Module 4 Flashcards
module 4 Rothenberg (measures of association) and lecture 4
What type of study involves looking at the present?
cross sectional
What are the 3 basic structures of study design?
start with diseased / non-diseased and determine exposure
go to exposed / non-exposed and follow for disease
ask about exposure and disease at the same time
What type of study involves retrospective analysis?
Surveillance
What type of study involves prospective analysis?
cohort study
What are longitudinal studies?
repeated queries of a population
What are the two types of longitudinal studies?
repeated studies of the same people (cohort or panel)
repeated studies, but with different people (trend study)
What is a retrospective cohort?
Look into previous studies for exposure and consequent disease, then follow up in a present study
What is an experimental study?
The investigator provides the exposure (usually in some form of treatment).
Start with a group who has a given disease and give some treatment A, some treatment B.
What is a “cross-over design” in research?
each subject acts as his / her own control
(e.g. subjects use a certain drug, then stop for a “wash out period,” then try a different drug).
What is incidence versus relative risk?
We observe incidence when we see who of the exposed and unexposed contract disease.
When we compare incidence in exposed to incidence in unexposed, we determine relative risk.
Describe the formula for Relative Risk (in terms of a two by two table)
How do we calculate relative risk in cohort and case control studies, respectively?
For cohort, we estimate incidence and ratio of incidence directly
For case control, we estimate relative risk with the Odds Ratio.
In its most generic form, what is Relative Risk?
Incidence of Exposed / Incidence in unexposed
It is just a number, without “dimension” (or unit)
What is difference of incidences?
Incidence in exposed - incidence in unexposed (cases / person-time)
Retains the dimension / unit
What was Doll and Peto’s approach to attribution research?
obtaining a crude estimate of “avoidable cancer” by comparing incidence in a community and comparing it to a community with the lowest, reliably reported incidence
e.g. lung cancer in Connecticut versus NIieria (little tobacco use in the 70s)
How did Levin differ from Doll and Peto?
he focused on incidence in the exposed rather than incidence in the unexposed. His works led to the development of the odds ratio.
Describe the derivation of attributable risk.
It is the total incidence minus incidence in the unexposed.
How do we refer to groups of participants (regarding Attributable Risk) in a case control study?
Proportion of exposed cases are P(cases) (instead of P1)
Since the controls have also been exposed (but receive no / different treatment), they are referred to as P(controls) instead of P(0)
What is the importance of multi variable PAR (Population Attributable Risk?)
The total percentage never exceeds 100%. Rather than separating exposures, it recognizes that risks interact and recognizes that each combination of exposures has its own Attributable Risk and its own proportion in the population.
Describe the Prevented Fraction Formula
The proportion of disease prevented in those exposed to the factor or intervention. (PF = 1 - RR)
I(0) is the group that was NOT exposed to preventative (so their infection rate will be higher). I(t) is total incidence in the population (whether exposed to preventative or not).
What are some important factors of Attributable Risk?
If X% of a disease is attributable to a factor, removing that factor does not mean it remove X% of the disease. (It measures the proportion of disease that may not have occurred if there were no exposure)
Prevalence and relative risk are population specific.
Describe the formula for the ratio of odds (cross product ratio)
