Cohort Studies Flashcards
Definition of cross sectional studies
Exposure status and disease status of an individual are measured at 1 point in time
Compared to those without disease/different exposure levels
Definition of prevalence ratio
Prevalence of disease in exposed divided by prevalence of disease in unexposed
Definition of odds
Odds that a diseased person was exposed or unexposed
Definition of odds ratio
Ratio of prevalence odds in exposed to prevalence odds in unexposed
Definition of risk
Probability of occurrence of disease in a disease free population during a specified time period
Definition of incidence rate/density
How rapidly cases occur
Definition of incidence proportion/cumulative incidence
No of cases at the end of the given period
Definition of survivor bias
Selection bias where results of the survivors of an outcome are disproportionately evaluated
Those who didn’t survive are ignored
Definition of selection bias
Bias introduced by selection of individuals so sample obtained is not representative of the whole population
Definition of sampling bias
Not all populations represented in the sample
Definition of ascertainment bias
People know what group they are in
definition of participation bias
Participant acts in a way that they think the researcher wants them to act
Definition of misclassification bias
People have been assigned to the wrong groups
Definition of ecological fallacy
The results of 1 person assumed to be true for the whole grou[
What are cross sectional studies
What groups are compared
When are these studies used
Exposure status and disease status of individual ensured at 1 point in time
Disease prevalence in those with and without exposure/at different exposure levels are compared
Used in health planning
What are the 3 ways of analysing the data in a cross sectional study
2x2 tables developed and measures calculated
Calculate
-prevalence ratio (prevalence of disease in exposed/prevalence of disease in unexposed)
- odds (disease exposed)x(no disease unexposed)/(disease unexposed)x(no disease exposed)
- odds ratio
How do you interpret the prevalence ratio
People with (condition) are (prevalence ratio) more likely to be infected with (disease) than people without (condition)
How do you interpret the odds ratio
Odds that a person with (disease) is (odds ratio) more/less likely to have (condition) than than those without the (disease)
What are the 5 advantages of a cross sectional study
Quick and easy Data collected once Measure prevalence for many factors Multiple outcomes and exposures Can generate many hypotheses
What are the 6 disadvantages of cross sectional studies
Difficult to determine time order
Unsuitable for rare diseases (sample size too small)
Reflects determinants of survival and etiologies
Can’t measure incidence
Hard to interpret results
Susceptible to bias
What are cohort studies
Prospective follow up studies
A group of people who share a common experience/condition are followed up to determine disease incidence
How would you use cohort studies to investigate etiologies
Study risk factors for developing disease
- behavioural
- occupational
- environmental
How would you use cohort studies in prognoses
Factors that predict mortality or disability
Factors that alter health outcomes
-treatments
-study of those diagnosed with a disease
What are the 3 assumptions made in a cohort study
Representative of population
Absence of exposure well defined
Comparability of outcomes
What are the 3 types of cohort study
Prospective
Retrospective
Ambidirectional
What are the 3 main characteristics of a prospective cohort study
Identifies population and exposure status
Follows up for disease development
Completion of study takes a long time
What are the 4 main characteristics of a retrospective study
Existing data identifies exposure status
Identifies present disease status
Short time to assemble study population
Determine disease at present
What are the 4 main characteristics of an ambidirectional study
Existing data identifies population and exposure
Follow up into future for disease
Short time to assemble study population
Additional time to follow disease development
How would you calculate risk and the risk ratio
Probability of occurrence of disease free population during specified time period
Risk = new cases over time period/population initially at risk
Risk ratio = risk in exposed/risk in unexposed
=a/(a+b) / c(c+d)
How do you interpret risk in cohort studies
Over a (time period), (no of new cases) out of (population initially at risk) at risks persons developed the disease
How do you calculate the rate and rate ratio in cohort studies
Measures the frequency of new cases
Can account for changes in population size (use midpoint population)
Rate = new cases in a population in a time period/total no of person years
Rate ratio = rate in exposed/rate in unexposed
What is the difference in incidence rate and incidence proportion
Incidence rate
-how rapidly cases occur
Incidence proportion
-no of cases at the end of the given period
How would you interpret the relative risk figures
=1
-no association
> 1
-Risk in exposed > risk in unexposed (risk factor)
<1
-Risk in exposed < risk in unexposed (protective factor)
What are the 5 advantages of cohort studies
- Temporal sequence between assumed cause and outcome
- Can investigate many outcomes associated with 1 exposure
- valuable in study of rare exposure
- reduce risk of survivor bias
- allow calculation of incidence rates, RR, CI
What are the 4 disadvantages to a cohort study
- Inefficient for rare diseases (can take a long time to develop)
- Require large populations
- Loss to follow up
- Expensive and time consuming
What are the 3 main types of bias in cohort studies
Selection bias
Information bias
Other
What are the 3 sources of selection bias
Sampling bias
Ascertainment bias
Participation bias
Wat are the 2 forms of information bias
Misclassification bias
Ecological fallacy
What are the 2 other sources of bias
Confounders
Chance
What 6 factors should you consider in an evaluation of association
Temporal sequence
Strength and consistency of association
Dose response relationship
Biological plausibility
Bias
- Amount of selection bias
- Completeness of follow ups
- assessed and controlled for confounders
- sample size
