Cohort Studies
what are the advantages of cohort studies?
- best way to establish causality -temporal relationship can be established (ie: exposure preceded the outcome) -many outcomes and exposures can be evaluated within a cohort -can investigate time to occurance of outcome -may be used to investigate rare exposures (ie: occupational cohorts)
what is the rate ratio equation? also called the incidence density ratio
Incidence density in exposed group / incidence density in non-exposed group **most appropriate for cohort studies where f/u is NOT complete for all subjects**
what is the relative risk equation?
The cumulative incidence (risk) in exposed group / cumulative incidence (risk) in non-exposed group ** most appropriate for cohort studies with complete f/u of cohort** An answer will usually come out to 1.1 or 2.o that means group A has a 2x greater chance of developing X than group B
define a cohort study
an OBSERVATIONAL study in which a group of people (cohort) who is initially FREE OF DISEASE (outcome) and followed over time to determine their outcome. Cohort members will be classified according to their exposure status (exposed or not exposed--NOT randomized) ie: nurses health study
what is a fixed cohort design?
cohort is defined at the start of the study, no new members allowed. some people may be lost to f/u so you can calculate the cumulative incidence if f/u is complete for all members OR incidence density is cohort members followed for different periods of time exp: nurses health study
what is ambi-directional cohort design?
data collected both retro and pro spectively on same cohort
define prospective cohort
investigator enrolls cohort and follows them forward as outcomes occur. OUTCOMES HAVE NOT YET OCCURED. this is more time consuming, more expensive. also called a longitudional study or concurrent cohort
what is a dynamic cohort?
new cohort members allowed into the study, each contributes the person-time they are observed. ie: a multi year study of the incidence of asthma in a school. New children allowed into the study. data analysis is based on person-time (incidence density)
define retrospective cohort
outcomes have already occurred when investigator starts the study. OUTCOMES HAVE OCCURRED when study starts -you rely on data already collected, done more quickly, less expensive also called historical cohort
what is a retrospective cohort?
same design as prospective except the investigator goes back to define the cohort at a certain point in time, determines exposure at that point then follows the cohort forward to see if the outcome occurs **this relies on using already collected data (medical charts, insurance claims)
what about the size of the cohort?
usually thought of as "mega studies" can have small cohorts though size of cohort depends on: frequency of outcome and how common the exposure is (rare conditions may need a smaller cohort)
can a cohort study be prospective?
yes- you could mail a questionnaire to ask someone if they took birth control for example and developed breast cancer later on
disadvantages of cohort studies?
-time consuming, may take years -expensive -loss to f/u -exposure info may not be complete if info is collected only at enrolment and not updated over time -not efficient for rare outcomes or diseases with long latency periods
what is the incidence density (or incidence rate) equation?
# of new cases or disease (or outcome) / total person-years at risk
what is the cumulative incidence (risk) equation?
# of new cases or disease (or outcome) / total population at risk
what is a closed cohort design?
cohort is defined at the start of the study, no new members allowed and NO losses to f/u exp: study of food poisoning after a picnic, uses cumulative incidence for analysis
