Study designs (epidemiology) Oct5 M2 Flashcards
2 types of study designs (broad categories)
descriptive vs analytical studies
3 types of descriptive studies
case series, cross-sectional studies, ecological studies
case series study def
present cases (exposures, outcomes), no control, low stat impact
ecological studies def
superpose 2 descriptive studies (disease and exposure), correlation but not causality, good for hypothesis generation
cross-sectional studies def
snapshot at one moment. prevalence.
2 types of analytic studies
observational and experimental
3 types of observational studies
cross-sectional, case-control, cohort
cross-sectional study design
take population, take data n exposure and disease, classify patients in positive negative disease and exposure groups
cross-sectional studies good and bad
good: helps for fixed exposure (age, gender)
bad: don’t know if exposure or outcome came first (bc snapshot in time)
case control study design
look back in time at defined pop. of cases, who was exposed. of controls, who was exposed
case-control study rare disease and exposure capabilities
can determine exposure in rare diseases
can’t deal with rare exposures
case-control study answers what question in lung cancer
in people with lung cancer, likelihood of smoking
cohort study: 2 designs and important point
retrospective (study after exposure and disease occur)
prospective (study before disease)
cohort study general design
take population, have expose and not exposed. look at who will get disease or not
cohort study: rare disease and exposure capabilities
can’t look at rare disease
can look at rare exposure
problem in cohort studies**
loss to follow up can create selection bias
experimental study: 1 type
RCT
RCT vs observational studies
produce same odds ratio on average
relative risk general deff
ratio of probability of disease in exposed group to non exposed group
relative risk includes what terms and word for all these terms
rate ratio, hazard ratio, risk ratio, odds ratio: MEASURES OF ASSOCIATION
rate ratio def
incidence rate (x per person year) in exposed div. incidence rate in non exposed
hazard ratio def
time to disease in exposed div. time to disease in non exposed
risk ratio def
cumulative incidence in exposed div cumulative incidence in non exposed
odds ratio def
odds of exposure in the cases vs odds of exposure in the controls
4 measures of impact
attributable risk, attributable fraction, population attributable risk, population attributable fraction
what we mean by relative risk
odds ratio
attributable risk def
diff in incidence RATES or in risks (cumulative incidence). how much disease we could prevent by removing exposure
attributable fraction def (2)
(incidence in exposed - incidence in unexposed) div. incidence in exposed
attributable risk div. incidence in exposed
population attributable risk def
incidence in all pop. - incidence in unexposed
population attributable fraction def
(incidence in all pop. - incidence in unexposed) div. incidence in all pop
how attributable fraction vs pop. attributable fraction differ generally
pop. attrib. fraction is lower bc you include some people without the exposure
how to calculate attributable fraction in case control study
(odds ratio - 1) div. odds ratio
how to calculate attributable risk in case control study
CAN’T CALCULATE IT. (need incidence but didn’t follow)
how to calculate the population attributable fraction in case control studies
AF (attrib. fraction) x prevalence of exposure in disease (of diseased, how many have been exposed)
how to calculate attributable fraction in cohort studies (2)
(incidence in exposed - inc. in unexp.) div inc. in exposed
(RR-1) div. RR
which observ study is better to study rare disease
case control
which observ study is better to study rare exposures
cohort study
which observ study is better to study multiple outcomes
cohort study (bc case control have to pick 1 outcome)
which observ study is better to study multiple exposures
case control study (bc cohort have to pick 1 exposure)
