Can I get the definition please?
The study of the distribution and determinants of health-related events in specified populations and the application of this study to control health problems.
Epidemiology
One way to increase this, which leads to narrower confidence intervals, is by increasing sample size
Precision
A study finds that women who consume alcohol heavily have a 1.5 times higher risk of breast cancer compared to women who do not drink (RR: 1.5; 95% CI: 1.25-1.67; p=0.0006). This result is considered what with regards to statistical significance?
Statistically significant
Bonus points: what do we typically set our p-value cut-off at?
A study in which children are randomly assigned to receive either a newly formulated vaccine or the currently available vaccine and followed to monitor for side effects and effectiveness of each vaccine, is an example of this type of study.
Randomized controlled trial
Bonus: If some of the children randomized to the currently available vaccine end up receiving the new formulation and some in the new formulation group don't end up receiving the vaccine, what different ways can you analyze the data?
A population-based cross-sectional survey in 2014 of 9,813 U.S. adults found that 3,905 individuals reported a BMI >=30. What is the prevalence of obesity (BMI >= 30) among those who took the survey?
Prevalence=3905/9813=39.8%
Bonus points: Is this a point prevalence or period prevalence estimate?
Studies that aim to characterize patterns of disease occurrence in terms of person, place, or time are considered descriptive studies. These types of epidemiologic studies aim to identify potential causes of disease.
Analytic
P-values are influenced by the magnitude of the association (stronger association -> smaller p-value). This also influences p-values.
Sample size
P-values and confidence intervals are used for testing whether the results of a study are consistent with this.
The null hypothesis
Note: this is called hypothesis testingThis study design is considered the strongest among observational studies for determining the evidence of a causal relationship between an exposure and disease.
Cohort studies
A cohort of 10,000 60 year-old women includes 5,000 who are regular exercisers and 5,000 who do not exercise regularly at the start of the study. Over the course of 10 years, 632 of the women who don't exercise regularly died of cardiovascular disease. Calculate the measure of disease frequency for this study among those who are not regular exercisers.
Risk: 632/5000=0.1264
Diseases that are ongoing at a usual level within a given population or geographic area are considered endemic. When the occurrence of disease rises to higher than normal levels, it is considered this.
An epidemic
Sample size determination is based on type I error, study power (aka 1-type 2 error), and this.
Effect size you wish to detect (magnitude of difference between your exposed and unexposed groups).
Assuming a p-value (or alpha) of 0.05, a study presents an association between an exposure and outcome with a p-value >=0.05. This value will be included in the corresponding 95% confidence interval.
Null value (1 or 0)
Bonus points: When is the null 1 and when is it 0?
This type of study design is efficient for studying rare diseases and multiple exposures.
Case-control studies
Bonus: What does a disease have to be in both the exposed and unexposed for an odds ratio to estimate a risk ratio?
150 children from Flint, Michigan with known exposure to high levels of lead in their water supply and 150 children from Atlanta, GA without exposure to water with high lead levels are followed for 10 years to assess the impact of exposure to high lead levels in water and academic performance. 125 children from Flint, MI had no drop in IQ levels. 9 children from Atlanta, GA had a drop in IQ levels. Calculate the appropriate measure of association for this study and interpret it.
Risk ratio=(25/150)/(9/150)=2.78
The risk of decline in IQ over ten years was 2.78 times higher among children exposed to water with high lead levels compared to children not exposed to high lead levels.
Often used to determine statistical significance in a study, this is the probability that an association at least as strong as observed might have arisen by chance alone assuming the H0 is true.
P-value
This is the ability of a study to detect a true difference between two groups.
Power
Bonus points: What do we typically set this at?
P-values and 95% confidence intervals both provide information about statistical significance. Confidence intervals additionally provide information about this by how wide they are.
Precision
For this type of study, the exposure and outcome are ascertained simultaneously.
Cross-sectional study
Prevalence is a measurement of the existing cases of disease at a particular moment or period in time. This is a measurement of the new cases of disease that occur during a specified period of time.
Incidence
This type of error occurs when one concludes that there is a difference between groups when there is no true difference.
Type I error
This type of statistical test asserts an alternative hypothesis that the occurrence of disease is not the same in the exposed and unexposed groups.
Two-sided test
This is an efficient way to "balance" factors that may also be associated with the outcome of interest in a case-control study that changes our calculation of the odds ratio.
Matching
Bonus points: how does it change our calculation of the odds ratio?
You are interested in examining a possible association between spontaneous fetal loss and antiepileptic drug (AED) use during pregnancy among women with epilepsy (WWE). You use hospital records to identify and recruit 150 WWE who had a spontaneous fetal loss and 300 WWE who had a full-term birth. Recruited women were then interviewed and asked to report all AEDs taken during their pregnancy. 60% of WWE who experienced spontaneous fetal loss used AEDs while only 32% of WWE who had full-term births used AEDs. Name, calculate, and interpret the measure of association for this study.
Odds ratio=(90*204)/(96*60)=3.19
The odds of taking AEDs during pregnancy among WWE who had a spontaneous fetal loss is 3.19 times the odds of taking AEDs during pregnancy among WWE who had a full-term birth.