Principles of Epi

Question 1

Uses of Epi

Answer 1

quantitative and qualitative methods of inquiry to study the health of populations to inform evidence based public health planning and policy to improve health

Question 2

5 W’s of Epidemiology

Answer 2

1. What is the outcome - infection, injury, health gain, death etc - can be positive or negative
2. Who is at risk - risk factors or protection factors
3. Where does it happen - geographical location, sub-populations, description of the populations
4. When does it happen - seasonal? age? over time spane?
5. Why does it happen - reasons for association between x, y, z.

Question 3

Approaches to Epidemiology?

Answer 3

Many approaches divided into:

Observational studies - nature has assigned the exposure. e.g. behaviour, insect bites etc.

Intervention - investigator assigns the exposure - always is a comparison group here.

Question 4

Types of Epi Studies?

Answer 4

Many types:

1. Descriptive - investigates WHAT is happening - Patterns of disease: to who - pop, age etc. , distribution of outcome, when does it happen - overtime, seasonal etc.
   - Used to determine how much resources is needed, ID high risk groups, estimate caseloads, programme planning etc. or to generate hypotheses for further investigation via analytical studies. Studied at both local level and wider levels
   - EXAMPLE - Case reports, Case series, Incidence
   - useful if little is known - used death, birth regs, hosp records etc
2. Analytical - WHY something happens - cause and effect, not typically conducted at local level due to level of resources needed -
   - Assesses determinants of disease and risk factors
   - Key: Uses comparison groups NB. ‘Exposed v Unexposed
   - used for: Testing hypotheses and/or ID’ing & measuring assocations
   EX

Question 5

Examples - A vs D

Answer 5

Descriptive - SARS - first 100 patients - who, when…

Analytical - SARS - Measure risk factors for SARS

Question 6

Counterfactual (analytical design)

Answer 6

Unexposed group - what would have happened if the exposure was not there. Cannot be observed

Question 7

What are the types of causes?

Answer 7

Three types:

1. Necessary Cause:
   - Essential to the outcome happening
   EXAMPLE: bacteria, virus, fungi in respective infections
   abuse, stress, accidents, violence etc.
2. Sufficient Cause:
   - A set of factors or conditions that lead to an outcome
3. Component Cause:
   - Contribute to the outcome occuring (part of S.C. above).

Question 8

Define Inferring Causality?

Answer 8

To infer causality is to relate a causative agent to an outcome of interest.
Note: statistically significant association between an exposure and an interest is not enough alone to infer causality

Question 9

How do we infer causality?

Answer 9

Satisfy Bradford-Hills (1965) considerations for gathering evidence of a causal relationship.

Question 10

Temporality

Answer 10

The cause has to happen before the effect.

If there is an expected delay between the cause and the effect, the effect must happen after the delay.

EXAMPLE: If a person had cholera symptoms before drinking from a well, then they did not get it at that well.

Notes:

• Can be difficult to asses for slowly developing outcomes
• Difficult to establish in case-control or cross-sectional studies (where the outcome and the exposure is measured at the same time)

Question 11

What is reverse causality?

Answer 11

When the risk factor could be the consequence of the outcome

Example: Smoking - depression - smoking - depression

Question 12

Strength

Answer 12

The stronger the relationship between exposure and the outcome, the less likely the outcome is to be caused by another agent (decrease in liklihood of confounding)

Example: Smoking + Lung Cancer

Question 13

Repeatability

Answer 13

Getting similar results repeatedly increase the likelihood of causality between x and y

Question 14

Dose-Response

Answer 14

Is there a greater risk of outcome with a greater level of exposure? and vice versa.

Example: risk of lunch cancer in heavy smokers

Notes:

• Does not confirm causality alone
• No dose-response does not mean causality is not there

Question 15

Plausability

Answer 15

The existence of a reasonable biological explanation for a cause and effect lengs weight to an association

Dependent on existing knowledge

Question 16

Reversibility

Answer 16

Removing the exposure (or reducing) removes or reduces the outcome

Question 17

Coherence

Answer 17

Does it fit logically with other information?

EXample - the increase in smoking as a trend saw an increase in the incidence of lung cancer and this supported an association

Question 18

Analogy

Answer 18

Is it similar to other cause-effect relationships

Question 19

Specificity

Answer 19

Relationship must be specific to the outcome of interest.

EXAMPLE - wearing a helmet decreases the incidence of head trauma in cyclists, but not trauma to to other parts of the body. This lends weight the the protective efficiency of helmets to the skull.