1997

Food No (%) of exposed cases

Odds Ratio CI 95%

(N = 42) (N = 42)

Patés 11 (26) 17 (40) 0.5 0.2 – 1.3

Sausage 24 (57) 28 (67) 0.7 0.3 – 1.6

Beef 32 (78) 33 (79) 1 0.3 – 3.5

Pork 23 (59) 29 (76) 0.5 0.2 – 1.5

Veal 22 (54) 19 (46) 1.4 0.6 – 3.4

Chicken 30 (71) 34 (81) 0.6 0.2 – 1.7

Munster (cheese)

4 (10) 1 (2) 4.0 0.5 – 35.8

Bleu de Gex (blue cheese)

12 (35) 10 (24) 3.0 0.6 – 14.9

Comté (cheese) 36 (86) 37 (88) 0.8 0.3 – 2.7

Fromage A(cheese)

33 (83) 23 (55) 6.5 1.4 – 28.8

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

34

7. Compare the hypothesis with the results

• Compare the results– clinical observation– biological examinations– epidemiological studies– statistical tests

• The hypothesis should be:– plausible – biologically acceptable– explain causal agent, source,

mode transmission, time of exposure

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

35

Example: Comparison of hypotheses with observed facts,

Jura, May-June 1997

Cheese A• Raw milk (plausible)• Consumed by children (data persons)• Regional product (data place)• Collect cheese among the cases (data

microbiological)– S. Typhimurium identified in 3 cheeses

A– Other cheeses negatives

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

36

8. Complementary studies

• At the same time, and oriented by the epidemiological survey– Environmental survey– Microbiological survey

• Plan more systematic studies– More cases, more controls– Dose-effect, facilitating factors..

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

37

Example: Complementary studiesEpidemic of S. Typhimurium

Jura, May-June, 1997

• Microbiological survey:– Food collection among cases– Sample collection among cases suppliers– Comparison of human specimens and food

products

• Survey on the distribution network of cheese A

• Survey among the producers: – Veterinary– Labour medicine– Environmental

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

38

Survey distribution network

CREMERIE

Wholesaler

Cheese dairy

Example: Complementary studiesEpidemic of S. Typhimurium

Jura, May-June, 1997

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

39

9. Writing a Report

• To be written on site• Promotes synthesis (of the objectives)• Documents the event (for evaluation/

legal purposes)• Allows communication of results• Provides recommendations• Pedagogical tool (training material)

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

40

10. Take some measure of control and prevention

• Don’t wait for the end of the investigation :– General measures at beginning– Specific measures according to the results

• Kinds of measures to control :– The source (e.g.: chlorination of water)– The transmission (e.g.: hygiene measures)– The carrier (e.g.: recall a lot of suspected cheese)– Reduce the susceptibility of host (e.g.: vaccination)

• Example Jura:– Personal Hygiene– Adequate cooking of meat– Recall of the incriminated product (Fromage A)

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

41

" The art of epidemiological reasoning is to make some reasonable conclusions

starting from imperfect data"

George W. Comstock

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

42

But better information… leads to better results

• This means having:– A good description of Time, Place, Person

(TPP) – Good data collection and preservation of

samples – A well coordinated multidisciplinary team

Immediate detection

Immediate response

Reduced morbidity and mortality

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

43

0

10

20

30

40

50

60

70

80

90

1 4 7 10 13 16 19 22 25 28 31 34 37 40

DAY

CASES

LabConfirmation

Outbreak Detection and Response

Response

Opportunity for control

Detection/Reporting

First Case

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

44

0

10

20

30

40

50

60

70

80

90

1 4 7 10 13 16 19 22 25 28 31 34 37 40

DAY

CASES

Outbreak Detection and Response

First CaseDetection/

ReportingLabConfirmation Response

Opportunity for control

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

45

Ethical Aspect

• Securing consent (participants)• Informing local authorities,

communities• Ethical treatment of animals

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

46

Best practices

• Establish clear and concise policies and procedures

• Careful documentation and proper recording of events/results

• Effective communication skills• Evaluate and review responses• Expect the unexpected

– Key people away, new, emerging pathogen, demystifying rumour, etc.

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

47

References

• De Valk, Henriette, French Institute for Public Health Surveillance (Institut de veille sanitaire, InVS)

• European Programme on Intervention Epidemiology Training

Outbreak Investigation – Best Practices/ Methods SEARO – CSR Training on Outbreak Investigation

48