CAT: CASE CONTROL

Question 1

Whats the benefit of a population-based method for recruiting cases?

Answer 1

It minimises selection bias by including all cases

Question 2

Why is it good to use an expert system to confirm cases?

Answer 2

It ensures accurate case classification - reduces misclassification bias

Question 3

Why is using recruitment criteria good for selecting cases?

Answer 3

It ensures comparability of cases

Question 4

Explain in general terms the key issue in the recruitment of controls in relation to the recruitment of cases in case-control studies?

Answer 4

The controls and cases must come from the same population

Question 5

How can you increase the validity of recruitment of controls?

Answer 5

Match them with the cases on key variables

Question 6

What is confounding?

Answer 6

A distortion of an association between an outcome and exposure by a third factors which has an association with both exposure and outcome

Question 7

Explain in general terms what thr consequence would be for a research study if all important confounders are not identified?

Answer 7

Association between exposure ans outcome could be biased

Question 8

What does ‘adjusted’ mean?

Answer 8

The analysis has taken into account the effect of other possible confounding variables

Question 9

How can you increase the precision of a study?

Answer 9

Increase sample size
More cases
More controls
Less measurement error

Question 10

Give examples of different information you might look for in the main paper to help you answer the following question “can results be applied to your local population”?

Answer 10

Description of the study sample to allow comparison with the local population
Consideration of whether pattern of exposures in the study was likely to be similar to the local population
Information to allow assessment of internal study validity - if not valid then its not possible to apply externally

Question 11

What is selection bias?

Answer 11

occurs when individuals or groups in a study differ systematically from the population of interest leading to a systematic error in an association or outcome.

Question 12

What is measurement bias

Answer 12

any systematic or non-random error that occurs in the collection of data in a study

Question 13

What is protopathic bias?

Answer 13

Aka confounding by indication
A form of bias that can occur in observational studies
It happens when a treatment or exposure seems to be causing an outcome, but the treatment was actually initiated in response to early signs or symptoms of the outcome, not the other way around. I.e. occurs when a pharmaceutical agent is inadvertently prescribed for an early manifestation of a disease that has not yet been diagnostically detected.

Question 14

What is recall bias?

Answer 14

a systematic error that occurs when participants do not remember previous events or experiences accurately or omit details

Question 15

How can you reduce recall bias?

Answer 15

Reduce the time interval between the event under study and its assessment
Whenever possible use objective pre-existing records rather than relying on participant experiences
Use a control group

Question 16

What is a case-control study?

Answer 16

A type of observational study that compares a group with the condition you are interested and one group that does not have the condition. Researchers study the history of these people in each group to learn what factors may be associated with the disease

Question 17

What is a nested case-control study?

Answer 17

A case control study where cases and controls are drawn from the population of a cohort study

Question 18

Important points on selecting cases for a case-control study?

Answer 18

Should be done using explicitly defined inclusion and exclusion criteria
Cases should be directly sampled from the source population
Case selection process and data sources from which they were selected should be described in detail
When cases are selected from a secondary data source the case definitions should be supported by previous validation studies

Question 19

Important points on selecting controls for a case-control study?

Answer 19

May be more than 1 control for each case
Should be selects from the same shire population as the cases
Case ans controls are often matched on 1 or more pt characteristic to control for potential confounding. Dont over match as this can cause bias. Dont match with variables that the investigator is interested in examining!

Question 20

Important points on defining drug exposure grous for a case-control study?

Answer 20

Drug exposures and their definitions should be clearly specified in methods
The look back period must be defined and should be comparable for controls
If investigators are collecting exposure data themselves, the outcome status should be blind

Question 21

How are results usually expressed for a case-control study?

Answer 21

Odds ratios

Question 22

Strengths of case-control studies?

Answer 22

Statistically efficient as uses less data to quantify a drug-outcome association than would be required for a cohort study
Efficient for rare outcomes
Efficient for long latency periods
Efficient for handling the time varying nature of drug exposure
Low cost
Less resources and time required
Less ethical concerns when studying harmful exposures

Question 23

Weaknesses of case-control studies?

Answer 23

Not efficient for rare exposures
Difficult to select unbiased controls
Unable to directly calculate incidence rates
Can’t be used to measure causal effects of drugs as observational studies are too susceptible to bias and confounding
No randomisation or blinding
Subject to selection bias
Subject to recall bias as retrospective
Dont allow calculation of incidence

Question 24

What bias are case-control studies exposed to?

Answer 24

Selection
Measurement
Recall
Classification
Observer bias

Question 25

What is berksonian bias?

Answer 25

Berksonsonian bias is a type of selection bias.
It can arise when the sample is taken not from the general population, but from a subpopulation. E.g. jus taking cases from a hopsital can cause a systematic error where exposed cases may be more likely to be admitted and therefore selected into the study