Lecture 6- Study Design 3

Question 1

What is a case control study?

Answer 1

-Observational study, generally carried out to test hypotheses (analytic)
-Participants are chosen on the basis of their outcome status: a group
with the outcome (cases) and a group without (controls). Information is collected from people with and without outcome about exposures that occurred in the past (retrospective). i.e. in general before disease was diagnosed.

Question 2

What are the advantages of case control studies?

Answer 2

• Relatively quick as you are working at the outcome level so don’t have to wait for the condition to develop
• Smaller than cohort studies, particularly for rare outcomes.
• Can examine the effects of multiple exposures.

Question 3

What are the disadvantages of case control studies?

Answer 3

• Events have already occurred so the potential for bias is higher (asking people to recall events from the past is difficult due to memory difficulties + wanting to view themselves in the best light possible)
• It is very hard (if not impossible) to remove all the effects of confounding.

Question 4

Which of these three hypotheses are the best and why?
1 Vitamin D deficiency is associated with increased risk of
cardiovascular disease. Are people with lower vitamin D levels more likely to develop cardiovascular disease?
2 Vitamin D deficiency causes an increased risk of cardiovascular
disease. Do low vitamin D levels cause an increase in risk of cardiovascular
disease?
3 In people with vitamin D deficiency, supplementation with 400 IU of
vitamin D per day over a 3 year period reduces the risk of
cardiovascular disease

Answer 4

The third one as it is specific enough and is actually asking what we want it to ask. For the first question the problem is that we want to determine cause not a correlation.

Question 5

How do you calculate and interpret relative risk?

Answer 5

e.g. for risk of disease while taking a pill.

divide the proportion of those who took the pill and developed the disease by the proportion who didn’t take the pill and developed the disease.

1= risk in one group is the same as the other
If less than 1 then risk is reduced
If greater than 1 have an increased risk

Question 6

What factors influence whether a calculated relative risk will match the true population relative risk?

Answer 6

Confounding, Bias, Chance (random variation)

Question 7

What is confounding?

Answer 7

A distortion of the association between exposure and
outcome caused by the presence of a third factor. A confounder is a
variable which causes this distortion. To be a confounder, a variable must
be both associated with the exposure (independent of outcome);
and associated with the outcome (independent of exposure);
Furthermore, it must not just be an intermediate link in the ‘causal
pathway’.

Question 8

What two types of bias can occur in an analytic study?

Answer 8

• Selection bias

- Information bias

Question 9

What is selection bias?

Answer 9

-Systematic error arising from the way participants are selected for
inclusion in the study.
-In an analytic study, selection bias occurs if the selection processes
cause a systematic difference between the groups of participants
selected for the study.

Question 10

Is ‘selection bias’ a thing in prospective analytic studies?

Answer 10

Prospective analytic studies rarely obtain participants through random
sampling from a population (self selection occurs). Because of this the issue of representativeness must be considered, but we consider it a generalizability (e.g. the doctor study- are the findings applicable to the wider population?) issue rather than bias.

Question 11

What is information bias? What might it be introduced by?

Answer 11

-Systematic error arising from the way study information is obtained,
interpreted and recorded.
-In an analytic study, information bias is a particular problem if there is a difference in the level of information bias across two groups. If they both have the same level then comparison is not as much of an issue
-Information bias may be introduced by the:
Observer
Study individual (respondent)
Instruments used to collect the data (e.g. badly-designed questionnaire)
Missing measurements (e.g. from loss to follow-up in a prospective
study)

Question 12

What is a participant not responding to a question considered as and how does this differ from a participant simply not wanting to participate in the study?

Answer 12

If don’t answer question= information bias

If refuse participant= selection bias

Question 13

What are the two types of classification used in defining studies?

Answer 13

• Classification by purpose of study:
  e. g. Descriptive (describe things) vs. analytic (testing hypotheses).
• Classification by form of the design:
  e. g. Experimental (researcher intervenes) vs. observational (researcher
  observes) .

Question 14

Using the key words for designing studies RCTs are….

Answer 14

Analytic, experimental, prospective

Question 15

Using the key words for defining studies cohort studies are…

Answer 15

Analytic, observational, usually prospective

Question 16

Using the key words for defining studies case-control studies are…

Answer 16

Analytic, observational, retrospective