Module 5: Epidemiological Research Flashcards
Epidemiology
The study of the distribution and determinants (causes, risk factors) of health-related states and events in specified populations
What are the 2 types of Epidemiology?
- Descriptive
- Analytic
What is cross-sectional epidemiology research design?
Collects information at a SPECIFIC POINT IN TIME (versus over time)
Can be used to describe a population and its characteristics (including disease prevalence) at a SPECIFIC POINT IN TIME
Define Prevalence.
Refers to the proportion of persons who HAVE a condition at or during a specific time period
Define Incidence.
Refers to the proportion or rate of persons WHO DEVELOP the condition during a particular time period
When analyzing cross-sectional data, what are the 2 options?
- You can split up the groups initially, OR
- You can look at them after you collect your data
2x2 Contingency Tables
Used to interpret Epidemiological data to determine RISK.
It has a very specific format, the disease or the variable of interest is listed across the top, with the exposure listed on the left side column. You then have the numbers for each combination plugged into the associated cells below. Then you can calculate the prevalence ratio (which is an estimate of the magnitude of the association between the exposure and the disease).
Cohort
A large group of individuals who share a common characteristic and are OBSERVED OVER TIME to gather information about exposures, that can affect an outcome
What the 3 Types of Research Designs in Epidemiology?
- Cross-Sectional
- Cohort (prospective or retrospective)
- Case-Control
(TRUE/FALSE)
Cohort studies have an intervention and are randomly assigned.
FALSE
There is no intervention or randomization in cohort studies
(TRUE/FALSE)
Cohort studies gives incidence of a disease.
TRUE
What are 2 limitations of a Cohort study?
- More expensive
- More time-consuming
Compared to cross-sectional studies because you now have to hire enough people to follow these cohort individuals over a period of time
What is this type of study?
Identifying the population, recruiting, and enrolling participants, getting baseline measurements, then following them over time to evaluate exposures and disease incidence.
Prospective Cohort study
What is this type of study?
Identifying the population prior to developing the disease or outcome, then collecting health information and exposure incidences collected in the past
Retrospective Cohort Study
Define Risk Ratio (Relative Risk).
Another way to interpret data, especially in Epidemiology
RR tells who is likely to develop the disease in an exposed/treated group versus a non-exposed group, BASED ON INCIDENCE DATA
**WHAT WOULD AN RR OF 1.0 MEAN?
There is NO difference between groups
**WHAT WOULD AN RR > 1.0 MEAN?
RR GREATER THAN 1.0
There is an INCREASED risk in the exposed group.
**WHAT WOULD AN RR < 1.0 MEAN?
RR LESS THAN 1.0
There is a DECREASED risk in the exposed group
How do you interpret an RR of 3.0? **
It would mean the exposed group would be 3x more likely to have the outcome, OR by 200% more than non-exposed group
(RR-1) x 100
How would you interpret an RR of 0.25? **
It would mean that the exposed group would be 75% LESS risk of having the outcome
(0.25-1) x 100
(TRUE/FALSE)
When interpreting the RR, the next step is to look at the confidence interval.
TRUE
What is the Confidence Interval?
95% of data points would fall within this range
What are the 2 considerations when interpreting CI?
- NARROWER IS BETTER (so CI of 0.2 to 0.4 would be stronger than 0.2 to 0.7)
- IF IT CONTAINS 1.0, THEN THE RR IS NOT LIKELY TO BE STATISTICALLY SIGNIFICANT AND MAY SIMPLY BE A MATTER OF CHANCE.
(A 1.0 means there is no difference between the exposed and the non-exposed)
Explain Case-Control as a Epidemiological research design.
Start with the outcome of interest (disease); make groups of with (CASE), or without (CONTROL) outcome; then use retrospective data on each to identify exposures to determine ODDS RATIOs
What is Odds Ratios (OR)?
The odds that a case has experience a risk/exposure versus the odds that a control has experienced that same risk/exposure
Does not show causality, but helps identify possible causal factors
OR = 1.0
no difference b/t cases and controls
OR > 1.0
Increased frequency of exposure among cases and exposure is POSITIVELY RELATED TO A DISEASE
OR < 1.0
Exposure is NEGATIVELY RELATED TO THE DISEASE, meaning exposure is protective against the disease
