Deck 15: More on Causal Inference: Bias, Confounding, and Interaction

Question 1

In a case-control study,a researcher found no association between exposure to A and disease B.Later she found out that there was a misclassification when measuring exposure to A,in both cases and controls.What is a potential explanation for the results (no association) of this case-control study?

A) Nondifferential misclassification
B) Differential misclassification
C) Selection bias
D) Confounding
E) Effect-modification

Answer 1

A
This is an example of nondifferential misclassification.The fact that the misclassification affected both cases and controls indicates a nondifferential misclassification,which dilutes the association between A and B.

Question 2

Investigators think that income might be a confounder in the relationship between Hispanic ethnicity and cardiovascular disease.They conduct a case-control study and they match cases and controls based on income.They find that Hispanics are more likely to have cardiovascular disease as compared with non-Hispanics.What would you conclude about income as a potential confounder in this association?

A) Income is a confounder.
B) Income is not a confounder.
C) Income is an effect modifier.
D) Income is not associated with cardiovascular disease.
E) Income is not associated with ethnicity.

Answer 2

B
In a case-control study,if,after matching for a potential confounder,we still find an association between the exposure and the outcome,we can say that the potential confounder is actually not a confounder.

Question 3

Epidemiologists tried to investigate the role of factor Z in the exposure-disease relationship.They developed three tables to show the exposure-disease relationship based on the status of factor Z.Based on the information given,what is the role of factor Z in the exposure-disease relationship?Overall
$\begin{array}{|l|l|l|} \hline& \text { Case } & \text { Control } \\\hline \text { Exposed } & 1,050 & 500 \\\hline \text { Unexposed } & 300 & 700 \\\hline\end{array}$

Factor X present
$\begin{array}{|l|l|l|}\hline & \text { Case } & \text { Control } \\\hline \text { Exposed } & 700 & 250 \\\hline \text { Unexposed } & 200 & 350 \\\hline\end{array}$

Factor X absent
$\begin{array}{|l|l|l|}\hline & \text { Case } & \text { Control } \\\hline \text { Exposed } & 350 & 250 \\\hline \text { Unexposed } & 100 & 350 \\\hline\end{array}$

A) Factor Z is a confounder.
B) Factor Z is an effect measure modifier.
C) Factor Z is a mediator.
D) Factor Z is an outcome.
E) Factor Z does not have any role in the exposure-disease relationship.

Answer 3

Factor Z is a confounder.

Question 4

Epidemiologists tried to investigate the role of factor Y in the exposure-disease relationship.They developed three tables to show the exposure-disease relationship based on the status of factor Y.In this analysis,what is the odds ratio in the overall population? $\text { Overall }$
$\begin{array}{|l|l|l|}\hline & \text { Case } & \text { Control } \\\hline \text { Exposed } & 1,000 & 500 \\\hline \text { Unexposed } & 200 & 500\\\hline\end{array}$

$\text { Factor } Y \text { present }$
$\begin{array}{|l|l|l|}\hline & \text { Case } & \text { Control } \\\hline \text { Exposed } & 800 & 250 \\\hline \text { Unexposed } & 100 & 250 \\\hline\end{array}$

$\text { Factor } Y \text { absent }$
$\begin{array}{|l|l|l|}\hline & \text { Case } & \text { Control } \\\hline \text { Exposed } & 200 & 250 \\\hline \text { Unexposed } & 100 & 250 \\\hline\end{array}$

A) 0.5
B) 1.0
C) 2.0
D) 5.0
E) 8.0

Question 5

Investigators think that income might be a confounder in the relationship between Hispanic ethnicity and cardiovascular disease.They conduct a case-control study,and they match cases and controls based on income.They find that Hispanics are more likely to have cardiovascular disease as compared with non-Hispanics.What is the measure of association used in such study?

A) Risk ratio
B) Relative risk
C) Incidence rate ratio
D) Attack rate
E) Odds ratio

Question 6

Epidemiologists tried to investigate the role of factor Y in the exposure-disease relationship.They developed three tables to show the exposure-disease relationship based on the status of factor Y.Based on the information given,what is the role of factor Y in the exposure-disease relationship? Overall
Factor Y present
Factor Y absent

A) Factor Y is a confounder.
B) Factor Y is an effect measure modifier.
C) Factor Y is a mediator.
D) Factor Y is an outcome.
E) Factor Y does not have any role in the exposure-disease relationship.

Question 7

This table shows incidence rates (per 100,000) of groups exposed to neither risk factors or to one or two risk factors for lung cancer.What is the expected value of incidence rate X on asbestos exposure group among smokers in multiplicative scale? Incidence Rates of Lung Cancer by Asbestos and Smoking Exposure Status
$$\begin{array}{r}\text { Smoking status }\quad\\\text { Asbestos exposure }\begin{array} {|l|l|l|} \hline& \text { Yes } \quad \quad& \text { No }\quad\quad \\\hline \text { Yes (regular exercise) } & 2.0 & 2.6 \\\hline \text { No (no exercise) } & 4.0 & x\\\hline\end{array}\end{array}$$

A) 4.0
B) 6.0
C) 8.0
D) 12.0
E) 24.0

Question 8

Epidemiologists tried to investigate the role of factor X in the exposure-disease relationship.They developed three tables to show the exposure-disease relationship based on the status of factor X.Based on the information given,what is the role of factor X in the exposure-disease relationship? $\text { Overall }$
$\begin{array}{|l|l|l|}\hline & \text { Case } & \text { Control } \\\hline \text { Exposed } & 1,000 & 500 \\\hline \text { Unexposed } & 200 & 500\\\hline\end{array}$

$\text { Factor } Y \text { present }$
$\begin{array}{|l|l|l|}\hline & \text { Case } & \text { Control } \\\hline \text { Exposed } & 800 & 250 \\\hline \text { Unexposed } & 100 & 250 \\\hline\end{array}$

$\text { Factor } Y \text { absent }$
$\begin{array}{|l|l|l|}\hline & \text { Case } & \text { Control } \\\hline \text { Exposed } & 200 & 250 \\\hline \text { Unexposed } & 100 & 250 \\\hline\end{array}$

A) Factor X is a confounder.
B) Factor X is an effect measure modifier.
C) Factor X is a mediator.
D) Factor X is an outcome.
E) Factor X does not have any role in the exposure-disease relationship.

Question 9

Smoking participants of a cohort study are more likely to drop out.It is known that the effect of asbestos on getting cancer is much stronger among smokers.When analyzing data for this study,researchers find no association between asbestos and cancer.A potential explanation for the results of the study is

A) confounding.
B) effect modification.
C) selection bias.
D) information bias.
E) specificity of association.

Question 10

This table shows incidence rates (per 100,000) of groups exposed to neither risk factors or to one or two risk factors for lung cancer.What is the expected value of incidence rate X on asbestos exposure group among smokers in additive scale? Incidence Rates of Lung Cancer by Asbestos and Smoking Exposure Status
$$\begin{array}{r}\text { Smoking status }\quad\\\text { Asbestos exposure }\begin{array} {|l|l|l|} \hline& \text { Yes } \quad \quad& \text { No }\quad\quad \\\hline \text { Yes (regular exercise) } & 2.0 & 2.6 \\\hline \text { No (no exercise) } & 4.0 & x\\\hline\end{array}\end{array}$$

A) 4.0
B) 6.0
C) 8.0
D) 12.0
E) 24.0