Interpreting Epidemiological Findings (2)

Question 1

What is confounding?

Answer 1

The effect of an extraneous variable that wholly or partially accounts for the apparent effects of the study exposure, or that masks an underlying true association

a confounding variable is a third variable which leads to bias in the estimate of association between outcome and exposure

Question 2

What can confounding lead to?

Answer 2

Biased estimates

Question 3

How can you identify confounding?

Answer 3

Knowledge of the subject matter: you could undertake an evidence review and see what other people have proposed or found

Test by the three criteria: associated with exposure in the source population; associated with the outcome in the absence of the exposure; not a consequence of the exposure

Stratification: we can look at the difference of apparent effect within different population strata

Compare the crude and adjust statistical estimate

Question 4

What are the three conditions for confounding?

Answer 4

Associated with the exposure in the source population

Associated with the outcome in the absence of the exposure

Not a consequence of the exposure

Question 5

How do you conduct stratification?

Answer 5

Compare stratum specific estimates with the estimate you get when you analyse the date from the study

Question 6

What after conducting stratification might indicate confounding?

Answer 6

Pooled estimate is considerably different from what you expect from stratum specific estimates

Question 7

What is effect modification?

Answer 7

Where the magnitude of the effect of an exposure variable on an outcome variable differs depending on a third variable
- not a problem. It is a natural phenomenon.

Effect modification – exists when the strength of an association varies over different levels of a third variable
Shouldn’t be controlled for + happens naturally
Take into account when reporting
When detected – conduct stratified analysis

Question 8

How can you test for effect modification?

Answer 8

• Breslow-Day test
• Q test
• Interaction terms in regression models

Question 9

What is synergism in relation to effect modification ?

Answer 9

The effect modifier potentiates the affect of the exposure

Question 10

What is antagonism in relation to effect modification?

Answer 10

The effect modifier demeans the effect of the exposure

Question 11

What are adjusted models used for?

Answer 11

Identification of potential confounding

And used to account for it

Question 12

What is a crude model?

Answer 12

Univariate analysis of exposure vs outcome

Question 13

What does the crude model do?

Answer 13

Simply looks at the impact of the exposure on the outcome – with no consideration of anything else

Question 14

What is an adjusted model?

Answer 14

Multivariate analysis of a range of exposures vs. outcome

Question 15

What is multivariate analysis?

Answer 15

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 16

What are other ways of naming crude vs adjusted models?

Answer 16

Hazard ratio (HR) vs Adjusted Hazard Ratio (Adj. HR)
Odds ratio (OR) vs Adjusted odds ratio (AOR)
Univariate model vs multivariate model

Question 17

Going back to the ICSM Lifestyle Tracking Study last year, 20% of the participants reported using a smart watch.

Do you think this is representative of the first-year student body overall? Why?

Answer 17

Probably not. People with smartwatches may differ systematically from those who do not: it may be that they’re more interested in lifestyle and health behaviours. They may also be more comfortable and confident with technology. This is an indirect form of selection bias.

Question 18

Do you think participants with a smartwatch (the exposure) will report more, less or the same number of steps (the outcome) on average than those participants without a smartwatch? Why?

Answer 18

I’d hypothesise that participants with a smartwatch will report more steps than average for two reasons:

Measurement error among those without smartwatches: The smartwatch is always worn and therefore will capture more steps than smartphone which may be left behind during physical activity. This is an example of differential misclassification.

Actual differences: People with smartwatches are more likely to be physically active: and this might be the driver of them buying a smartwatch, or the smartwatch may cause them to be more physically active.

Interestingly the actual results showed the precise opposite which is that those with smartwatches on average reported less time being physically active than those without smartwatches. It’s possible that people without the technology over-estimate their physical activity.

Question 19

four ways to potentially identify confounders

Answer 19

1. Knowledge of the subject matter: you could undertake an evidence review and see what other people have proposed or found.
2. Test by the three criteria: associated with exposure in the source population; associated with the outcome in the absence of the exposure; not a consequence of the exposure.
3. Stratification: we can look at the difference of apparent effect within different population strata.
4. Compare the crude and adjusted statistical estimates.

Question 20

study showed birth order associated with downs syndrome..what is the confounding variable

Answer 20

• maternal age is the confounding variable.
• Further analysis (shown in the 3D column plot) demonstrates that actually birth order is not associated with Down’s if you account for maternal age.

Question 21

Study showed alcohol consumption associated with lung cancer…what is the confounding variable?

Answer 21

• alcohol is associated with tobacco consumption = tobacco consumption is associated with lung cancer - people who drink are more likely to smoke
  confounder: cigarette use.

Question 22

Several observational studies have demonstrated an association between playing games of cognition (such as Sudoku) and reduced risk of dementia in later life. what is the confounding variable?

Answer 22

It’s possible (indeed likely) that people who choose to undertake cognitively-stimulating exercises may be people who are at lower risk of dementia in many other ways: through educational level, occupation or other socioeconomic characteristic.

Question 23

What do you think is the effect modification of the following:

In the Million Women Study,1 post-menopausal women were observed in two groups: those using continuous combined hormone replacement treatment (HRT) and those not using HRT. The overall estimate suggested a risk reduction for endometrial cancer of approximately 30% for women using HRT.

Answer 23

breaking down the results by BMI category, the association was only observed among women with a BMI ≥30kg/m2. There was no apparent difference for women in the other BMI groups.

BMI effect modifier

Question 24

In a study of alcohol consumption and associated harm, what do you think is the effect modifier

Answer 24

A Scottish research group has identified socioeconomic status as a powerful effect modifier. Irrespective of alcohol consumption (volume or behaviour), socioeconomic status seems to be the most influential determinant of coming to alcohol-related harm. This is described by some as the alcohol harm paradox: even though more-affluent people drink more on average than those less-affluent, it is the socioeconomically more vulnerable who suffer the greatest harm.

Question 25

What is adjustment

Answer 25

Adjustment is how we identify and account for potential confounding

Question 26

What is multivariate regression

Answer 26

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

Question 27

Define confounding and confounding variable

Answer 27

Confounding = the effect of an extraneous variable that wholly/partially accounts for the apparent effect of the study exposure or masks an underlying true association
Confounding variable – a third variable which leads to bias in the estimate of association between outcome + exposure

Question 28

Explain critical reasoning will help identify a confounding variable if present

Answer 28

Knowledge of subject matter
E.g. knowledge of plausible biological pathways
Not always possible (e.g. novel associations)

Question 29

explain the test by condition criteria that will identify a confounding variable if present

Answer 29

Test by condition criteria

1. Associated with exposure in source pop.
2. Associated with outcome without exposure
3. Not a consequence of exposure

Question 30

explain process of stratification as a test/critical reasoning that will identify a condoning variable

Answer 30

Stratification

• Stratify by variable of interest
• Look at difference of apparent effect within different population strata
• Calculate odds ratio within each sub group
• Pooled estimate considerably different to stratum specific estimates

Question 31

explain process of comparing crude and adjusted statistical estimates as a test/critical reasoning that will identify a condoning variable

Answer 31

Comparison of crude and adjusted statistical estimates

• Adjusted odds ratio differs from crude odds ratio by ≥ 50%
• Not optimal method

Question 32

key points

Answer 32

Key points
a. Confounding is where a third variable which biases the estimate of association
between outcome and exposure.
b. Confounding can be identified and mitigated in a number of different ways, but you
need to look for it to find it.
c. Effect modification is similar to confounding: it’s a third variable, but where the
modification of association is a real phenomenon.
d. Stratification and multivariate regression are both approaches which can be used to
examine the relationship between an exposure and an outcome in the presence (or
absence) of a third variable (or more variables, in the case of multivariate regression).