Cohort Studies Flashcards
What are the two main types of studies?
Observational and interventional
Give an example of interventional studies.
RCTs
Given examples of observational studies. (3)
Cohort
Case-control
Cross-sectional
How are cohort studies done?
Identify individuals
Measure exposures in each individual
Follow-up individuals to determine disease/disorder occurrence
Relate information on disease occurrence to exposure
Explain a cohort study simplistically. (2 words)
Cohort –> disease
How do cohort studies differ from case-control studies?
Cohort –> disease FOR COHORT
Exposure
Why might there be bias in cohort studies? (3)
Loss to follow-up
Exposure usually measured at just one time point
Selection of cohort
What are the disadvantages of cohort studies? (3)
Take a long time
Need a lot of people
Very expensive
What was the British Doctors Study?
Information from 34,439 male doctors initially collected in 1951 and they were followed up for 50 years. They were asked about smoking. Death certificates obtained when they died.
What is incidence?
What is it usually taken to be a measure of?
Number of new cases (or deaths) of a disease per 100,000 people per year.
Risk
How do you calculate relative risk?
Incidence of disease in exposed population/incidence in unexposed population
How can we tell if the results have arisen due to chance alone?
Look at p-values and confidence intervals
If the risk of a disease is the same for the exposed and for the unexposed populations, what is the relative risk?
1
Explain what a 95% CI of a sample relative risk means.
The 95% CI of a sample relative risk contains the population relative risk with a probability of 95%
Explain what a 95% CI of a sample mean means.
The 95% CI of a sample mean contains the population mean with a probability of 95%
When can you tell that the results (relative risks/ratios) are statistically significant looking at the 95% CI?
If the CI doesn’t include 1 (for means, it is 0 but for risks it is 1).
What is a confounder?
A factor that is associated both with the exposure and also with the disease.
How do you account for confounders?
Statistical adjustment
e.g. with alcohol and smoking. Analyse alcohol consumption and risk of lung cancer in non-smokers, and alcohol consumption and risk of lung cancer in smokers. You might expect these risks to be similar, in which case you could combine them.
Which should you use – unadjusted or adjusted RR?
Adjusted - this is adjusted for confounders.
What curves are used for survival analysis? What do these plot?
Kaplan-Meier curves - they plot proportion of people surviving over time.
What does the log rank of a Kaplan-Meier curve compare?
The two curves for the exposed/unexposed population
How do hazard ratios differ from relative risks?
The take into account the time taken for the event to occur
What are hazard ratios?
They give the risk of dying at any time point in one group compared to the other
What two measures of importance are given in this lecture?
Absolute excess risk
Attributable proportion
How do you calculate the absolute excess risk?
Risk in exposed – risk in unexposed
How do you calculate the attributable proportion?
Incidence in population attributable to exposure/incidence in population
What is the formula for attributable proportion?
1 + p (RR-1)
where p = proportion exposed in population
