Cohort Studies
Attrition Bias
Loss to follow up - don't know if these people developed disease or not
Open Cohort Study
People can be measured for different lengths of follow-up
Healthy Worker Effect
Problem with occupational cohorts. Form of selection bias. Those that are at work and generally in better health than those who aren't, so tend to have lower mortality, less disability etc
How to overcome Attrition Bias?
Retrospective Cohort Study (using a cohort where there is existing data on exposure and disease) Quicker and cheaper Good for long latency diseases (ones that take a long time to develop after exposure)
Survival Analysis
Used in studies where there is follow up over time Measures the number of people who survive for x years Data is shown in a survival curve (aka Kaplan Meier plot)
Strengths of Cohort Studies (5)
1) Can measure incidence of multiple diseases 2) Can measure multiple exposures 3) Can assess temporal relationship as exposure is measured before disease develops (prospective cohort studies) 4) Can be used to measure effects of rare exposures 5) Selection and Recall Bias is low
What is a Cohort Study? (5)
1) Measures incidence of disease 2) Prospective - people are followed forwards through time 3) Study population is free from disease at the start 4) Exposure variables are common and measured 5) Risk of disease between exposed and unexposed is compared
Limitations of Cohort Studies (4)
1) Prospective = time consuming, very expensive 2) Retrospective = relies on accuracy of existing records 3) Attrition Bias 4) Inefficient at assessing rare diseases
Closed Cohort Study
Everyone is followed up for the same period of time