week 5 cohort and case control studies

Question 1

what is the time dimension of a cohort study

Answer 1

looking at people over time from exposure to outcome

Question 2

how is a cohort study designed?

Answer 2

two groups exposed and unexposed
then divided into who develops the disease and who doesn’t in each group

Question 3

when does a prospective cohort study start

Answer 3

after the exposure before the outcome or before the exposure and outcome

Question 4

when does a retrospective cohort study start

Answer 4

after exposure and outcome have occured

Question 5

what are potential biases in cohort studies

Answer 5

• from non-response or loss of follow up
• from bias in assessment of outcome
• information bias - different info gathered for exposed and unexposed

Question 6

the major potential biases in cohort studies is

Answer 6

the cofounding variable that alters the association between exposure and disease

Question 7

cohort study advantages

Answer 7

• Can establish population-based incidence
• Can examine rare exposures (e.g., asbestos →
  lung cancer)
• Temporal relationship can be inferred (prospective design)
• Time-to-event analysis is possible
• Multiple outcomes can be studied

Question 8

cohort studies disadvantages

Answer 8

• Lengthy and expensive
• May require very large samples
• Not suitable for rare diseases or diseases with long latency
• Unexpected environmental changes may influence the association
• Loss-to-follow-up biases
• Sub-clinical disease may go undetected at baseline

Question 9

design of case control studies

Answer 9

opposite of cohort studies
you take two groups - one has the disease “cases” and one does not have the disease “controls”
* then you determine whether or not members of each group were exposed or not to exposure

Question 10

can we estimate the prevalence of disease in a population from a case control study?

Answer 10

No, you cannot directly estimate the prevalence of a disease in a population from a case-control study. Normall you would use population surveys or cross sectional studies?

Question 11

can case control studies study multiple exposures

Answer 11

yes

Question 12

Case control studies are good for studying….

Answer 12

rare diseases, new diseases, diseases of little know etiology can look at multiple exposures that contribute, diseases that need medical care

Question 13

what does a case definition need to be

Answer 13

Need clear diagnostic criteria with high sensitivity and specificity
* Sensitivity = correctly identify those with disease
* Specificity = correctly identify those without disease

Question 14

what is an ideal case study

Answer 14

ALL cases in a defined population over a specified period of time
(or a representative sample of ALL)
* Population can be defined geographically or by membership (e.g., occupational group)
* Selection process can be facilitated by surveillance registry
* But, can be very expensive and thus unfeasible

Question 15

how do you conduct case selection participants

Answer 15

Most common:
* All (or representative sample of) cases diagnosed (or receiving care) in certain facilities within a specified time period or death certificates

Question 16

how to conduct control selection

Answer 16

• Controls should be “free of disease, similar in all other respects”
• Controls should be selected in an unbiased manner from those individuals who would have been included in the case series, had they developed the disease under study

Question 17

most common control selection for case studies is

Answer 17

people at home nonhospitalized
or hospital or clinic patients with unrelated cases of disease

Question 18

pros of case control studies

Answer 18

Relatively inexpensive
* Compared with prospective cohort studies
Can study multiple exposures at once
* Including investigations of interactions among exposures
Can be conducted in relatively short time period * Usually at least 1 year – up to a few years
Generally require relatively small numbers of cases and controls for study

Question 19

cons of case control studies

Answer 19

• Not well suited to study weak associations
• • Hard to distinguish between a true weak association and one due to bias
• You need a fair sample size
• • Often potentially differential response rates by exposure status for cases and controls, leading to selection bias
• Misclassification of exposure
• • Recall bias – diseased folks remember more exposure than controls
• • Poor recall – cases and/or controls can’t remember their exposures well enough
• • Reporting accuracy bias - differential reporting due to disease
• • Other forms of information bias

Question 20

Comparing cohort vs case control studies

Answer 20

cohort study compares incidence rates of disease in exposure and nonexposed
case control study compare the proportions exposed in people with the disease and without
* cohort studies can determine incidence and case control cannot