CAM201 Lecture Notes - Lecture 16: Prospective Cohort Study, Cohort Study, Sampling Bias
Observational Studies
Definitions
E= exposure: can be a multitude of things (risk factor, intervention).
D= disease: can include other outcomes (e.g. obesity, injury, complications)
Eo and Do= individuals without exposure or disease
Ei and Di= individuals with exposure or disease.
What is the purpose of the study?
• Descriptive: establishes the burden of disease (D) or the extent of exposure to a risk factor.
E.g. how many people in the community smoke.
• Analytical: tests hypotheses about disease causation or effectiveness of intervention. E.g. does
smoking cause cancer?
Level of Control
• No control over the exposure status of individuals
• Whereas when you do an experimental study the investigator determines who does and
does’t get the eposure.
Most studies of the cause of diseases in humans are observational (as often randomised control
trials are unethical!).
Directionality
• Determined by the time sequence used to collect information on E and D.
a. Single point in time: cross sectional study.
b. Forwards from E to D: (Follow forwards): cohort studies (categorise them, and then
follow them into the future to determine if the get the disease)
c. Backwards from D to E: case-control study (retrospective).
Sample Selection
The sample can be selected on the basis of their exposure status, disease status, or neither.
1. Cohort: exposure status
2. Case-control: disease status
3. Cross-sectional: neither (represent target population without regard to E or D initially).
Timing in Study Design
Relationship between when the study begins and the occurrence of exposure and outcome
• Retrospective: e.g. Case-control, cross-sectional, and retrospective cohort studies. When the
study starts today but E and D status is in reference to the past.
• Prospective: e.g. prospective cohort study. When the study starts today and exposure status is
determined as subjects are recruited into the study. Disease status is followed into the future.
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
E= exposure: can be a multitude of things (risk factor, intervention). D= disease: can include other outcomes (e. g. obesity, injury, complications) Eo and do= individuals without exposure or disease. Ei and di= individuals with exposure or disease. What is the purpose of the study: descriptive: establishes the burden of disease (d) or the extent of exposure to a risk factor. E. g. how many people in the community smoke: analytical: tests hypotheses about disease causation or effectiveness of intervention. Level of control: no control over the exposure status of individuals, whereas when you do an experimental study the investigator determines who does and does(cid:374)"t get the e(cid:454)posure. Most studies of the cause of diseases in humans are observational (as often randomised control trials are unethical! The sample can be selected on the basis of their exposure status, disease status, or neither: cohort: exposure status, case-control: disease status, cross-sectional: neither (represent target population without regard to e or d initially).
