Epidemiology Flashcards
Which study design is the most scientifically sound and the best measure of exposure?
Clinical trial
Which study design is the most accurate observation study and a good measure of exposure?
Cohort study
Which study design can study rare diseases and is relatively less expensive/fast?
Case-control
Which study design is the fastest and least expensive and measures multiple exposures and outcomes?
Cross-sectional
Which study design is time consuming and the most expensive?
Cohort study
Which study design causes possible time-order confusion and possible error recalling exposures?
Case-control
Which study design causes possible time-order confusion, the least confidence in findings, and measurement bias?
Cross-sectional
Which study design is time consuming, not generalizable, unethical for harmful exposures, and most expensive?
Clinical trial
What happens to incidence and prevalence if a new effective treatment is initiated?
Incidence stays the same, prevalence decreases
What happens to incidence and prevalence if a new effective vaccine gains widespread use?
Incidence decreases, prevalence decreases
What happens to incidence and prevalence if number of persons dying from the conditions decreases?
Incidence stays the same, prevalence increases
Case fatality rate equation
Number of deaths due to disease/number of cases of a disease
Proportionate mortality rate equation
Number of deaths due to disease/total deaths in the population
Attributable risk equation
Incidence rate among exposed - incidence rate among unexposed
Relative risk equation
Incidence rate among treated / incidence rate among non-treated
Attributable risk percent equation
AR / incidence rate among treated
Number needed to treat
1 / attributable risk
Relative risk equation
[a/(a+b)] / [c/(c+d)]
Odds ratio equation
ad/bc
T/F: For a rare disease, odds ratio is a good measure of the relative risk.
True
Interpret a RR of 5.0
50% chance of developing a disease if exposed
P value of 0.05. What does this mean?
There is a 5% chance that the results were just do to chance
Limitations of p-values
dependent on sample size
larger effect sizes more likely to produce significant sample size
only assess the null hypothesis, not the alternative hypothesis
tells us about statistical significance but nothing about magnitude of effect
T/F: P-values more likely to be statistically significant for larger effects
True.
Reject null hypothesis when it is, in fact, true.
Type I error. Convicting someone who is innocent.
Failing to reject the null hypothesis, when in fact, it is false.
Type II error. Acquit someone when they are guilty.
When odds ratio overestimates relative risk, what type of error is this?
Type I
The probability that a person with a positive result actually has the disease
Positive predicted value
The probability that a person with a negative result actually does NOT have the disease
Negative predicted value
Equation for positive predicted value
PPV = True positives/(True positives + False positives)
Equation for negative predicted value
NPV = True negatives/(TN+FN)
If the prevalence increases, does PPV increase or decrease? What about NPV?
It increases! And Negative predicted value decreases…
What is selection bias?
Choosing participants for a study who are most qualified to provide the results you are looking for.
Is the level of exposure in the controls the level expected in the population?
What is confounding by indication?
A bias frequently encountered in studies of drug effects. Because the allocation of treatment is not randomized and the indication for treatment may be related to the risk of future health outcomes, there are different risks for treated vs untreated groups, which can generate biased results
Ex: drug given to highly selected individuals who needed the drug because of their poor prognosis
What causes recall bias?
caused by differences in the accuracy or completeness of the recollections retrieved (“recalled”) by study participants regarding events or experiences from the past. Ex: case vs control mothers may recall differences in exposures.
How is a confounding factor similar to “guilt by association.”
If you fail to account for a confounding factor, you may develop a perceived relationship between an independent variable and a dependent variable that has been misestimated
A logical fallacy in the interpretation of statistical data where inferences about the nature of individuals are deduced from inference for the group to which those individuals belong
Ecological falacy
Interpretation and limitations of P-value: interpretation
Interpretation: % of random variability; limitation: large sample size gives skewed p-value
(blank) (exposure before disease) is a strength of clinical trials and cohort studies
Temporality
