Study Designs and Measures of Association Flashcards

1
Q

Define a cohort study.

A

Comprises of people with common characteristic.

Exposed and unexposed cohorts.

Purpose of cohort is to measure occurrence of one of more specific diseases during follow-up with aim of comparing disease rates.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What is an experiment?

A

A study in which incidence rate or risk of disease in two or more cohorts is compared after assigning exposure to people.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What is the aim of a clinical trial?

A

To evaluate the incidence rate of disease complications in cohorts assigned to difference treatment groups. through randomization.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What is incidence rate?

A

Number of new persons developing disease within a time period / person years.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What is a field trial?

A

Test whether intervention works in real life settings.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What is community intervention trial?

A

Exposure is assigned to groups of people rather than individuals.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What is one requirement of a population at risk?

A

They must all be free of disease at outset.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What is the difference between a closed an an open cohort?

A

A closed cohort has a fixed membership. Once members are identified, there can be no new members. Will always become shorter with time.

An open cohort can take additional members as time passes.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

How do you count disease events in a closed versus open cohort?

A

Closed cohort: risk.

Open cohort: incidence rate.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Define a case-control study.

A

Two groups differing in outcomes are identified and compared on the basis of a supposed causal attribute.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What is the purpose of a case control study?

A

To determine if exposure is associated with an outcome.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What is calculated from a case control study?

A

Odds ratios.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What is a case crossover study?

A

Each PPT acts as their own control. Control is a sample of time before exposure.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What is a longitudinal study?

A

Information pertains to more than one point in time. It is implicit that the causal action comes before the development of the disease as consequence to exposure.

All cohort studies are longitudinal.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What is a cross-sectional study?

A

Measures prevalence of characteristics. Cannot measure incidence because risk calculation requires information across time.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What do case control studies help us answer?

A

If QUANTITY of one variable is different based on value of another.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

What do cohort studies help us answer?

A

Compare MAGNITUDE of difference in RISK in one group versus another.

18
Q

What is incidence rate?

A

Number of new cases/person time

19
Q

What is incidence proportion/cumulative proportion/cumulative incidence?

A

Number of new cases/total number of people at risk

20
Q

What is risk ratio?

A

Compares two incidence proportions

21
Q

What is incidence rate ratio?

A

Ratio of two incidence rates.

22
Q

Are cohort and case control observational or experimental?

A

Observational.

23
Q

What do cohort studies measure?

A

INCIDENCE (both incidence rate and incidence proportion).

Exposed versus unexposed.

24
Q

What is risk ratio not a fair comparison?

A

Differential loss of follow up (such that loss of follow up is due to exposure and/or outcome status).

25
Q

Describe how PPTs are chosen in a cohort study?

A

Based on exposure.
They must be exposed to a disease.
Ex: COVID-19, everyone at risk. Lung cancer-smokers.

Followed forward in time and see how many people develop an outcome of interest. Must be at risk and free of disease at beginning of study.

26
Q

What does it mean to be at risk?

A

Healthy but no immunity from previous disease occurrence.

27
Q

What do cohort studies measure?

A

Incidence of disease.

28
Q

What is the risk ratio?

A

Compare incidence proportion in exposed group to unexposed group.

Number of new disease/number at risk divided

Incidence proportion/cumulative incidence/cumulative proportion

29
Q

What is incidence rate ratio?

A

Compare incidence rate of exposed group to that of unexposed group.

Number of new disease/person time

More nuanced way of looking at risk as time incorporated into the rate.

30
Q

When is incidence proportion/cumulative incidence appropriate?

A

When everyone completes the study OR if there are similar proportions of people in each group completing study.

31
Q

Why is incidence rate better assessment of risk?

A

Different patterns of follow-up in exposed and unexposed populations. Measure equalizes TIME AT RISK.

32
Q

Why should we NOT use odds ratio in cohort study?

A

Higher prevalence of disease, odds will be more divergent. Will overstate the risk.

Odds are trickier to understand and will overstate association when prevalance is higher than 20%.

33
Q

How do we choose people for case-control?

A

Based on outcome. Move backwards in time to see if people exposed or not.

34
Q

Describe PPTs in case control?

A

Cases must be similar to people with disease in underlying population. Controls must be similar to people without disease in underlying population.

Source population needs to be representative.

Controls should be similar to cases in every way except disease outcome.

BUT they must have similar risk of exposure.

35
Q

What is the cardinal rule of case control?

A

All PPTs must have equal risk of having been exposed.

36
Q

When should we use case control?

A

For rare disease.

When emerging.

Outbreak needs answers quickly.

Cannot wait for cohort study.

37
Q

Why can we not estimate risk in case control?

A

Because we choose PPTs based on outcomes. Only OR can be calcualted to measure association between outcome and exposure.

Distort true ratio of people with and without disease in source population.

Magnitude problem but association true.

38
Q

Explain bias in case control.

A

Selection bias:
Controls may be more or less likely than underlying population to have disease.
Cases more likely to agree to be in study BC the study is about them.
Cases not representative of all cases.

Information bias:
Cases more likely than controls to remember exposure
Cases harder to reach if very ill or dead

39
Q

How to calcualte OR?

A

Can cross multiply.

P/1-P = Odds
Odds/1+Odds = P
40
Q

What is towards the null?

A

Towards 1.

41
Q

What is away from null?

A

Away from 1.