Introduction to Epidemiology and Surveillance

An Introduction to Epidemiology pt 01

Dr Alex Keenan, Epidemiology and Surveillance Analyst, Cheshire & Merseyside HPU

27th April 2010

Learning Outcomes

• Basic Understanding of Epidemiology including analysis during outbreak situations

• Exercise 3

• Importance of Surveillance How? Why? What?

• Exercise 1

• Importance of Data including integrity, consistency, accuracy and limitations of data sources

• Exercise 2

Aims of Session 1

1. Understanding of Epidemiology

2. Importance of Surveillance

3. Surveillance Systems

4. Importance of Data Quality

5. Interpretation of Data

6. Data Sources

Definitions of Epidemiology

•Epidemiology is the study of the distribution and determinants of health-related states or events in specified populations, and the application of this study to control of health problems. A Dictionary of Epidemiology, Last J. (Ed.)

•Epidemiology is the study of factors affecting the health and illness of populations, and serves as the foundation of logic of interventions made in the interest of public health and preventive medicine. Wikipedia

Surveillance

Why Bother?

WHO recommended surveillance standards, Second editionhttp://www.who.int/csr/resources/publications/surveillance/WHO_CDS_CSR_ISR_99_2_EN/en/

The core functions in surveillance of any health event are:

Case Detection

Reporting

Investigation & Confirmation

Analysis & Interpretation

Action•Control / Response•Policy•Feedback

Surveillance – functions

Case definition

Agreed system/process

A means to follow up cases

Effective (secure) storage and skills to analyse

Political will to act – perceived importance

Requires…

Surveillance Systems

Database

Laboratory /clinic

Data AnalysisData Analysis

Policy makers

PCTs / LAs / SHAs

Health Practitioners

SpecialistSpecialistLaboratoryLaboratory

Dissemination

Supplementary data

Data Quality

Input

• Consistency • Data accuracy• Data reliability (integrity)

Output

• Standardised Outputs• Consistency of Outputs• Routine e.g. monthly, annual• Ad Hoc for outbreak situations e.g. swine flu

Weekly Surveillance Data

Campylobacter – Weekly Surveillance (1)

Campylobacter - Reports per Week

0

10

20

30

40

50

60

70

80

90

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51

Week Number

Nu

mb

er

of

Re

po

rts

2006 2007 2008

Campylobacter – Weekly Surveillance (2)

Campylobacter - 4 weekly rolling average of reports

0

10

20

30

40

50

60

70

80

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51

Week Number

4 w

ee

k a

ve

rag

e n

um

be

r o

f re

po

rts

2006 2007 2008

Data Interpretation

Interrogating Datasets

• Define whether increase in reports or increase in cases

Compare with Previous Years (Temporal)

• Decide if increase is higher than expected

Compare Other Geographical Areas (Spatial)

• Is increase confined to one area

Compare Age Groups

• Is increase confined to one particular age group

Campylobacter – Weekly Surveillance (3)

Campylobacter - Central & Eastern Cheshire PCT

0

5

10

15

20

25

30

35

40

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51

Week Number

Nu

mb

er

of

Re

po

rts

2006 2007 2008

Temporal Distribution

Mumps - 4 Weekly Rolling Average of ReportsCheshire & Merseyside 2005 - 2008

0

50

100

150

200

250

300

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51

Week number

Nu

mb

er

of

Re

po

rts

2005 2006 2007 2008

Measles – Outbreak (1)

Measles – Outbreak (2)

Measles Cases by Age Groups (final)

0

5

10

15

20

25

30

35

40

45

50

under 1 1 - 4 5 - 9 10 - 14 >14

Age Groups

Nu

mb

er

of

Ca

se

s

Probable Confirmed Possible Negative Equivocal

Exercise (1) - ~15 mins

Each group has a graph of data to interpret. For each one, assess: What is the data showing? What else would you like to know? What action might it inform? What are the strengths and weaknesses of the data?

Nominate someone to feed back from each group

Exercise (2) ~ 15 mins

What routine data sources do you think that you might use when studying an epidemiological issue?

• What sources are available?

• How is the data accessed?

• How easy is it to access the data?

• What is the timeliness of accessibility?

Hospital Utilisation Data

• Hospital Episode Statistics

• Presentation at Accident & Emergency

• Attendance at outpatients clinics

• Patients registered for Specialist Clinics

• Korner data

• Laboratory data

General Practice Data

General Practice Research Database

http://www.gprd.com/

• Established in 1987

• Largest computerised database of medical records in world

• Currently 450 PCT practices

• Records for 3.4 million patients (13 million total)

• 46 million patient years of validated data

• Includes data on demographics, symptoms, therapy, referrals and lifestyle factors

NHS Direct

• Organisation started with 3 pilot sites 1998

• National 2000

• Online 24 million visitors per year

• Phone service 7 million calls per year

• Interactive TV to 16 million households

• 2.5 million users per month

• Record sex, age, postcode, primary symptom, time and date of call

Other sources of Routine Data

• Cancer registries

http://www.ukacr.org.uk/

• Surveys

http://www.dh.gov.uk/PublicationsAndStatistics/PublishedSurvey/fs/en

• Mortality Figures

http://www.statistics.gov.uk/

• National Poisons Information Service

http://www.npis.org/

Routine Environmental Data

• Air Quality

http://www.airquality.co.uk/archive/index.php

• Pollution Inventory

http://www.environment-agency.gov.uk/maps/

• Radiation

http://www.hpa.org.uk/radiation/understand/radiation_topics/ultraviolet/uv_data/index.htm

• Contaminated Land

• Local Authority public registers

Congenital Abnormalities

European Surveillance of Congenital Abnormalities

• Started in 1979

• More than 1.5 million births surveyed per year in Europe ~ 29% of European Birth Population

• 43 registries in 20 European countries

• Structural defects

• Chromosomal abnormalities

• Inborn metabolism errors

• Hereditary diseases

http://www.eurocat.ulster.ac.uk/

Merseyside and Cheshire Congenital Anomaly Survey

• Started as foetal anomaly survey in 1992

• 1995 - Member of EUROCAT

• Approximately 1200 notifications of congenital anomalies per year

• Reporting voluntary

• Delivery within geographic area irrespective of place of residence

• ~ 27000 births each year

[email protected]

The key to successful analysis of routine data for epidemiological studies

•Good case definition

•Knowledge of the limitations of routine data

•Careful selection of non exposed population

•Care with use of small numbers

Session 02

Outbreaks, Clusters and some Maths

Aims of Session 2

• Outbreaks

• Clusters

• Studies

What is an outbreak?

• Observed number of cases greater than expected for a defined place and time period

• Two or more cases with a common exposure

• One case of serious/rare disease e.g. Ebola/plague/smallpox

How do clusters arise?

Human pattern recognition

Desire to explain things

Genuine clusters in time, space and person

Clusters can be…

In Timee.g. cases of legionella

In Place e.g. meningococcal cases in same school class

In Persone.g. cases of breast cancer in a family

The term cluster denotes the suspicion of an increased frequency of some event occurring, not that any increase has been demonstrated

How do outbreaks come to light?

• Acute/unusual event:

call to HPA, NHS or LA from a health professional, school, public, hotel staff, media etc.

• Routine surveillance:

data show an increase over the normal background level for the particular place and time of year

Types of outbreak

• Common source:

• Point – peak one IP after exposure• Intermittent – irregular pattern• Continuous – irregular pattern

• Propagated (person to person):

• Successive series of increasing peaks about one IP apart

• Mixture of the above

Figure. Measles cases by date of onset of rash. Region of Madrid, March 2006.(Cases reported until 16th March, 2006)

Example of propagated outbreak(see www.eurosurveillance.org/ew/2006/060330.asp)

Example of Point Source Outbreak

Epidemiological Curve Outbreak July 2009Point Source Outbreak

0

20

40

60

80

100

120

18/07/2009 19/07/2009 20/07/2009 21/07/2009 22/07/2009 23/07/2009 24/07/2009 25/07/2009 26/07/2009 27/07/2009

Reported date of onset

nu

mb

er

of

ca

se

s

Why investigate outbreaks?

• Control of disease

• Get new evidence about:

• optimal outbreak management• prevention of outbreaks• behaviour of novel organisms

• Political, legal or public concerns

How to use epidemiology in outbreaks

• Descriptive• Date of Onset – Epidemiological Curve• Age Groups• Sex

• Analytical• Single Variable Analysis e.g. Odds Ratios• Multi Variable Analysis

• Awareness of Bias e.g. Recall – collect evidence ASAP

• Agree Questionnaire before asking questions to avoid multiple calls

Examples of Types of Outbreak Investigation

• Case Control

• Matched• Unmatched

• Cohort

• All attendees

Risk Factors – some examples

• Age

• Sex

• All foods consumed

• Toilet Visited

• Foreign Travel

• Other restaurants / takeaways / parties attended

• Swimming Pools

• Farms Visited

Odds Ratios (I)

The odds ratio is a calculation that is used to measure the strength of a relationship between 2 variables.

Odds Ratio (Cross Product) =

Ate / Exposed Didn’t Eat / Not Exposed

Ill / Disease a b

Not ill / No Disease

c d

bc

ad

Odds Ratios Example

During a Wedding people became ill and we tried to ascertain if there was a link to any particular food.

Odds Ratio (Egg Salad) = = = 12

Egg Salad Ate Didn’t Eat

Ill 8 1

Not ill 1 4

bc

ad11

48

Exercise (3) ~ 30 mins

Each group has a table of data to interpret. For each one, assess: What is the data showing? What action might it inform?

Then analyse the data as you see fit

Nominate someone to feed back from each group

Outbreak at an event in Liverpool July 2009

• 1000 - 1200 Attendees at event

• Several Reports of illness associated with people who attended event and ate food from event

• Questionnaires returned and completed by 200 people

• 148 of those were ill

• Unmatched Case Control Study

Investigation

• Environmental

• Environmental Health Officers visited premises where food was prepared to investigate possible issues

• Samples taken from premises

• Microbiological

• Samples taken from premises analysed• Stool samples analysed from those who were ill

• Epidemiological

• Epidemiological Curve• Identify Risk Factors• Perform Odds Analysis• Further more detailed analysis

Epidemiological Curve

Epidemiological Curve Outbreak July 2009

0

20

40

60

80

100

120

18/07/2009 19/07/2009 20/07/2009 21/07/2009 22/07/2009 23/07/2009 24/07/2009 25/07/2009 26/07/2009 27/07/2009

Reported date of onset

nu

mb

er

of

ca

se

s

Risk Factors Identified

• Age

• Sex

• Spring Rolls

• Chicken Lollipop

• Beef Cutlet

• Vegetable Samosa

• Chicken Curry

• Rice

Single Variable Analysis

Odds Ratio Confidence Interval

p – value

Age 1.02 0.99 – 1.04 0.056

Sex 1.30 0.71 – 2.67 0.34

Spring Roll 4.87 2.60 – 9.77 <0.01

Chicken Lollipop

6.05 2.80 – 10.52 <0.01

Beef Cutlet 6.21 3.12 – 12.56 <0.01

Vegetable Samosa

3.61 1.99 – 7.22 <0.01

Chicken Curry 15.40 6.99 – 31.84 <0.01

Rice 15.70 6.45 – 32.47 <0.01

Further Analysis

• More detailed analysis involving (stepwise) multivariate logistic regression

• People Ate more than 1 food so can take into account several foods eaten

• Bias – people from same household likely to all reply or not reply

Multivariate Analysis

Odds Ratio Confidence Interval

p- value

Chicken Curry 3.63 1.54 – 8.51 0.003

Rice 12.63 4.34 – 36.88 <0.0001

Summary of Epidemiological Findings

• Unmatched case control analysis performed

• Point source outbreak

• Age not a risk factor

• Sex not a risk factor

• All foods identified as risk factors during single variable analysis

• Chicken Curry and Rice identified as most likely risk factors to be associated with illness

A Selection of Useful Reference Books

• A Dictionary of Epidemiology, Last J. (Ed.)

• Research Methods in Health, Bowling A.

• A – Z of Medical Statistics a companion for critical appraisal, Pereira-Maxwell F.

• Essential Public Health, Donaldson & Donaldson

• Oxford Handbook of Public Health Practice, Pencheon D. et. al. (Eds.)

[email protected]

[email protected]