Week 5 - Evidence Based Anesthesia Practice Flashcards
Describe the research pyramid from the base to the top (9 levels)
- In vitro “test tube” research
- Animal research
- Editorials and opinions
- Case reports
- Case series
- Case-control studies
- Cohort studies
- Randomized controlled trials
- Meta-analysis
*increasing clinical relevance as you move through the levels
What is a parametric test?
- The observations must be independent
- The observations must be drawn from normally distributed populations
- These populations must have the same variances
- The means of these normal and homoscedastic populations must be linear combinations of effects due to columns and/or rows
Ex: z test or t test, t test between or within groups, 1 way ANOVA between groups, 1 way ANOVA within or repeated measure, Pearson’s r
What is a non-parametric test?
Observations are independent
Variable under study has underlying continuity
Ex: x^2, Macnarmar’s x^2, Cochran’s Q, Mann Whitney U, Spearman’s rho, Kruskal Wallis H, Friendman’s ANOVA
What are considerations for applying evidence to your patient?
- Are my patients similar to the study population?
- Is the suggested intervention one with which I am familiar or skilled at performing?
- What are the risks? Risk/Benefit analysis?
- Is the patient well informed and do they understand the risks and benefits?
- Did the patient make an independent decision to undergo this intervention?
What is a Confidence Interval?
Range within something is likely to occur
- 95% CI means that 19 of 20 times the true value will land within the specified range
- Narrower CI is better than wider CI
What is an Odds Ratio?
Likelihood (odds) of something happening to two different groups
Ex: What are the odds of death in group A? What are the odds of death in group B? What is the ratio of death in group A to group B?
What is the NNT (number needed to treat)?
How many patients must be treated with an intervention to prevent 1 additional bad outcome?
NNT 8 means treat 8 to prevent 1
What is a Power Analysis?
How many subjects are needed to make sure you found what you think you found and didn’t make an error
- Type 1 Error: incorrectly accept the null hypothesis that a difference between two groups exists when one did NOT
- Type 2 Error: incorrectly reject the null hypothesis that a difference between two groups did NOT exists when one actually DID
*involve setting a alpha level (p-value)
What is a Meta-Analysis?
Combination of multiple studies on the same topic or intervention and combine those results to increase the power, validity, effect size
-use high quality RCTs, extract relevant data, integrate studies findings using common statistical metrics, summarize and report
Internal Validity vs External Validity
Internal Validity: the measure is measuring what it is supposed to be measuring
External Validity: to whom can these measured results be applied
