Study Design Flashcards
goals of epidemiological studies
describe disease
identify associations
a good study is:
scientifically sound
valid
precise
efficient
descriptive study designs
not comparing groups
case reports
case series
cross-sectional descriptive studies
analytical study designs
groups are compared
experimental
observational
analytical study: experimental
clinical trials (treated/exposed)
analytical study: observational
compare groups of populations- ecological
compare groups of individuals- cross-sectional analytical studies, case-control, cohort (prospective and retrospective)
characteristics of descriptive studies
describe disease
no hypothesis is tested
no groups compared
characteristics of analytical studies
determine if there is an association and if so the strength
designed to test hypothesis
always compares groups
descriptive study: 1 subject
case report
descriptive study: a few subjects (6-12)
case series
descriptive study: lots of subjects (several dozen to hundreds)
descriptive cross sectional
disadvantages of case reports and case series
small number of cases
findings are not generalizable to the population
strictly descriptive
disadvantages of case reports and case series
small number of cases
findings are not generalizable to the population
strictly descriptive
descriptive cross-sectional study
sample of population
estimate the amount and distribution of disease
measure of disease occurrence is usually prevalence
steps of descriptive cross-sectional study
select subjects from the source population
measure the disease i each study subject
calculate the measure of disease occurrence
advantages of descriptive cross-sectional study
can generalized to the population
fast and cheap
provide good descriptive or baseline data for future
disadvantages of descriptive cross-sectional study
not good for causality
prevalence is of limited value
what is an association
when one changes so does the other
exposure and outcome are dependent on one another
outcome
a result or response, usually a disease or some other change in health status
exposure
potential determinant of disease or health status
T/F an exposure may increase, decrease, or have no effect at all
true
determinant (risk factor)
once an exposure is shown to be associated with disease
T/F analytical studies compare groups on the basis of either exposure or outcome
true
T/F controls are essential for measuring the effect of exposure
true
T/F controls are the reference group
true
how are analytical groups compared
exposure or outcome
observational studies
observes real life situations and draw inferences from them
not given a treatment or exposure
experimental studies
investigator manipulates subjects
subjects are selected and the allocated to receive a treatment or an exposure
selected
subjects selected because of their characteristics
allocated
assigned- experimental studies only
located to receive a treatment or exposure in order to study the effect
subjects selected because..
represent a population
exposure status
outcome status
randomization
process of making something random
each subject has an equal chance of being selected or allocated
sampling
subset of a population/group
goal is to select sample to represents population or group studies
T/F in experimental studies subjects randomly allocated to groups
true
experimental studies as also known as
randomized control trials
clinical trials
T/F experimental studies can be done in the lab or field
true
advantages of experimental studies
may establish causality
well-controlled studies are essential free of bias and confounding
statistically powerful
exposures and outcomes are clearly measured during the study
disadvantages of experimental studies
expensive and very narrow in scope
not always ethical
placebo effect
loss of follow-up can be higher for some treatments
T/F in observational studies are selected to participate in the study and then classified into groups
true
what are 2 broad types of observational designs
populations
individuals
ecological study designs
compares groups of populations
T/F no measurements are made on individuals-all measured at a population level
true
steps in ecology study
select a sample of populations
find population-level statistics on exposure and outcome for each population
compare groups of populations
T/F exposure and outcome are not measured at an individual level, but at the population level
true
advantages of ecological study design
can be done quickly and inexpensively
analysis and interpretation are relatively simple
can assess a wide range of exposure levels
disadvantages of ecological study design
ecologic fallacy
cannot detect subtle of complicated relationships
what is an ecologic fallacy
relationship observed at the population level may not hold true at the individual level
observational study on individuals
based on reason for selecting study subjects: represent a populations, have a certain outcome or exposure
analytical cross-section studies
individuals selected to represent a population
only selected one
T/F analytical cross-section studies measure both exposure and outcome (usually at the same time)
true
advantages of analytical cross-sectional studies
can generalize to population
fast and cheap
provide good descriptive or baseline data for future studies
disadvantages of analytical cross sectional studies
not good for causality
prevalence is limited (disease of long duration will have high prevalence even if incidence is low)
T/F diseases of short duration and high mortality will be under-represented in an analytical cross-sectional study
true
prevalence is limited in nature
case-control studies
individuals are selected to represent outcome
selected because of outcome status
T/F with case-control studies there is an association if the amount of exposure is different between the 2 outcome (disease) groups
true
number of subjects with the exposure are counted in each outcome and compared
selection in case-control studies
selection of cases
selection of controls
case definition
clearly describes and defines cases of the disease of interest
advantages of case control studies
short timeline
can look at multiple risk factors at one time
good for studying rare disease
disadvantages of case-control studies
poor selection of control can invalidate entire study
depends on accurate assessment of exposures that happened in the past
cohort studies
selected to represent exposure groups
used to study the effect of an exposure
what are they two types of cohort studies
prospective
retrospective
prospective cohort
forward in time
subjects selected based on exposure status and then followed over time to see if they develop the outcome
measure of association for prospective cohort
relative risk
retrospective cohort
back in time
selected based on exposure status, then outcome is determined from history, records, questionnaires, tests etc to determine if had/have outcome
measure of association for retrospective cohort
prevalence ratio (disease has already occurred)
advantages of prospective cohort studies
better to establish causality
can look at multiple exposures at one time
good for studying rare exposures
disadvantages of prospective cohort studies
can take a long time
have to deal with changing study populations
depends on accurate assessment of exposures and consistent assessment over time
advantages of retrospective cohort study
short timeline
can look at multiple exposures at one time
good for rare exposures
disadvantages of retrospective cohort studies
not as good for establishing causality
poor selection of controls can invalidate the entire study
depends upon accurate assessment of diseases that happened in past