cohort study

Question 1

cohort study

Answer 1

group of people followed overtime to measure the incidence oif disease
compare incidence rates of exposed and unexposed

Question 2

uses

Answer 2

follow up people with a particular condition/look at netural history/survival rates
identify causal factors
explore association between factors to identify causal pathways
identify biomarkers
evaluate outcomes in people with different diseases

Question 3

selection of cohort

Answer 3

random selection 
defined by life event 
occupational group
specific exposure 
defined by one disease to measure the incidence of other or of complications

Question 4

2 types of cohort

Answer 4

prospective - identify people and follow them up
retrospective - disease already occurred but the group is defined by exposure
both longitudinal

Question 5

measurement of exposure

Answer 5

questionnaires
records
baseline tests
need to collect repeated measurements of exposures to categorize patients

Question 6

meaurements of outcome

Answer 6

record linkage - details tagged
link medical records
regular re-interview or examination

Question 7

analysis and interpretation

Answer 7

calculate incidence in exposed and unexposed groups

Question 8

incidence(cumulative incidence)

Answer 8

number of new cases of a disease in a group or population over time

Question 9

incidence rate (cumulative incidence rate)

Answer 9

number new cases of disease in a defined period divided by the size of the population
proportion rather than a true rate

Question 10

relative risk =

Answer 10

incidence in exposed/incidence in unexposed

Question 11

attributable risk

Answer 11

absolute measure with the same units as incidence rate eg per person per year
indicates number people among the exposed group that could theoretically be prevented if the exposure was eliminated - assuming risk is causal

Question 12

incidence =

Answer 12

number new cases/number disease free people at the start of the time period

Question 13

attributable rate =

Answer 13

incidence rate in the exposed - incidence rate in unexposed

Question 14

dynamic cohort

Answer 14

variation if follow up time is accounted for
people recruited at different times and followed to point when leave the cohort eg: acquire disease, lost to follow up, study ends
each person contributed a certain amount to the cohort - summed to make a total time - ‘person years at risk’ this is used as a denominator for incidence which is tehn a true rate rather than a proportion and the association is described as the rate ratio

Question 15

things to watch out for in a cohort study

Answer 15

usually gold standard but there are things to look out for:
selection bias - reduce generalisability
exposure measurement accurate, unbiased, detailed and comparable quality in both groups
misclassification of the outcome could lead to bias
how long was the follow up
confounding
magnitude of association between exposure and disease - could association be due to chance
dose response gradient?

Question 16

strengths

Answer 16

temporal sequence 
measure incidence 
multiple outcomes 
rare exposures 
minimise selection and recall bias

Question 17

weakness

Answer 17

long time
non-participation if require a lot of surveys
loss to follow up - bias
expensive
inefficient for rare outcomes unless LARGE cohort