Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
82 Cards in this Set
- Front
- Back
Purpose of Clinical Trials:
|
To study the effectiveness of treatments in terms of causal relationships
|
|
Purpose of Cohort and Case Control studies:
|
To look at associations between risk factors and disease
|
|
Purpose of Diagnostic Testing studies:
|
To estimate accuracies and predictive values of diagnostic tests
|
|
What is an "independent variable"?
|
A CAUSE on which the EFFECT (dependent variable) is dependent.
|
|
In Evidence-Based-Medicine, what are we looking for evidence OF?
|
Causality - that a variable causes a changed effect.
|
|
Why aren't Koch's Postulates very good for determining causality?
|
Because they are so stringent that they don't allow for the multiple factors that can be associated with an effect.
|
|
What is the requirement of Temporality to provide evidence of causality?
|
The cause must precede the effect.
|
|
What is the requirement of Association to provide evidence of causality?
|
The cause and effect must be seen together more than expected by chance.
|
|
What other alternative explanations have to be considered?
|
Confounders
|
|
What is a Dose-Response pattern?
|
A relationship between cause and effect where the greater the cause, the greater effect.
|
|
8 Factors of Evidence for Causality:
|
1. Temporality
2. Strength of association 3. Dose Response 4. Consideration of alternative explanations 5. Consistency across studies 6. Biologic Plausability 7. Cessation of exposure 8. Specificity of association |
|
2 Main categories of study design:
|
1. Experimental
2. Observational |
|
How is an experimental study conducted?
|
Something is done to a group of patients and the impact is measured in comparison to a control group.
|
|
What are experimental studies generally referred to as?
|
Randomized controlled trials
|
|
How are observational studies conducted?
|
The researcher does not control what is done to the subjects, associations are just measured.
|
|
4 types of Observational Studies:
|
1. Ecologic
2. Cross-sectional 3. Cohort 4. Case Control |
|
What is measured in an Ecologic study?
|
Associations between groups
|
|
Unit of analysis in an ecologic study:
|
Group
|
|
Disadvantage of ecologic studies:
|
Few conclusions can be made, and they are generally very poor and circumstantial.
|
|
Advantage of ecologic studies:
|
Allows you to look at larger geographic trends.
|
|
How is a Cross-sectional Study designed?
|
By starting with a sample and looking at what proportion is exposed, and what proportion is diseased.
|
|
What does a Cross-sectional study measure?
|
Prevalence
|
|
What is Prevalence?
|
The # of cases/Total population at a GIVEN POINT IN TIME.
|
|
What can/cannot be done with a Cross-sectional study?
|
CAN: build evidence for associations
CANNOT: establish causality |
|
Why can't a cross-sectional study establish causality?
|
Because a CS study is weak
|
|
How is Measure of Association calculated from a cross-sectional study?
|
Prev of disease in exposed
-------------------------- Prev of disease in unexposed |
|
For asthma, what would a prevalence ratio of 3 for heavy smokers compared to nonsmokers tell you?
|
That heavy smokers are 3 times more likely to have asthma than nonsmokers.
|
|
What is an example of a cross-sectional study?
|
A survey
|
|
Advantages of cross-sectional studies:
|
Quick, easy, and cheap
|
|
What does a Cohort Study analyse?
|
Risk and prognosis
|
|
How does a Cohort study start?
|
By chosing a group of patients based on clear criteria that are free from the outcome at the start.
|
|
After choosing a study population, how does a cohort study proceed?
|
By measuring the exposure or risk factor of interest.
|
|
How is it decided who gets exposed or not?
|
It isn't decided; the groups form naturally.
|
|
What happens after groups are differentiated into exposed vs nonexposed?
|
The researchers look forward in time and measure the incidence of the outcome in each group.
|
|
What is a key factor of evidence ofr causality that is demonstrated by cohort studies?
|
Temporality - because the investigators look into the future to look for the outcome.
|
|
What is a confounder?
|
A factor that is associated with (and effects) both variables but has not been accounted for.
|
|
2 ways that confounders can affect study results:
|
1. Cause spurious/false associations
2. Influence the size of the effect |
|
What is the confounder in a cohort study that shows heavy drinkers are more likely to develop lung cancer?
|
Smoking - because heavy drinkers are more likely to be smokers.
|
|
What is the measure of outcome in a cohort study?
|
Incidence
|
|
What is Incidence?
|
The proportion of an at-risk population that DEVELOPS the factor of interest over a certain period of time.
|
|
What is Prevalence?
|
The proportion of an at-risk population that HAS the factor of interest at ANY GIVEN TIME.
|
|
For the flu, what is higher, incidence or prevalence?
|
Incidence - more people develop the flu over one year than have it at one given moment.
|
|
Why is incidence higher than prevalence for the flu?
|
Because it is such a short-lived illness.
|
|
What types of diseases have a higher prevalence than incidence?
|
Chronic, long-lasting diseases. Because the illness lasts so long that patients accumulate.
|
|
What is the measure of association in a Cohort study?
|
Relative Risk, aka Risk Ratio
|
|
Formula for Relative Risk:
|
Incidence of Outcome in Exposed
------------------------------ Incidence of Outcome in Unexposed |
|
What does a RR=1 say?
|
There is no difference in exposed vs unexposed, hence no association with the outcome.
|
|
What does a RR>1 say?
|
Exposure causes the outcome.
|
|
What does a RR<1 say?
|
Exposure prevents the outcome.
|
|
Disadvantages of Cohort studies:
|
-Can be expensive
-Can be time-consuming -Difficult to study the incidence of a rare disease |
|
Why is it difficult to study the incidence of rare diseases?
|
Because you need such a huge sample population.
|
|
How do we practically study associations for rare occurrences?
|
Via Case-control studies.
|
|
What does a Case-control study start with?
|
Selecting 2 groups of patients:
1. Cases - already have the disease or outcome 2. Controls - don't have the disease or outcome |
|
What do you want to be the same between cases and controls?
|
The population from which they came.
|
|
What happens after sample selection in a case-control study?
|
You look retrospectively to obtain information on risk factors, then see if there is an association with the outcome.
|
|
3 Advantages of a Case-control study compared to cohort:
|
-Quicker
-Cheaper -Require no long follow-up period |
|
3 Disadvantages of a Case-control study:
|
1. Can't establish temporality bc it's retrospective
2. Still have issues of confounding variables 3. Selection bias |
|
What is the measure of association for a cohort study?
|
Incidence
|
|
What is the measure of association for a case-control study?
|
Odds ratio
|
|
Why can't we calculate incidence for a case-control study?
|
Because we START with the disease, so we can't determine what the risk of getting the disease is.
|
|
What do we do with the incidence calculated from a cohort study?
|
Calculate a Risk Ratio
|
|
What do we do with the odds that we calculate in a case-control study?
|
Calculate an Odds Ratio
|
|
How do you calculate Odds for a case-control study?
|
# with factor
------------- # w/out factor |
|
What is the probability of rolling a six with a die?
|
1/6 = 0.17
|
|
What are the odds of rolling a six with a die?
|
1/5 = .20
|
|
How does the Odds Ratio work as a measurement of association?
|
If the outcome is rare, the OR approaches the relative risk and is a good estimate of it.
|
|
A case-control study is:
|
always retrospective; the history of exposure is measured after the outcome is known.
|
|
Comparison groups in a case-control study are selected based on:
|
outcome
|
|
Case-control studies are useful when:
|
the outcome is rare
|
|
The measure of association for a case-control study is:
|
the Odds Ratio
|
|
What is the gold standard of study design?
|
Randomized Controlled Trials
|
|
How does an RCT start?
|
By selecting a group of subjects that meet clear inclusion criteria
|
|
What happens in an RCT after the study population is selected?
|
Allocation into 2 groups:
1. Experiment/intervention 2. Control |
|
What is one of the strengths of the RCT?
|
Temporality can be established because you measure the outcome after causing the cause.
|
|
Why is it advantageous to "control" who gets the experimental intervention or not?
|
Because you can control alternative variables and keep them constant.
|
|
What is the measure of association in an RCT?
|
Incidence - calculated as relative risk.
|
|
How is Relative Risk calculated from an RCT?
|
Incidence of outcome in treated
------------------------------- Incidence of outcome in control |
|
What is meant by:
RR=1 RR>1 RR<1 |
RR=1 means no difference
RR>1 means treatment causes the outcome RR<1 means treatment prevents the outcome |
|
What stipulation must be met in order to use the relative risk ratio?
|
The outcome must be dichotomous - yes/no, better/worse, etc.
|
|
The RCT is the gold standard for:
|
Studies about treatment
|
|
3 Disadvantages of the RCT:
|
1. Expensive
2. Time-consuming 3. May not be ethically feasible |
|
The RCT measures
|
Incidence/Risk ratio
|