EBP Review

Spring Semester Y1

Prevalence Calculation

(Total number of people who have condition/Total number in study)

Sensitivity Calculation

(Number of true positives/Total number of people who have condition)

What do sensitivity calculations tell us?

The number of people correctly identified to have a condition (who had X condition and tested positive)

Used to rule OUT conditions (higher sensitivity means higher chance we can rule out that diagnosis)

Specificity Calculation

(Number of true negatives/Total number of people who do not have condition)

What does specificity tell us?

The number of people correctly identified to not have a condition (who did not have X condition and tested negative)

Used to rule IN conditions (higher specificity means higher chance we can rule in that diagnosis)

Positive Predictive Value

Probability that those who tested positive actually have the condition

Negative Predictive Value

Probability that those who tested negative do not actually have the condition

Pre-Test Probability

Same thing as prevalence

(Total number of people who have condition/Total number in study)

Positive Likelihood Ratio

Sensitivity / (1-Specificity)

How much a positive test result increases the likelihood of a condition being present

Negative Likelihood Ratio

(1 - Sensitivity) / Specificity

How much a negative test result reduces the probability of having a condition

While reading a randomized controlled trial, which value is the best indicator of the magnitude of effectiveness of an intervention?

Cohen’s d = 0.5 (aka Effect Size)

Cohen’s d

Standardized measure of effect size used to quantify the difference between two group means

Expressed in standard deviation units

P Value

Tells us if there is a significant difference between Group A and Group B

Level of Significance (alpha)

Maximum amount of acceptable error you can have

Cronbach’s alpha

Used exclusively in questionnaires

Measure of internal consistency

Tells us how well the items in a set fit together (agreeableness with one another)

Intraclass Correlation Coefficient (ICC)

Reliability statistic used for continuous variables (such as numerics like goniometry)

Example: ROM, height, weight (continuous variables)

Kappa

Reliability statistic used for non-continuous (categorical, dichotomous) variables

Example: FABER test (either positive or negative, no range)

What p values will lead to the rejection of the null hypothesis (level of significance = 0.05)

Any values less than or equal to the p value (<= 0.05)

Which variable is an indicator of excellent internal consistency?

Cronbach alpha = 0.9

Which value represents the smallest amount of change that can be considered above threshold of error expected in the measurement?

Minimal detectable change (MDC)

How does a systematic review differ from narrative reviews?

Systematic reviews conduct quality appraisal of studies to report bias

I²

Degree of inconsistency or heterogeneity

Proportion of variability in effect estimates across studies that is due to true heterogeneity (differences between studies) rather than chance

What does a low I² value indicate?

There is no heterogeneity, which we want to have low values

Indicates a positive result