1What is an epidemiological study?
A study used to obtain estimates of epidemiologic measures, such as associations between exposures and health outcomes.
2What is the major difference between observational and experimental studies?
In an observational study, the investigator does NOT assign exposure; in an experimental study, the investigator DOES assign the exposure or treatment.
3Observational study
The investigator observes participants and records their exposures and outcomes without assigning the exposure.
4Experimental study
The investigator intervenes and assigns an exposure or treatment to subjects.
5When is an experimental study appropriate?
When assigning the exposure is ethical and feasible.
6What is a randomized controlled trial (RCT)?
An experimental study in which subjects are randomly assigned to treatment/exposure groups to estimate the effect on an outcome.
7Why is a control group important in an experimental study?
It provides a comparison group so investigators can determine whether the treatment or exposure affects the outcome.
8Randomized Control Trial
An experiment that attempts to answer a medical question with the goal to extrapolate to the general population
9What is blinding or masking?
Keeping study personnel or participants unaware of treatment-group assignment to reduce bias.
10What is confounding?
An outside variable that influences both the exposure and outcome, creating a false or distorted association.
CONFUSING
11What does a confounder do to an exposure-outcome relationship?
It can make the exposure appear associated with the outcome when some or all of the association is actually due to another variable.
12What is the purpose of randomization?
To make treatment groups more comparable and help prevent confounding.
13Does randomization guarantee that groups are identical?
No. Its effectiveness depends partly on sample size, so investigators should still compare baseline characteristics and potential confounders.
14Why does a larger sample size help randomization?
It increases the likelihood that potential confounders are distributed similarly between groups.
15What is a cross-over study?
An experimental design where the same subjects are studied under both exposed/treatment and unexposed/control conditions.
16What is a major advantage of a cross-over study?
Each subject serves as its own control.
17What important assumption must be met in a cross-over study?
There should be no carry-over effect from the previous treatment, or an adequate washout period must be used.
18What are the major observational study types?
Cohort, case-control, cross-sectional, ecological, and diagnostic assay evaluation.
19Why is confounding especially important in observational studies?
There is no randomization, so exposed and unexposed groups may differ in other variables besides the exposure.
20What is a cohort study?
A study that begins with disease-free subjects, groups them by exposure status, and follows them to determine who develops the outcome. HOLY GRAIL
21What direction does a cohort study follow?
Exposure → outcome.
22What do you compare when analyzing a cohort study?
The incidence or risk of the outcome in exposed versus unexposed subjects.
23When is a cohort study especially useful?
When studying a rare EXPOSURE.
24Prospective cohort study
Subjects are selected and then followed forward in time to determine whether the outcome occurs.
25Retrospective cohort study
The investigator uses existing records to identify past exposures and outcomes.
26What is a case-control study?
An observational study that starts with subjects based on OUTCOME status, cases versus controls, and then compares their previous exposures.
27What direction does a case-control study work?
Outcome → look back at exposure.
28When is a case-control study especially useful?
When studying a rare DISEASE.
29What are cases in a case-control study?
Subjects that have the disease or outcome of interest.
30What are controls in a case-control study?
Subjects without the disease/outcome that represent the source population from which the cases arose.
31How should controls be selected in a case-control study?
Independent of exposure status.
32Cohort vs. case-control high-yield rule
Cohort starts with EXPOSURE and looks for OUTCOME; case-control starts with OUTCOME and looks for EXPOSURE.
33Rare exposure vs. rare disease rule
Cohort = useful for rare EXPOSURE; case-control = useful for rare DISEASE.
34What is a cross-sectional study?
A study that examines a population or random sample at a particular point/period without selecting subjects based on exposure or disease.
35What does a cross-sectional study commonly measure?
Prevalence.
36What is a major limitation of a cross-sectional study?
Temporality is difficult to establish, so it may be unclear whether the exposure occurred before the outcome.
37Adult dairy cows are tested once for antibodies against Johne’s disease to estimate how common infection is on the dairy. What study design is this?
Cross-sectional study.
38What is an ecological study?
An observational study in which exposure and/or outcome are measured at the GROUP level, such as countries, states, counties, or classes.
39What is the ecological fallacy/ecological bias?
Assuming that an association observed at the group level also applies to individuals within those groups.
40Countries with greater chocolate consumption have more Nobel laureates. Can you conclude that individuals who eat more chocolate are more likely to win a Nobel Prize?
No. That would be an ecological fallacy because a group-level association does not necessarily represent an individual-level association.
41Fast exam clue: investigator assigns treatment
Experimental study.
42Fast exam clue: start disease-free, separate by exposure, follow for outcome
Cohort study.
43Fast exam clue: start with diseased vs. non-diseased and compare past exposure
Case-control study.
44Fast exam clue: measure disease prevalence at one point/period
Cross-sectional study.
45Fast exam clue: compare exposure/outcomes between countries, states, counties, or other groups
Ecological study.