cohort: group of persons sharing a common
characteristic (geogr, occup, sex, race,etc).
1. SOURCE POPULATION:se sacan entire or
random sample, se descartan enfermos YA con
outcome (exposure always preceed outcome) but
sometimes disease IS the exposure y se evaluan
muertes
2.FOLLOW OVER TIME
for new events,
calculating eg. risk v rate
useful for studying uncommon or rare exposures
measure incident cases -- can be PROSPECTIVE OR RETROSP., OPEN or CLOSE
CASE CONTROL STUDIES
for studying rare diseases(<10%).
Is INDUCTIVE (from effect to cause,
RETROSPECTIVE)--can really MIRROR a cohort
1. Define&select CASES (sick people)
2. Define&select CONTROLS (nonsick): restricted
source population from which cases arise
MATCHING: for every case,
one control has the same
variable--should be only
STRONG factors
3. Measure&compare EXPOSURE PREVALENCE
CANNOT use rate&risk, ONLY ODDS (the control is not representative of the population