Term
| What is the Exposure (E)? |
|
Definition
| Independent variable that may determine the chance (risk) of getting a Dz. Can be measured via questionnaires, blood tests, chart reviews, etc. Measurement may be problematic if over a long period of time. |
|
|
Term
|
Definition
| Dependent variable. Any deviation from health. This is more straightforward because we can rely on a Dx. However, the definition of the Dz is important and is based on the research question. |
|
|
Term
| What are the most commonly used Study Designs? |
|
Definition
1. Cohort (follow-up)
2. Case control
3. Clinical Trial (Gold Standard for causation) |
|
|
Term
| What are causes of a non-null RR? |
|
Definition
| Chance, bias, confounding, and causation |
|
|
Term
| When can you conclude causation? |
|
Definition
| When you can eliminate confounding, chance and bias as possible explanations. |
|
|
Term
|
Definition
The occurrence of NEW cases of Dz in a population of interest in a stated time period.
Equation:
I = (New Cases)/(# of People at risk) |
|
|
Term
| Define Cumulative Incidence of Risk (CI) |
|
Definition
States what proportion of the subjects under study got the Dz X within the observation period T
Equation:
CI = (# of new cases over T)/(population at risk) |
|
|
Term
|
Definition
The "commonness" of Dz in a given population at a stated point in time
Equation:
P = (# with Dz)/(# of people at risk) |
|
|
Term
|
Definition
Simply the incidence rate of death
Equation:
M = (# of new deaths)/(person-T) |
|
|
Term
| Define Cause Fatality Rate |
|
Definition
| Not a rate at all, it is a cumulative incidence (risk) of death |
|
|
Term
|
Definition
The probability of staying alive for a specific length of time
Equation:
S = (total pts - dead pts)/(total pts) |
|
|
Term
| Define Strength of Association |
|
Definition
| A small association does not mean that there is not a causal effect. When you have two intervals (continuous variables), Pearson's correlation coefficient (r) is used to measure the strength of the association. |
|
|
Term
|
Definition
| Consistent findings observed by different persons in different places with different samples strengthens the likelihood of an effect. Large N's, multiple centers |
|
|
Term
|
Definition
| The more specific an association between a factor and an effect is, the bigger the probability of the causal relationship. Causation is likely if a very specific population at specific site and Dz with no other likely explanation. |
|
|
Term
| Define Temporal Relationships |
|
Definition
| Effect has to occur after the cause |
|
|
Term
| Define Biological Gradient |
|
Definition
| Greater exposure should generally lead to greater incidence of the effect |
|
|
Term
|
Definition
| A plausible mechanism between cause and effect |
|
|
Term
|
Definition
| Laboratory findings, lack of such evidence cannot nullify the epidemiological affect on associations |
|
|
Term
|
Definition
| The effect of similar factors may be considered |
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
| Equation for Positive Predictive Value |
|
Definition
|
|
Term
| Equation for Negative Predictive Value |
|
Definition
|
|
Term
|
Definition
| Sensitivity/(1-Specificity) |
|
|
Term
|
Definition
| (1-Specificity)/Sensitivity |
|
|
Term
|
Definition
Likelihood ratio for a positive test result is the probability of a positive test result for a person WITH the Dz of interest divided by the probability of a positive test for a person WITHOUT the Dz.
Large LR+ values indicate better diagnostic value of the test (<10) |
|
|
Term
|
Definition
Likelihood ratio for a negative test result is the probability of a negative test result for a person with the Dz of interest divided by the probability of a negative test result for a person without the Dz.
Small LR- values indicate better diagnostic value of the test (<0.1) |
|
|
Term
|
Definition
| A variable that distorts the apparent relationship between an exposure and a Dz of interest |
|
|
Term
|
Definition
| A systematic error in a study that arises from mixing of the effect of the exposure of interest with other associated correlates of the Dz outcome |
|
|
Term
|
Definition
| Nonrandom error in a study that leads to a distorted result |
|
|
Term
|
Definition
| Sample distorted by selection process; cases/controls/both |
|
|
Term
| What is Information Bias? |
|
Definition
| Misclassifications of the variables |
|
|
Term
| What is Misclassification Bias? |
|
Definition
| Incorrect characterization of the status of subjects with regard to study variable, leading to a distorted conclusion |
|
|
Term
| What is Differential Misclassification? |
|
Definition
| Incorrect categorization of the status of subjects with regard to one variable (exposure) that is unrelated to another characteristic of interest (Dz status) |
|
|
Term
| What is Non-Differential Misclassification? |
|
Definition
| Incorrect categorization of the status of subjects with regard to one variable (exposure) that is unrelated to another characteristic of interest (Dz status) |
|
|
Term
| What is Ecological Fallacy? |
|
Definition
| Fallacy assumes that individual members of a group have the AVERAGE characteristics of a group at large (stereotypes) |
|
|
Term
|
Definition
| Cases are more likely than controls to over-report |
|
|
Term
| What is Confidence Interval? |
|
Definition
The correct interpretation of a study with a 95% confidence interval which is (-0.81 -- +1.41) is as follows:
Crosses 1 (truth) - anytime the confidence interval crosses 1, it is not usually significant |
|
|
Term
|
Definition
| The point on the scale of a measure of association that corresponds to no association (1 for the risk ratio and odds ratio and 0 for the risk difference and the attributable risk %) |
|
|
Term
| What is the Null Hypothesis? |
|
Definition
Null is necessary for statistical testing and possibly conceptually and for keeping causality in perspective.
--All swans are white. |
|
|
Term
| What are the 3 Steps of a Manuscript Review? |
|
Definition
1. Read the manuscript thoroughly
2. Wait...
3. Review Manuscript |
|
|
Term
|
Definition
1. Process of analyzing data
2. Description of a field of study
3. Data or numbers |
|
|
Term
| Describe the Types of Data |
|
Definition
Nominal: dichotomous, binary, or categorical
Ordinal: stages
Numerical: continuous |
|
|
Term
| Describe the Measures of Control Tendency |
|
Definition
Mean: average of observations
Median: the middle observation
Mode: most common value
Distributions: Symmetric- same shape on both sides of mean (use mean); Skewed- not the same shape (use median) |
|
|
Term
| Describe the Measures of Spread |
|
Definition
Range: difference between smallest and largest
Standard Deviation: measure of spread of data about their mean
Variance: the square of the standard deviation. Measures the "spread" of a distribution |
|
|
Term
| Describe the Confidence Interval |
|
Definition
A range (interval) around point estimate. Helps determine the "reliability" of an estimate. CI - based on mean of population, distribution, statistic, and standard error. Used when comparing between two groups, it is a measure of how "confident" we are that the statistics are TRUTH.
If CI passes 1, it means it is less reliable; if it doesn't pass 1, it's the truth |
|
|
Term
|
Definition
The probability of observing a result as extreme or more extreme than the one actually observed from chance alone (if the null is true). Takes into account the mean and variance of a distribution and estimates how "likely" or "unlikely" your observed outcome is.
Statistical significance: usually defined as p<0.05
--Smaller P value = stronger association
--Larger the sample size = smaller the P value |
|
|
Term
|
Definition
A statement about a population
-Test a null hypothesis (Ho) and an alternative hypothesis (H0 or H1)
-Alternative Hypothesis: what do we believe to be true if the null is rejected |
|
|
Term
| What are the steps to testing a Hypothesis? |
|
Definition
1. Evaluate data
2. Review Assumptions
3. State hypothesis
4. Select test statistic
5. Determine distribution of test statistic
6. State decision rule
7. Calculate test statistic
8. Make decision
9. Conclusion: Have we found the truth |
|
|
Term
| What is a Type I error? Type II? |
|
Definition
Type I: false positive, if we do a test 20 times, 1 will be wrong
Type II: false negative, did not have enough information to tell the truth |
|
|
Term
| How do you compare Mean values? |
|
Definition
| t-Test (Student's t-Test) |
|
|
Term
| How do you compare Median values? |
|
Definition
Non-parametric tests (Wilcoxon rank sum, Kruskall Willis)
--Just means the sample size was not a bell curve |
|
|
Term
| How do you compare Proportions? |
|
Definition
| Chi Square or Fisher's Exact test - test categorical data "yes/no", %'s |
|
|
Term
| What are Measures of Effect? |
|
Definition
| Express how two groups or populations differ with regard to a measure of Dz frequency (I, CI, P) |
|
|
Term
|
Definition
| Relative risk is a ratio measure (not rate or proportion). A ratio is a measure of something in one population divided by the measurement of that same something in another population. |
|
|
Term
| What does it mean when RR does NOT = 1? |
|
Definition
There is a statistical association between E (exposure) and D (Dz)
-E is predictive of or associated with D |
|
|
Term
| What does it mean when RR DOES = 1? |
|
Definition
There is no association between exposure and Dz
-E is NOT predictive of D |
|
|
Term
| What is Medical Surveillance? |
|
Definition
The ongoing, systematic collection, analysis, and interpretation of outcome-specific data for use in the planning, implementation, and evaluation of public health practice
--Active: system in which the public health unit contacts reporting sources to elicit reports
--Passive: system in which information is received from physicians as mandates by law |
|
|
Term
| What are the goals of serveillance? |
|
Definition
-Estimate magnitude of problem
-Detection of Dz outbreaks
-Development of clues about possible risk factors
-Finding of cases for further investigation
-Anticipation of health service needs |
|
|
Term
|
Definition
| An epidemic that occurs suddenly and within a relatively defined geographic area |
|
|
Term
| What are the three components of an outbreak? |
|
Definition
1. A pathogen in sufficient quantities to affect multiple persons
2. An appropriate mode of transmitting the pathogen to susceptible persons; person-to-person; common source exposure
3. An adequate pool of susceptible persons who are exposed to the pathogen |
|
|
Term
| What are the important investigational steps with an outbreak? |
|
Definition
1. Verify that there is an outbreak
2. Survey cases and contacts
3. Formulate hypotheses for causation and spread
4. Plan and conduct a field investigation
5. By a Hazmat suit and a machine gun
6. Analyze data using appropriate statistical tests
7. Test hypotheses against data
8. Formulate conclusions
9. Initiate control measures
10. Prepare a report for comparison to other outbreaks
11. Media interaction |
|
|
