Week 3 - Case-Control Studies

Question 1

Why type of study is a Case-Controlled Study?

Answer 1

Observational.

Question 2

Is there any intervention from the researcher in a Case-Controlled Study?

Answer 2

No.

Question 3

What does a Case-Controlled Study start with (of participants)?

Answer 3

Stars worth the outcome of interest (those with the diagnosis).
And then they look back to see what it may have been caused by.

Means it is a retrospective study.

Question 4

What are the groups like in a Case-Controlled Study?

Answer 4

The Cases -
These are the patients with a condition that is being studied.

These are compared to…

The Controls -
Participants who don’t have the condition.

Question 5

Who are the patients of the cases group compared with?

Answer 5

The control patients, who don’t have the condition.

Question 6

Who gets matched together?

Answer 6

Patients with the particular disease/condition are identified and compared with others who dont have the disease/condition (the controls).

Question 7

How should you select the study’s control group?

Answer 7

Pick people who are a good representation of the whole population that produced the cases group.

Question 8

What do patients of the control group represent?

Answer 8

The proportion of the population two have the exposure and are at ‘risk’ of becoming cases.

(They don’t have the disease under investigation, but would be in the study’s cases group if they had).

Question 9

What happens if a particular exposure was more common in cases than controls?

Answer 9

Then it may be a risk factor for that outcome.

E.g. linking lung cancer with smoking.

Question 10

In a Case Control Study, we should assess and compare the exposure of both groups to the risk factor in equation.

Question 11

What would we compare?

Answer 11

How frequently the exposure to a risk factor(s) is present in each group.

This will determine the relationship between the risk factor and the disease.

Question 12

There can be multiple risks factors.

Question 13

Is the risk factor always bad?

Answer 13

No.
Sometimes it can be beneficial to the patient (e.g. aspirin, vitamin d) which would reduce the incidence of the disease.

Question 14

What thing can we compare of between both groups to look for any associations?

Answer 14

Lifestyles.
Medical histories.

E.g. possible exposures to the risk factor.

Question 15

How is data obtained in Case Controlled Studies?

Answer 15

On past exposures.

Question 16

What does it mean if a particular exposure is more common in cases than controls?

Answer 16

Then it may be a risk factor for that outcome.

Question 17

What is a Case Controlled Study useful for?

Answer 17

Studying rare diseases.
And looking at the outcomes they may have.

As they can easily recruit a large number of patients who already have the disease, but not ones developing or about to develop it (which a cohort study would struggle with).

Question 18

Design of a Case Controlled Study -

Answer 18

Group 1 - Cases:
They have the disease.
Half were exposed to the risk factor for the disease in the past.
Half weren’t exposed to the risk factor for the disease in the past.

Group 2 - Controls.
They do not have the disease.
Half were exposed to the risk factor for the disease in the past.
Half were not exposed to the risk factor for the disease in the past.

Question 19

A Case Control Study is backwards tracing.

Answer 19

It’s retrospective.

Each groups current status is acknowledge from past data.

Question 20

What will identifying the cases help to avoid?

Answer 20

Bias.

Otherwise, if a participant is misallocated, then it may be substantially influence the results.

Question 21

Cases and Controls should be selected irrespective of what?

Answer 21

Their exposure status.

Question 22

Why may it be hard to recruit suitable controls?

Answer 22

The control group should need similar characteristics to the cases (except with regard to their disease status).

Question 23

There should be an equal number of cases and controls.

Question 24

Controls should be matched to the cases using specific criteria.

Question 25

How is the information in this type of study collected?

Answer 25

Retrospectively.

Question 26

What will collecting data retrospectively create?

Answer 26

Bias.

Question 27

What 2 types of bias may it cause?

Answer 27

Recall bias. Observer bias.

Question 28

What’s Recall bias?

Answer 28

Participants don’t remember past events accurately. This may be due to the extreme state of illness they were in at the time affecting memory, like past medications they were on at the time. And the researcher may need them to use more than one method to confirm their exposure to the disease etc.

Question 29

What’s Observer bias?

Answer 29

Condition of the patient is known to the researcher. So researcher may act differently around them. But it can be hard to avoid this sometime, e.g. if the participant is clearly ill.

Question 30

Do Case Control Studies show causation or association?

Answer 30

May show association. But not causation (i.e. that the certain risk factor definitely caused the disease). So have to think - e.g. Does unemployment trigger depression? Or… Does depression make it more difficult for people to find work?

Question 31

What are Case Control Studies useful for studying?

Answer 31

Possible risk factors for uncommon / rare diseases or outcomes.

Question 32

What is data obtained from?

Answer 32

Past exposures. Done through looking through e.g. medical records.

Question 33

What should we look out for?

Answer 33

Confounding factors.

Question 34

What’s a confounding factor?

Answer 34

An additional factor associated with both exposure and outcome.

Question 35

What is a way to account for confounding factors?

Answer 35

Matching cases with controls very carefully. E.g. based on age, sex, socio-economic status.

Question 36

Examples of some clinical questions which could be answered by a Case Control Study?

Answer 36

Does the prone sleeping position increase the risk of cot death in babies? Does whooping cough vaccine cause brain damage? Did eating at a particular pizza restaurant in a particular time period cause hep A? Is smoking linked with lung cancer? Do overhead power cables cause leukaemia? Does alcohol consumption increase breast cancer?

Question 37

Look at onenote for a Case Control Study example.

Answer 37

And for ones in medical imaging too.

Question 38

Statistics - Odd Ratios - When are they used?

Answer 38

In case control studies.

Question 39

Statistics - Odd Ratios - What is meant by Odds?

Answer 39

The number of times an event happens / number of times it didn’t happen.

Question 40

Statistics - Odd Ratios - What does the Odds Ratio compare?

Answer 40

The odds of exposure to the risk factor amongst other cases to the factor amongst controls.

Question 41

What does an odd ratio of 1.0 mean?

Answer 41

Odds of exposure among the cases group is the same as the odds of exposure among the controls group. This means the exposure isn’t associated with the disease.

Question 42

What does an odds ratio of a lot less than 1.0 mean?

Answer 42

The odds of exposure among the cases group is lower than the odds of exposure in the controls group. So the exposures may be protective against the disease.

Question 43

Look on onenote for table of statistics: odds ratios for formulas on how to calculate them.

Question 44

Look at onenote for a case control study worked example.

Question 45

What are some Advantages of a Case Control Study?

Answer 45

It’s good for studying rare conditions/disease. It’s good for studying the effect of exposures on disease outcomes with long latency periods. It takes advantage of already existing medical records. Less time is needed to conduct this type of study (because cases are identified at the start) (and because it is retrospective - so no long periods of follow ups). They’re good for exploring many risk factors or exposures that may affect disease outcome. They’re useful in cases with cohort studies.

Question 46

What are some Disadvantages of a Case Control Study?

Answer 46

They often rely on participant memory, but people with a condition will be more motivated to recall only risk factors (recall bias occurs with this). It’s limited to investigating the effect of exposures on only one outcome. It may be hard to find a suitable control group. They don’t provide data on incidence within the population - as the study doesn’t find every case possible. They’re don’t establish causations with the risk factor. They aren’t useful for the effect of rare exposures on only one outcome.

Question 47

Look at table in onenote for comparing factors between a Cohort Study and a Case-Control Study.

Question 48

A case control study is a form of an observational study.

Question 49

A case control study aims to identity risk factors for developing an outcome of interest.

Question 50

Subjects with the outcome (cases group) and those without the outcome (controls) are selected. Risk factor exposure measurements are collected retrospectively from both groups (either from the subject or any valuable records).

Question 51

Cases selected should be representative of all cases of the outcome in the population.

Question 52

Cases and controls should be as similar as possible to each other. Age, gender, demographics.

Question 53

The odds ratio indicates the increased (or decreased) odds of the disease being associated with the exposure of interest.