13 Deriving Inferences from Epidemiological Studies

Question 1

Association

Answer 1

The co-occurrence of two or more variables at a frequency that is more than expected by chance

Question 2

Spurious Association

Answer 2

When a common cause results in both the exposure and disease

Question 3

Reverse Causality

Answer 3

When the disease causes the exposure

Question 4

Confounder

Answer 4

Associated with both the exposure and the outcome

Question 5

4 types of causal relationships

Answer 5

Necessary and sufficient
Necessary, but not sufficient
Sufficient, but not necessary
Neither sufficient nor necessary

Question 6

Necessary and sufficient causal relationships

Answer 6

Without the exposure, the disease never develops
With the exposure, the disease always develops
Rarely ever occurs

Question 7

Necessary, but not sufficient causal relationships

Answer 7

Each factor is necessary but by itself is not sufficient to cause the disease
Multiple factors are required
Ex: like TB (need the bacillus but also poor sanitation, etc)

Question 8

Sufficient, but not necessary causal relationships

Answer 8

The factor alone can produce disease, but so can other factors acting along
The criterion of sufficient is rarely met by one factor alone

Question 9

Neither sufficient nor necessary causal relationships

Answer 9

A factor by itself is neither sufficient nor necessary to produce the disease
A complex model which likely best depicts the causal relationships for most chronic diseases

Question 10

Koch’s 3 postulates

Answer 10

1. The organism must be present in all who have the disease
2. The organism is not found with any other disease
3. The organism can be isolated, cultured and capable of producing the disease in others

Question 11

Sir Bradford Hill

Answer 11

Statistician on the MRC Streptomycin in Tuberculosis Trial (first RCT)
Studies on smoking and lung cancer

Question 12

9 Criteria for Causation

Answer 12

Temporality
Plausibility
Consistency
Strength
Dose-response relationship
Specificity
Reversibility
Coherence
Analogy

Question 13

Criteria for Causation: Temporality

Answer 13

The factor must have occurred before the disease developed
Measurement of the exposure is not required to precede measurement of the outcome though
The only criteria that is absolutely essential
Ex: smoking, then lung cancer

Question 14

Criteria for Causation: Plausibility

Answer 14

Has coherence with the current body of biologic knowledge
Makes sense from the perspective of known biological mechanisms
Ex: carcinogens cause tumour induction
Associations often precede knowledge of underlying mechanisms though

Question 15

Criteria for Causation: Consistency

Answer 15

Also called replication of findings
An association is more likely to be causal when it is observed in multiple epidemiological studies using a variety of locations, populations, and methods

Question 16

Criteria for Causation: Strength of Association

Answer 16

Most commonly measured by risk, rate, or odds ratio
Strong association = strong risk factor
But can’t rule out causation if a weak association exists

Question 17

Criteria for Causation: Dose-response relationship

Answer 17

The frequency or intensity of the outcome increases when the frequency or intensity of the exposure increases
Ex: more smoking = higher risk for lung cancer
Need to consider that a threshold may exist

Question 18

Criteria for Causation: Specificity of association

Answer 18

When an exposure is associated with only one disease
Weakest of all guidelines
Rarely applicable, should probably be deleted

Question 19

Criteria for Causation: Reversibility

Answer 19

Also called cessation of exposure
Risk of disease declines when the exposure is reduced or eliminated
Ex: stopping smoking reduces risk of lung cancer

Question 20

Criteria for Causation: Coherence

Answer 20

Also called consistency with other knowledge
Consistent with other data
Should not conflict with what is known about natural history, distribution in time or place

Question 21

Criteria for Causation: Analogy

Answer 21

Also called consideration of alternative explanations
When one causal agent is known, the standards of evidence are lowered for a second causal agent that is similar in some way

Question 22

2 steps in the basic process to move to guidelines for causal inferences in use today

Answer 22

1. Categorize existing evidence by the quality of its sources
2. Evaluate the evidence of a causal relationship using standardized guidelines