FoS Flashcards
5 A’s in evidence based medicine
Ask Acquire Appraise Apply Assess
Critical appraisal
Systematic approach to identify the strengths and weaknesses of a research article to assess the validity and usefulness of research findings
Requires knowledge of research methods and application
Why is there a positive correlation between impact factor and retraction index
Most highly cited journal in medicine has the most number of retractions
This is because people really want to publish there, so they look the other way when their hypothesis is wrong
3 questions to ask when critically appraising a study
Are the results valid?
What are the results?
Will the results help me care for my patient?
6 sources of bias
Randomization Blinding Allocation concealment Intention-to-treat Follow up Conflicts of interest
When examining the results, what 3 questions do you want to ask?
Size of the effect
Relative vs absolute numbers
Confidence intervals around the effect seen
What is MeSH?
A very specific type of thesaurus that describes people, places, publication types, and anatomical/physiological states and conditions
Captures otherwise hard to express concepts and unites the literature on common topics under a single umbrella
PICO
Way to format a research question (quantitative) Population Intervention Comparison Outcome
PEO
Qualitative research question
Population
Exposure
Outcome
Confidence interval
Tells us how sure we are that the true mean is in the interval
P value
Tells you how far above a value you need to get before you decide that it probably isnt the mean
The likelihood that getting X or greater is true (the probability of getting our estimate or more extreme, if the null hypothesis is true)
Don’t just look at p value!
3 values you want to look at to determine if a number is valid
P value
Confidence interval
Sample size
Null statement
There is NO relationship between X and Y
Prevalence vs incidence
Prevalence: how many people currently have the disease
Incidence: how many new cases per year
Can you measure incidence or prevalence in:
- Cross sectional studies
- Cohort studies
- Case control studies
- Prevalence
- Incidence
- Neither
Standard error
On average, how far will an estimate move from the true/population value
If the confidence interval crosses what value, then you reject the null hypothesis?
1
If the confidence interval crosses 1, then it is basically saying that the true OR could show either an increase in the exposed or unexposed people
3 processes/outcomes of CBPHR
All members engaged in systematic research investigation, developing new knowledge together
Actively involving all members in co-learning processes
For the purposes of action conducive to health
