research study designs Flashcards
study done by CDC when disease outbreak
cases: have disease
control: no disease
* retrospective: look at recent food intake, prior environmental exposures, travel hx to determine how groups are similar/different → identify exposure/risk factor linked to disease
* observational: no intervention by investigator, experimental arm
goal: starts with DISEASED population and look for prior risk factors*/exposures for a disease (identify an association between risk factor and disease - not able to say causal)
case-control study
outcome of case-control study yields
odds ratio
group of people with common exposure/outcome compared to group of people without exposure/outcome
- observational: follow both cohorts and compare the outcomes among the exposed and non-exposed, no intervention
- retrospective: form cohorts based on exposures in past and determine if outcome ever developed in past
- prospective: follow cohorts with recent exposure into future to determine future outcomes
goal: starts with EXPOSURE and see if exposure ↑ likelihood of outcome
cohort study
outcome of a cohort study yields a
relative risk ratio
*observational: collect data from a population at one point in time, identify number of people with disease and/or risk factors/exposures for the same point in time
*NOT retrospective or prospective: one point in time
goal: prevalence: estimate frequency of disease and known risk factors in a population at a particular point in time
doesn’t establish causality = how exposure and disease outcome are related over time
cross-sectional study
outcome of cross-sectional study
prevalence: frequency of disease and known risk factors in a population at a point in time
*observational: compare concordance rates of diseases/traits in identical twins/fraternal twins to measure heritability of disease/traits
twin concordance study similar to adoption study (twins raised in separate homes: study impact of genetics vs environmental factors “nature vs nurture”)
*interventional: experimental study where investigator is intervening to draw a conclusion on HUMANS!
one group given experimental therapy
other group given comparison treatment or placebo
clinical trial
at least 2 study arms:
experimental group receives experimental treatment and
control group receives placebo or comparator drug
crossover study: patient serves as own control (limit selection bias and limit amount of confounding)
controlled clinical trial
random assignment to either group: experimental or control
limits selection bias + confounding (2 groups have similar characteristics)
randomized clinical trial
participant or investigator don’t know whether the participant is receiving experimental treatment or control drug
expectation of investigator (pygmalian effect) and participant won’t effect the evaluation of the study intervention
double-blinded clinical trial
phases of drug trials
phase 1: is the drug safe?
phase 2: does the drug work?
phase 3: does the drug work better?
phase 4: postmarketing surveillance
give drug to healthy volunteers to determine if drug is safe (toxicities) + pharmacokinetics
phase 1
give drug to few patients with disease to measure efficacy, dosing, side effects
phase 2
compare experimental drug to drug on market or placebo: randomly assign to experimental drug or comparator/placebo - does it work better?
phase 3
