7 Sensitivity vs Specificity

SpPIN

Specificity Positive Rules In

Negative likelihood ratio formula

(1-sensitivity)/specificity

Likelihood ratios numbers to know

2,5,10 15, 30, 45% Positive LR of 2, shifts 15% Positive LR of 5, shifts 30 Positive LR of 10, shifts 45 Negative LR of 1/2, shifts 15% Negative LR of 1/5, shifts 30 Negative LR of 1/10, shifts 45

PPV formula

A/(A+B)

Sensitivity formula

A/(A+C)

McMurray's test .96 specific, meniscus tear test

Among people known to have no meniscus tear, this test is negative 96% of the time •False positives are rare •If the test is positive it is most likely torn

McMurray's test sensitivity .29, meniscus tear test

Among people with known meniscus tear, this test is positive 29% of the time (false negatives very common, negative test result doesn't help us much)

Contingency table used to

Arrange data according to test result vs truth. You want a positive test and positive gold standard, you want a negative test with a negative gold standard

LR positive greater than 5

Clinically significant, increases your confidence of ruling in with a positive finding

Specificity formula

D/(B+D)

NPV formula

D/(C+D)

Discrete data is

Data that can't be subdivided (yes or no)

LR negative less than -.2

Increases your confidence of ruling out with a negative

Likelihood ratios

Indicate the value of a test relative to our certainty of a diagnosis or outcome (more generalizable)

A highly sensitive test will rarely be ________ among people who have the condition

Negative (because we are testing it with people that have the condition)

High specificity improves ______

Our confidence ruling in the diagnosis/outcome

A negative test result for high sensitivity rules _____

Out the diagnosis

A highly specific test will rarely be _______ among people who do not have the condition

Positive (because we are testing it with people who don't have the condition)

Most tests in PT examination is

Positive or negative, normal or abnormal

PPV and NPV are highly sensitive to the

Prevalence of the condition (aka, running clinics will get more occurences of stress fractures than your average clinic because runners are much more likely to get stress fractures)

Negative likelihood ratio

Ratio of a negative test result in people with and without the pathology. The closer the LR is to 0, the less likely the disease/event

Positive likelihood ratio (LR+)

Ratio of a positive test result in people with and without the pathology (bigger the number, the more convincingly the finding suggests the disease/event)

A positive test result for a test with high specificity ___

Rules in the diagnosis/outcome

High sensitivity improves our confidence of _____-

Ruling out the diagnosis/outcome.

SnNOUT

Sensitivity Negative Rules Out

Positive LR (likelihood ratio) formula

Sensitivity/1-specificity

Negative predictive value (NPV)

percentage of those identified by the test as negative who do not have the diagnosis

Positive Predictive Value (PPV)

percentage of those identified by the test as positive who have the diagnosis

Sensitivity

proportion of individuals with a particular diagnosis who are correctly identified as positive by the test (only evaluated among people with a condition) True positives

Specificity

proportion of individuals without a particular diagnosis who are correctly identified as negative by the test Only evaluated among people without a condition True negatives