Critical Appraisal RCT Introduction Flashcards
Assessing Cause and Effect
Each individual is either exposed to the treatment or not
- Assess which ones had the outcome and which ones didn’t
All confounding variables should be kept the same with the only difference being exposure
How do biases arise
The only difference in groups should be the experimental intervention
- Biases results from differences in prognostic factors (Patient Characteristics)
Steps to practice EBM
- Ask
- Acquire
- Appraise
- Apply
- Act and Assess
Using pre-appraised resources
Have to be careful as individual studies can be misleading
Random Error
- What is it
Random Error: Pure chance causes differences
Systematic Error
- What is it
Systematic Error (Bias)
- Underestimation or Overestimation of underlying effects leading to Systematic Deviation from the underlying truth
Systematic Error
- Causes
Caused by a feature of the design or the conduct of the research study
Random Error vs Systematic Error
Random Error: Graph is spread out and more variable, mean is still the same
- Does not affect the mean
Systematic Error: Graph’s mean deviates from the original
- Affects the mean
EBCP definition
Evidence Based Clinical Practice
- Find the best available evidence from systematic reviews and meta analysis
- Determine the strength of those evidences
- How useful it the evidence’s result to the patient
Why is it important to study critical appraisals
To be able to assess the quality of studies and deal with misinformation
Be able to tell the difference from high and low quality research
Resources to find study synopses and critical appraisals
Wiki Journal Club
ACP Journal Club
RX Files
BMJ Evidence Based Medicine
Summary Resources
Dynamed
Lexidrugs
Major categories of bias
- Arises from randomization process
- Deviations from intended intervention
- Missing outcome data
- Measurement of the outcome
- Selection of the reported result
Steps to appraise a therapy trial (JAMA approach)
- Risk of Bias
- Clinical Relevance
- Applicable to Patient Care
Risk of Bias
- JAMA Approach
- Was prognostic balance maintained beginning, during the study, and the end
Clinically Relevance
- JAMA Approach
How large is the treatment effect
How precise is the estimate of the treatment effect
Applicable to Patient Care
- Were the study patients similar to my patients
- Were all important outcomes considered
- Are the benefits worth the harms and costs
Understanding Confidence Intervals
- 1.43 [0.51,4.05]
What does 1.43 mean
1.43 times / 43% more likely
For a good event
- Greater than 1 means more likely for good event
- Less than 1 means less likely for good event
Understanding Confidence Intervals
- 1.43 [0.51,4.05]
What does 0.51 mean
Lower limit
- Cure rate are up to 50% / 0.50 times lower with duct tape
Understanding Confidence Intervals
- 1.43 [0.51,4.05]
What does 4.05 mean
Upper limit
- Cure rate are up to 300% / 4 times higher with duct tape
Understanding Confidence Intervals
- Sufficiently precise
If 95% CI crosses a decision threshold then the results are not sufficiently precise
If 95% CI does not cross a decision threshold then the results are sufficiently precise
Relevance
- Does the info focus on outcomes patients care about
- Is the intervention feasible and is the problem it addresses common
- Would the information cause a change in practice
If yes to all 3 it is patient oriented evidence
Validity
Technical rigor of the work (Is it accurate?)
Work
Time, money, and effort to answer a question
