Unit 6 & 7: Evidence Based Practices Flashcards
SNOUT
Highly SENsitive rules OUT when NEGATIVE
SPPIN
Highly SPecific rules IN when POSITIVE
POSITIVE Liklihood Ratio (+LR)
measure of how well a test:
when POSITIVE rules IN
+LR = 1.0 is USELESS
2, 5, 10 rule
LR 2 = FAIR = +15%
LR 5 = GOOD = +30%
LR 10 = EXCELLENT = +45%
NEGATIVE Liklihood Ratio
-LR
measures how well a test when NEGATIVE rules the condition OUT
-LR = 1.0 USELESS
KAPPA (K)
measure of test reproducibility BEYOND CHANCE ALONE between INDEPENDENT examiners (not within the same examiner)
Acceptable Kappa reliability
greater or equal to 60%
Sensitivity + Specificity = 100%
completely USELESS
The Reciprocal (-LR)
- 5 recip of 2 Good -15% from prevalence
- 2 recip of 5 Fair -30%
- 1 recip of 10 Excellent -45%
-LR = 1.0 is completely useless
Standard Mean Difference (SMD)
raw difference between AVERAGE scores in the treatment group & control group
raw difference between average scores at beginning of RCT and end of RCT WITHIN the SAME GROUP
typically a negative number
When is Standard Mean Difference (SMD) statistically SIGNIFICANT
p < 0.05, then statistically SIGNIFICANT
LESS THAN
Relative Risk (RR) definition
Chance of OUTCOME occurring in the treatment group/chance of it happening in control group
OR treatment/control
Odds Ratio (OR) definition
Chance of EVENT occurring in the treatment group/chance of same thing happening in the control group
Relative Risk GREATER than 1.0
RR > 1.0 = GREATER chance of outcome occurring in TREATMENT group compared to control group
Relative Risk LESS than 1.0
RR < 1.0 = REDUCED chance of outcome occurring in the TREATMENT group compared to the control group
Relative Risk = 1.0
RR = 1.0: NO DIFFERENCE in chance of outcome occurring in either treatment or control group. the treatment is USELESS!
Odds Ratio GREATER than 1.0
OR > 1.0 = odds of event occurring in TREATMENT group is GREATER than same event in the control group
Odds Ratio LESS than 1.0
OR < 1.0 = odds of event occurring in TREATMENT group is LESS than same event in control group.
Odds Ratio = 1.0
OR = 1.0: NO DIFFERENCE in odds of event occurring in treatment and control group
2 Ways to Determine Statistical Significance
P-Value and Confidence Intervals (CI)
P-Value
measure of the probability that a result is PURELY due to CHANCE
(only implies the presence or absence of statistical significance)
Statistically significant/insignificant P-values
statistically significant <0.05
statistically insignificant >0.05
Confidence Intervals (CI)
An estimate of the RANGE OF VALUES that are likely to include the result of the study
The Narrower the CI…
The narrower the CI,
the more PRESCICE the point estimate is
the LARGER the RCT
thus, the SMALLER the trial, the WIDER the confidence interval, the LESS confidence
the LARGER the trial, the NARROWER the interval, the MORE confidence