Interpreting Epidemiological Findings

Question 1

What are the Bradford hill criteria?

Answer 1

• STRENGTH
• CONSISTENCY
• SPECIFICITY
• TEMPORALITY
• BIOLOGICAL GRADIENT
• PLAUSIBILITY
• COHERENCE
• EXPERIMENT
• ANALOGY

Question 2

What is the strength of a study?

Answer 2

The strength of the association and relationship

Question 3

What is consistency?

Answer 3

• Consistency (reproducibility): Consistent findings observed by different persons in different places with different samples strengthens the likelihood of an effect.

Question 4

What is specificity?

Answer 4

• Specificity: Causation is likely if there is a very specific population at a specific site and disease with no other likely explanation. The more specific an association between a factor and an effect is, the bigger the probability of a causal relationship.

Question 5

What is temporality?

Answer 5

• Temporality: The effect has to occur after the cause (and if there is an expected delay between the cause and expected effect, then the effect must occur after that delay).

Question 6

What is biological gradient?

Answer 6

• Biological gradient (dose-response relationship): Greater exposure should generally lead to greater incidence of the effect. However, in some cases, the mere presence of the factor can trigger the effect. In other cases, an inverse proportion is observed: greater exposure leads to lower incidence. (Hill used smoking as an example more packs confers greater risk)

Question 7

What is coherence?

Answer 7

• Coherence: Coherence between epidemiological and laboratory findings increases the likelihood of an effect. However, Hill noted that “… lack of such [laboratory] evidence cannot nullify the epidemiological effect on associations”.

Question 8

What is plausibility?

Answer 8

• Plausibility: A plausible mechanism between cause and effect is helpful (but Hill noted that knowledge of the mechanism is limited by current knowledge).

Question 9

What is experiment?

Answer 9

• Experiment: “Occasionally it is possible to appeal to experimental evidence”.

Question 10

What is analogy?

Answer 10

• Analogy: The use of analogies or similarities between the observed association and any other associations (rubella and thalidomide causing congenital abnormalities)

Question 11

What is internal validity?

Answer 11

Internal validity is the extent to which a study establishes a trustworthy cause-and-effect relationship between a treatment and an outcome.

Question 12

What is external validity?

Answer 12

External validity is the validity of applying the conclusions of a scientific study outside the context of that study. In other words, it is the extent to which the results of a study can be generalized to and across other situations, people, stimuli, and times

Question 13

What is bias?

Answer 13

Any trend in the collection, analysis, interpretation, publication or review of data that can lead to conclusions that are systematically different from the truth

Question 14

What is the difference between random and systematic error?

Answer 14

Any trend in the collection, analysis, interpretation, publication or review of data that can lead to conclusions that are systematically different from the truth

Question 15

What is external validity synonymous with?

Answer 15

Generalisability

Question 16

What is selection bias?

Answer 16

Selection bias is the bias introduced by the selection of individuals, groups or data for analysis in such a way that proper randomization is not achieved, thereby ensuring that the sample obtained is not representative of the population intended to be analysed; when participation of the study is associated with BOTH the exposure and outcome

Question 17

What is Berkson’s bias?

Answer 17

It can arise when the sample is taken not from the general population, but from a subpopulation i.e. a hospital based case control study with participants selected from the hospital population

Question 18

What is the health worker effect?

Answer 18

Bias in occupational epidemiology studies that is typically characterized by lower relative mortality and morbidity rates from all causes combined and from selected causes in an occupational cohort, possibly masking an increased risk of the disease under study

Question 19

What are the three ways you can decrease selection bias?

Answer 19

• Controls representative of target population
• Minimise non-response
• Compare respondents with non-respondents

Question 20

What are the three main types of bias?

Answer 20

• Selection bias
• Information bias
• Confounding

Question 21

What is a systematic error?

Answer 21

• Systematic error-we consistently over estimate, introduces bias, reducing validity

Question 22

What is information bias?

Answer 22

Misclassification-disease status/exposures or both

Question 23

Why does misclassification arise?

Answer 23

• Study variable are not properly defined

* Flaws in data collection

Question 24

Why does interview bias occur?

Answer 24

Example: Asking about an individual’s exposure status Exposure to smoking is investigated more thoroughly in participants that have lung cancer compared with controls (without lung cancer) because interviewers expect lung cancer patients are likely to be or have been smokers.

Question 25

How can interviewer bias be prevented?

Answer 25

• Can be prevented if interviewer does not know the disease status of the individual or if the collection process is standardized so that interviewers follow strictly-defined protocols.

Question 26

How is recall bias prevented?

Answer 26

Can be prevented by using objective ways to assess exposure, for example medical records or biomarkers. Refers to the differentially inaccurate recall of past exposure

Question 27

What is non-differential misclassification?

Answer 27

• When all cases and controls have difficulty remembering an exposure status and it has nothing to do with their disease. Or when exposure status is misclassified, but equally among cases and controls.
• Also applies to when there are errors in determining the outcome, but they occur equally among exposed and unexposed individuals.
• Odds ratio is then Biased toward the null

Question 28

In non-differential misclassification, the odds ratio tends to what?

Answer 28

Towards to null

Question 29

What is differential misclassification?

Answer 29

When exposure status is misclassified, but unequally among cases and controls.
Also applies if there are errors in the diagnosis of the disease which occurs unevenly among exposed and unexposed individuals.
• Bias towards or away from the null

Question 30

In which form of misclassification is bias towards or away from the null?

Answer 30

Differential misclassification

Question 31

What is response bias?

Answer 31

• Response- describes the tendency of participants to respond to questions in a way that is more socially acceptable. This not a type of selection bias.

Question 32

What is non-response bias?

Answer 32

• 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 33

What is the Lake Wobegon effect ?

Answer 33

• 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 34

What is the Hawthorne effect?

Answer 34

. The precise definition is controversial, but is usually described as the consequence of participants realising they are being observed and therefore acting differently. For example, if you knew you were being studied, you might behave in a slightly different way to normal!

Question 35

What is confounding?

Answer 35

Confounding- Is the effect of an extraneous variable that wholly or partially accounts for the apparent effect of the study exposure, or that masks an underlying true association.
Example-Down syndrome, birth order and maternal age.

Question 36

What are the three conditions of confounding?

Answer 36

a. Associated with the exposure in the source population
b. Associated with the outcome in the absence of the exposure
c. Not a consequence of the exposure

Question 37

What is stratification?

Answer 37

a. Stratified data by variable of interest. Compare stratified specific estimates with the estimate of the whole/pooled data set and the difference is considerable indicates confounding.

Question 38

How can comparing crude and adjusted estimates indicate confounding?

Answer 38

a. If adjusted odds ratio differs from the crude odds radio by 15% or more this may indicate confounding.
b. Not optimal method of identification of confounding

Question 39

What is effect modification/

Answer 39

Exists when the strength of the association varies over different levels of a third variable
It is a naturally occurring phenomenon and therefore should not be controlled for but stratified analysis should be undertaken

Question 40

What are the tests for effect modification?

Answer 40

Tests for effect modification
• Breslow- Day test
• Q test
• Interaction terms in regression models

Question 41

What is a crude model?

Answer 41

). This is simply looking at the impact of the exposure on the outcome – with no consideration of anything else.

Question 42

What is an adjusted model?

Answer 42

• Adjusted model (which is the multivariate analysis of a range of exposures vs. outcome). A multivariate analysis means that multiple potential exposures have been included. The inference is that the outputs of these analyses mean that holding all other adjusted variables equal, X is the association between exposure and outcome.

Question 43

What is a multivariate analysis?

Answer 43

multiple potential exposures have been included.

Question 44

What is multivariate regression?

Answer 44

Multivariate regression (which gives us this adjusted output) - is a highly effective way for us to identify possible confounding