Evidence cycle
Ask–acquire–appraise–apply
Hayward’s evidence cycle
Adds in assess
Making a health decision takes what into consideration
Patients circumstance, research evidence, clinical expertise, and patients values/preferences
Background question components
Question + condition
Ex: (what causes) +(Lyme disease)
Foreground question components
“PICO”
Patient +intervention+ comparison + outcome
Diagnostic PICO
P: describe the patient
I: new diagnostic test
C: gold standard test
O: condition to be detected
Prognostic PICO
P: describe patient
I: prognostic factor
C: no prognostic factor
O: health state
Harm PICO
P: describe the patient
I: exposure
C: lack of exposure
O: health state
What is EBCP
Recognize problems, asking questions, evaluating and applying evidence
Clinical research
Basic science research
6 Levels of the hierarchy pyramids
Likelihood of bias
Case control studies or cohort studies
Start with patients who have the disease
Match them with controls
Retrospective
(+) ethnically challenging diseases, rare conditions, limited resources
Best study to show causation
RCT
Meta-analysis or systematic review
Combines the results of many studies
Sensitivity
Probability of a positive test in someone who has the disease
“PID”
In a 4*4 it’s the top left/ total left column
Specificity
Probability of a negative test in someone who doesn’t have the disease
“NIH”
In a 4*4 it’s the bottom right/ total right column
Positive predictive value
Probability that a patient with a positive test really does have the disease
In a 4*4 it’s the top left/ top row
Negative predictive value
The probability that a patient with a negative test really does not have the disease
In a 4*4 it’s the bottom right/bottom row
Positive likelihood ratio
ratio (+in disease/ total in disease)/ (+ no disease/ total no disease)
What are examples of exposures
Survey, file review, interview
Interpretation of + likelihood ratio
1: test is 5 times more likely in patients w/ disease than without
Statistical Test to use if:
Describe one group
Normal: mean, SD
Non-normal: median