Odds
probability/(1-probability)
Likelihood Ratio (+) Survival analysis (Kaplan-Meier)
Odds of a finding in diseased versus Useful when subjects are followed for
healthy individuals. How many times different periods of time. Survival ratios
more likely a test result will be found in are remaining subjects over subjects at
diseased versus healthy people. last event. Subtract drop-outs from
(sensitivity)/(1-specificity) numerator and denominator at each
(probability present in diseased step. Cumulative survival is product of
pts)/(probability present in non-diseased survival ratio at each death/event.
pts) Calculate hazard ratio instead of
relative risk.
Likelihood Ratio (-)
Hazard ratio
(probability absent in diseased
pts)/(probability absent in non-diseased Basically, it's relative risk ratio specific
pts) to Kaplan- Meier survival analysis.
(1-sensitivity)/(specificity)
Intention-to-treat analysis
Clinical Trial
Patient's data analyzed with group
Experimental study. Prospective. originally assigned to
Interventional.
Bias
Cohort study
Systematic error that skews results
Observational. Usually prospective.
Identify exposure and observe incidence Selection bias
of outcome.
One group under/overrepresented.
Case-Control Unreasonable exclusion criteria.
Allocation concealment
Logistic regression
Parameter
Statistic
Alpha
Beta
(1-Beta)
The Power of the experiment to detect a
difference, rejecting the null hypothesis
correctly.
P value
Confidence interval
Evidence-based Medicine
Foreground question
Background question
1. Temporal
2. Strength of association
3. Dose-response relationship
4. Replication of findings
5. Biologic plausibility
6. Consideration of alternative
explanations
7. Cessation of exposure
8. Consistency with other knowledge
9. Specificity of association
Prevention
ROC curve