06. Diagnostic Testing

Question 1

Sensitivity

Answer 1

“True positive rate”; Proportion of people with disease who have a true positive test result; How good a test is at detecting people with disease; Insensitive to prevalence

Question 2

Specificity

Answer 2

“True negative rate”; Reflects how good the test is at IDing people without disease; unlikely to produce false positives; insensitive to prevalence

Question 3

True positive

Answer 3

Test positive & disease present

Question 4

False positive

Answer 4

Test positive & disease absent

Question 5

True negative

Answer 5

Test negative & disease absent

Question 6

False negative

Answer 6

Test negative & disease present

Question 7

Likelihood ratio (LR) definition

Answer 7

The likelihood that a given test result would be expected in a patient with the disease compared to the likelihood that the same result would be expected in a patient without the disease. Can be a positive or negative value.

Question 8

Importance of likelihood ratio

Answer 8

Useful for assessing the value of performing a diagnostic test. Use the sensitivity & specificity of the test to determine whether the test result will significantly change the probability that a disease exists.

Question 9

LR+ Equation

Answer 9

[Sensitivity/ (1-specificity)]; Amount by which the pretest probability is INCREASED in pts with POSITIVE test

Question 10

LR+ Interpretation

Answer 10

SUPPORTS TRUE DISEASE; Should be slightly significant above 1. LR+ >10 causes large shift in posttest probability of disease

Question 11

LR- Equation

Answer 11

[(1-sensitivity)/specificity]; amount by which the pretest probability of disease is REDUCED in patients with a NEGATIVE test

Question 12

LR- Interpretation

Answer 12

Supports NO DISEASE; Should be significant

Question 13

LR = 1

Answer 13

No shift in probability of disease

Question 14

Predictive value importance

Answer 14

Sensitive to prevalence; indicates what the test MEANS to the patient

Question 15

Positive predictive value

Answer 15

% of positives that are TRUE POSITIVES; [True positives/All positives]

Question 16

Negative predictive value

Answer 16

% of negatives that are TRUE NEGATIVES; [True negatives/All negatives]

Question 17

Parallel testing

Answer 17

Shotgun; Many tests at once. ANY positive is diagnostic. INCREASES SENSITIVITY. Reduces risk from lack of intervention.

Question 18

Series testing

Answer 18

Rifle; Sequence of diagnostic tests - ALL tests must be positive for diagnosis (think about tests needed to confirm +HIV status); INCREASES SPECIFICITY; Reduces risk of unnecessary intervention. Reduces sensitivity.