EBM

1
EBM

• Dr Adrian Burger
• 20 March 2007

2
EBM Intro

• Evidence based guidelines
• Evidence based care paths
• Evidence based Questions
• Evidence based Solutions
• Evidence based

3
Shift_Expectations

• Whether to implement
• How to - sensibly
• - efficiently
• - avoid misconceptions
• EBM CONCEPTS ARE NOW UNAVOIDABLE

4
(Mis)Concepts

• EBM DOES NOT RCT
• Cannot rely solely on research from RCT
• - lt14 are RCT
• - Observational studies overlooked
• - undervalued patient preferences, clinical
• circumstances, expertise

5
What You Need to Do

• Clearly define your question
• Thorough literature search
• Critically appraise the quality of the evidence
• Critically assess the applicability
• Balanced application

6
The evidence cycle
7
Balanced application of the evidence
8
Clinical circumstances

• Clinical expertise and judgment
• Assumes physician competence on basis of
• - clinical experience
• - development of clinical instincts

9
Patient preferences

• Individualizes needs
• Takes into account patients
• - experiences
• - values
• - expectations

10
Research evidence

• Ask the question
• Acquire the information
• Appraise the quality
• Apply the results
• Act on the patient

11
EBM Resources

• Own research, reviews
• - Guide to the Medical Literature
• Online
• - Cochrane Database
• - Best Bets

12
Levels of Evidence

• Hierarchy of evidence
• RCT
• Controlled observational studies
• Uncontrolled studies and opinion

13
Studies First Question

• Is the study
• - investigating results of a treatment?
• Therapeutic Study
• - outcome of a disease?
• Prognostic Study
• - Diagnostic test?
• - Developing an economic model/decision
• analysis?

14
Studies Categorize the study

• Level 1
• RCT or systematic review of RCTs
• Level 2
• Prospective cohort, poor quality RCT,
  systematic review
• Level 3
• Case control, retrospective cohort, systematic
  review of level 3 studies

15
Categorize the studies 2

• Level 4
• Case series
• Level 5
• Expert opinion

16
Level One and Two
17
Level 3-5
18
Errors in Hypothesis Testing

• Small sample studies
• - Type 2 Beta errors
• Probability of concluding that no difference
  exists when there is a difference
• Acceptable rate is 20
• This equals a power of 80
• If gt20 then unacceptably high risk of false
  negatives

19
Errors 2

• Type 1 or Alpha Error
• Concluding the results of a study are true
  when in fact it is due to chance or random
  sampling error
• Is an erroneous false positive
• Acceptable rate of 5

20
Power of a study

• Probability of concluding a difference between
  two treatments when one actually does exist
• Power 1- Beta
• Can be used before a study to determine study
  sample size
• Can be used after a study to determine if
  negative findings true or just due to chance

21
Article

• Injury, 37, 302 306
• Evidence-based Medicine What it is
• and what it is not