Y2 Session 6 - Interpreting Epidemiological Findings

Question 1

What is the purpose Bradford-Hill criteria?

Answer 1

Criteria that allow us to infer causation (from association) using both observational and interventional method.

Question 2

What are the subsections of the Bradford Hill criteria?

Answer 2

1. Strength
2. Consistency
3. Specificity
4. Temporality
5. Biological gradient
6. Plausibility
7. Coherence
8. Experiment
9. Analogy

Question 3

What is the strength aspect of the Bradford-Hill criteria?

Answer 3

A stronger association increases the confidence that an exposure causes an outcome

Question 4

What is the consistency aspect of the Bradford-Hill criteria?

Answer 4

Consistent findings across settings tends to rule out errors or fallacies that might befall one or two studies.

Question 5

What is the specificity aspect of the Bradford-Hill criteria?

Answer 5

Specific diseases affecting specific works supports causality, but its absence doesn’t actually convey much.

Question 6

What is the temporality aspect of the Bradford-Hill criteria?

Answer 6

An exposure must precede an outcome.

Question 7

What is the biological gradient aspect of the Bradford-Hill criteria?

Answer 7

A dose-response effect is a compelling argument for causality e.g. more cigarettes smoked, the great the risk of cancer.

Question 8

What is the (biological) plausibility aspect of the Bradford-Hill criteria?

Answer 8

The relationship should be biological plausible where the science is “understood”

Question 9

What is the coherence aspect of the Bradford-Hill criteria?

Answer 9

The association ought to be consistent with existing theory and knowledge (similar to plausibility)

Question 10

What is the experimentation aspect of the Bradford-Hill criteria?

Answer 10

Evidence from experimentation should be supportive of proposed link (difficult to be 100% with ethical practice)

Question 11

What is the analogy aspect of the Bradford-Hill criteria?

Answer 11

Drawing upon analogous findings, we may make inferences on the relationship.

Question 12

What is internal validity?

Answer 12

Findings accurately describe the relationship between exposure and outcome in the context of the study e.g. only tested in one hospital so can’t be proven outside of the study

Question 13

What is external validity/generalisability?

Answer 13

Where an association can be proven to be applicable outside the context of the study

Question 14

What is bias?

Answer 14

Any trend in collection, analysis, interpretation, publication or review of data that can lead to conclusions that are systematically (key word) different from the truth. Something cannot have validity if there is bias present.

Question 15

What is the difference between systematic and random error?

Answer 15

Random error over a large sample size can be assumed to cancel one another out.
However, systematic error is consistent error engrained in the study, so the overall results are skewed or just incorrect. It can cause bias, affecting validity.

Question 16

What is selection bias?

Answer 16

Where individuals have an increased or decreased likelihood of being selected for the study because they have an association to the exposure AND the outcome

Question 17

What kind of study is more susceptible to selection bias?

Answer 17

Case control studies.

Question 18

How can we reduce selection bias?

Answer 18

1. Pick controls representative of target population
2. Minimise non-response
3. Compare respondents with non-respondents

Question 19

What is information bias?

Answer 19

Where we have a misclassification of the disease status, the exposure or both.

Question 20

What is interviewer bias?

Answer 20

A type of information bias related to misclassification of the exposure. If an interviewer knows who has the disease and who doesn’t they may be more thorough with those who have it e.g. exploring light and heavy smokers for lung cancer patients but neglecting to ask those without lung cancer.

Question 21

How can we reduce interviewer bias?

Answer 21

Do not tell the interviewers who does and who doesn’t have the disease.

Question 22

What is recall bias?

Answer 22

Information bias where when patients get sick, they try harder to remember pieces of information e.g. patients with lung cancer may recall smoking as a teenager, whereas a healthy patient might not see this as relevant.

Question 23

How can we reduce recall bias?

Answer 23

Using objective ways to measure exposures e.g. biochemical markers

Question 24

What is non-differential misclassification?

Answer 24

Where there is a misclassification of exposure e.g. none of the participants can guarantee exposure or misclassification of the disease but it is equal for all groups.

This means it is always bias towards the null.

Question 25

What is differential misclassification?

Answer 25

Where there is a misclassification of the exposure or disease that is uneven between groups, so can be bias towards or away from the null.

Question 26

What is response bias?

Answer 26

Response bias describes the tendency of participants to respond to questions in a way that is more socially acceptable.

A type of information bias.

Question 27

What is non-response bias?

Answer 27

Non-response bias is the process through which people who do not respond are systematically different to the people who do respond. This is a type of selection bias.

Question 28

What is the lake Wobegon effect?

Answer 28

The Lake Wobegon effect is also sometimes referred to as illusory superiority. If you ask a class of students how good their driving is, most of them will believe themselves to be better than average.

Question 29

What are the Hawthorne effects?

Answer 29

The consequence of participants realising they are being observed and therefore acting differently.