ch 8 - cohort study Flashcards
what is a cohort study?
- subjects are defined according to their exposure levels and followed for disease occurrence
- typically examines multiple exposures in relation to a disease; subjects are defined based on their disease status, and exposure histories are compared.
- the observational equivalent of experimental studies but researchers do not allocate exposure but locate natural experiments to observe the relationship between exposure and disease
what are the three types of populations studied? what calculation is used to find them?
- open dynamic- changeable - IR
- fixed- loss may occur - IR
- closed - no loses - ci
what are exposed populations based on? what are the two types?
- depends on hypothesis under study, exposure frequency, feasibility
- general and special cohort
general cohort
- investigates the exposures that are common with a population - alochol use
- researchers can study the effects of multiple E on one or more outcomes
- selected by professional groups or residents from a well-defined grographic area
- ex: framingham study
special cohort
- investigates unsual and rare expsoures
- researchers can study the effects of a SINGLE exposure on one or more outcomes
- must go to where the expsoure has occured , will occcur
counterfactual ideal
- the concept referring to what would have happened to an individual in the absence of the exposure
- The counterfactual ideal is hypothetical since, in reality, the exposed and unexposed study participants cannot exist simultaneously.
- The rationale of behind the counterfactual ideal is the notion that the comparison group and the exposed group are exactly the same except for the exposure
dose-response relationship
risk of disease increases as intensity/duration of expsoure increases
how do epidemiologist select sets of poeple?
- as similar as possible with respect to:
other factors that can influence outcome
to collection of comparable and accurate information
what are the three types of comparison groups?
- internal- unexposed members of the same cohort
- general pop- based on preexisting pop data on disease incidence and mortality
- comparison - consists of members from another cohort
Internal comparison group? strength? weakness?
- are comprised of unexposed members of the same cohort
- is most comparable to the exposed group
- difficult to identify
general pop? strengths and weakness
- maybe used as a comparison group when it is not possible to identify a comparable internal comparison group. and occurs in occupational cohort studies
- Accessible/stable data
- weakness:
lack of comparison with the exposure group
results may suffer from healthy worker effect
data on key variables may be missing - herbicide factory works vs general pop
what is healthy worker effect?
why is it an issue?
- in occupational cohort studies the rates of morbidity and mortality among working population is lower than those in general population
- since its a health requirement for workers to be healthy
- issues is that it creates an underestimation of teh risk when comparing a worker to general popualtion
comparison group?
strength
- weakness
- example?
- comprised of unexposed study participants from another population
- fairly comparable
- results can be difficult to interpret
because comparison cohort often has other exposure - comparing landscapers with construction workers
strengths and weakness of compairson groups
- fairly comparable
- results often difficult to interpret because a comparison cohort has other E
what are the three sources of exposure information?
- preexisting records
- questionnaire , interview
- direct testing
how do we measure frequency and assoiation?
2x2 table (not realistic, in realiity we use statistical package)
CI or IR of E
RD or RR compairing E
what are strength of cohort studies?
- clear temporal sequence between E and D especially if prospective
- efficient for rare E
- good information on E, confounders, especially if prospective
- can study the effect of an E on multiple outcomes
- less vunerable to bias
what are the limitations of cohort studies?
- many need a large number of subjects to be followed for a long time
- expsenive and time consuming
- not good for rare disease or those with long latecny if prospective
- LTF can undermine validity
what is the goal of a cohort?
- obtain a complete follow up info on all subjects regardless of exposure status
- very time consuming porcess
what is LTF?
- loss to follow up - raises doubts about the validity of a study
- high LTF = bias
- decrease in sample size
- no magical number for it to be okay but preferred to be >80%
what is a prospective cohort study
- participants are grouped on teh basis of past or current exposure and followed into the future yo observe the outcome of interests
- looking forward in time
prospective cohort study limitations and strengths
- limitations
expensive and time consuming
not efficient for diseases with long latent period - strengths
better exposure and confounder data (real time collection of data)
less venerable to bias (since outcome has not happened yet)
what is a retrospective study?
- exposure and outcome has already occurred when the study began; studies past E and outcomes
- looking back on time and starting where no one has a disease in the beginning
what are the strengths and limitations of a retrospective study?
- strengths
cheaper and faster
efficient with disease with long latent period - limitation
exposure and confounder data may be inadequate
more venerable to bias
what is ambidirectional?
- both componets; goes back in time to assess the exposure of the pop (everyone will disease free)
- followed through current and future
Describe in one sentence what the key feature is that differentiates a Randomized Clinical Trial (otherwise
known as an Experimental Study) from a Cohort Study
Cohort studies are the observational equivalent to experimental studies with the exception that the researcher
does not allocate “exposure” (does not randomize exposure status). In experimental studies, the research
assigns exposure to the study participants.
Prospective cohort
study group?
Absolute measures of comparison?
Relative measures of comparison?
Temporal relationship between E and D?
Multiple associations?
Time required for study?
- exposed persons
- unexposed persons
- RD/AR PRD/PAR APe/AR% APt/PAR%
- relative risks and Disease OR
- easy to establish
- possible to study association of an E with multiple diseases
- generally long because of the need to follow the subjects
prospective study
- the cost of the study?
- estimate differences between groups studied?
- Is population size required?
- best when (E is rare or not/D is rare or not ) ?
- problems?
- potential bias?
- expensive
- medium to large
- relatively large
- exposure is rare and disease is frequent among exposed
- selection fo nonexposed comparison group is often difficult and changes over time in criteria and methods
- assessment of outcome
