Causal Inference Flashcards

(33 cards)

1
Q

What is causal inference?

A

Causal inference evaluates whether the exposure-outcome relationship found is truly that of cause and effect.

Causal inference is crucial in epidemiology and research to determine the validity of observed associations.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Define ‘Cause’ according to Rothman & Greenland.

A

An event, condition, or characteristic that preceded the disease event and without which the disease event either would not have occurred at all or would not have occurred until some later time.

This definition emphasizes the temporal relationship and necessity of causes in disease occurrence.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

List the four approaches to causal inference.

A
  • Bradford-Hill Criteria
  • Sufficient/Component Causal Pies
  • Counterfactual Ideal
  • Causal Diagrams or Directed Acyclic Graph (DAG)

Each approach provides a distinct framework for understanding causal relationships in epidemiological studies.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What is the Bradford-Hill Criteria?

A

A set of principles that can be used to determine whether an observed association is likely to be causal.

The criteria include factors such as strength, consistency, specificity, temporality, biological gradient, plausibility, coherence, experiment, and analogy.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What are Sufficient/Component Causal Pies?

A

A model that illustrates how various factors (components) can contribute to a sufficient cause that leads to an outcome.

This approach highlights that multiple factors can interact to produce a disease.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Explain the Counterfactual Ideal.

A

A conceptual framework that considers what would have happened to the same subjects in the absence of the exposure.

This ideal is often used in randomized controlled trials to assess causal relationships.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What are Causal Diagrams or Directed Acyclic Graphs (DAG)?

A

Visual representations that map out the relationships between variables in a causal framework.

DAGs help in understanding the direction and nature of causal relationships and are useful in identifying confounding variables.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What is the first criterion of the Bradford-Hill Criteria?

A

Strength of the association: There is a strong association between the exposure and the outcome (e.g., large RR or OR).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What does the consistency criterion in the Bradford-Hill Criteria refer to?

A

Multiple studies found similar results.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

What is meant by specificity in the Bradford-Hill Criteria?

A

The exposure/agent results in a specific outcome.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What does temporality refer to in the Bradford-Hill Criteria?

A

The exposure/agent precedes the outcome.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Fill in the blank: The Biological Gradient, also called ________, indicates that an increase in the magnitude or duration of the exposure increases the risk of the outcome.

A

dose-response relationship

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What does plausibility mean in the context of the Bradford-Hill Criteria?

A

There is a possible mechanism of action or pathway (e.g., biological) between the exposure and the outcome.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What is coherence in the Bradford-Hill Criteria?

A

The exposure-outcome relationship is in line with what is already known about the exposure and the outcome.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What does the experiment criterion in the Bradford-Hill Criteria suggest?

A

Removal of the exposure reduces or eliminates the outcome.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What is the analogy criterion in the Bradford-Hill Criteria?

A

The exposure-outcome relationship is similar to other exposure-outcome relationships previously studied.

17
Q

What is the only criterion that every epidemiologist agrees must be true to make a causal claim?

A

Temporality: the exposure/agent precedes the outcome.

18
Q

What is a sufficient cause in the context of disease?

A

A minimal set of factors that unavoidably produce disease

Sufficient causes can be thought of as the complete mechanism leading to a disease.

19
Q

How can sufficient cause be conceptualized?

A

As a disease process or disease pathway

This conceptualization helps in understanding the progression of disease.

20
Q

Can there be multiple sufficient causes for one disease?

A

Yes, there may be several sufficient causes of one disease

This indicates the complexity of disease etiology.

21
Q

What is a component cause?

A

Each factor in a sufficient cause

Component causes are essential elements that contribute to the occurrence of a disease.

22
Q

What is a necessary cause?

A

A component that is present in every sufficient cause

Necessary causes are indispensable for the disease to occur.

23
Q

Fill in the blank: A sufficient cause can also be thought of as a _______.

A

[disease process]

This highlights the pathway through which a disease manifests.

24
Q

True or False: A necessary cause is a type of component cause.

A

True

Necessary causes are specific types of component causes.

25
What are the visual representations used to depict sufficient causes of diseases?
Causal Pies ## Footnote Causal pies illustrate the relationship between different causes and their contributions to disease.
26
What is the risk of disease in a given group?
The proportion of individuals for whom a sufficient cause has formed. ## Footnote This concept emphasizes that risk is linked to the formation of sufficient causes within a population.
27
What does the apparent 'strength' of an exposure in causing disease depend on?
The prevalence of the other component causes. ## Footnote This highlights the importance of considering all contributing factors when assessing disease causation.
28
How are component causes in the same sufficient cause regarded?
They are considered to interact. ## Footnote This means that the effects of these component causes are not merely additive but may influence each other.
29
How are component causes that are not part of the same sufficient cause viewed?
They are considered independent. ## Footnote This indicates that such component causes do not influence each other in the context of disease causation.
30
What is the ideal experiment in the counterfactual model?
Take a population, expose it to the risk factor of interest, and follow it for disease outcomes. Then, set the clock back, take the same population, remove the exposure of interest, and follow it over the same period of time for disease outcomes.
31
What is the key comparison made in the counterfactual model regarding lung cancer risk?
The risk of lung cancer in a group of smokers compared to the risk of lung cancer in THE SAME GROUP, had they never smoked.
32
Why is the counterfactual model considered impossible?
It is contrary to fact (counterfactual) – it is impossible to remove past exposures and observe outcomes.
33
What alternative comparison is made instead of the ideal counterfactual model?
We compare the risk of lung cancer in a group of smokers to a group of non-smokers.