Evaluating Diagnosis

Question 1

What do studies of a diagnostic test usually evaluate the relationship between?

Answer 1

Being positive for standard diagnostic test vs being positive for evaluated screening test

Question 2

What are the 4 groups in diagnostic test studies

Answer 2

True positive = A
False positive = B
False negative = C
True negative = D

Question 3

First column for table for diagnostic test studies

Answer 3

Test positive
Test negative
Total

Question 4

Second column

Answer 4

Disease (gold standard positive)
A
C
A+C

Question 5

3rd column

Answer 5

No disease (gold standard -ve)
B
D
B+D

Question 6

4th column

Answer 6

A+B
C+D
A+B+C+D

Question 7

First row

Answer 7

Diseased, No disease

Question 8

2nd row

Answer 8

Test positive, A, B, A+B

Question 9

3rd row

Answer 9

Test negative, C, D, C+D

Question 10

4th row

Answer 10

Total, A+C, B+D, A+B+C+D

Question 11

Define sensitivity

Answer 11

Usefulness of a test for truly diseased population

Question 12

Define specificity

Answer 12

Usefulness of a test of non-diseased population

Question 13

Calculation for sensitivity

Answer 13

A/A+C

(true positive/total diseased)

Question 14

Calculation for specificity

Answer 14

D/B+D

(true negative/total non-diseased)

Question 15

Calculation for accuracy

Answer 15

A+D/A+B+C+D

(true positive + true negative/total population)

Question 16

Characteristics of a highly sensitive test

Answer 16

Will pick up cases that show even slightest of evidence
Makes more false positives

Question 17

When is a highly sensitive test preferred?

Answer 17

If diagnosis should not be missed but overdiagnosis is not harmful

Question 18

Characteristics of highly specific test

Answer 18

Will pick up cases only if definitive evidence noted
Makes more false negatives

Question 19

When is a highly specific test preferred?

Answer 19

If missing some cases is not bad but wrong labelling can be costly/harmful

Question 20

Impact of higher cut off on specificity and sensitivity

Answer 20

Low sensitivity
High specificity

Question 21

When is a receiver operator curve (ROC) useful?

Answer 21

Choosing between two diagnostic tests of different sensitivity and specificity
Choosing a cut off point for making a diagnosis

Question 22

X and Y axis of ROC curve

Answer 22

X: (1-specificity)
Y: sensitivity

Question 23

How do interpret a ROC curve

Answer 23

Elbow/knee of the curve is cut-off point with balance between sensitivity and specificity.
Among 2 curves, the one closer to the left upper corner is better screening test

Question 24

What is ROC analysis used for?

Answer 24

To select optimal cut off point
To dichotomize a continuous scale

Question 25

What does accuracy of ROC analysis depend on?

Answer 25

Quality of gold standard test

Question 26

Impact of increase in sensitivity on specificity?

Answer 26

Decrease in specificity

Question 27

How do we know if a test is not accurate in a ROC curve?

Answer 27

The closer the curve comes to the chance line

Question 28

What is the area under the ROC curve?

Answer 28

Accuracy

Question 29

What is Likelihood ratios for a positive test (LR+)?

Answer 29

Likelihood for testing positive rightly/wrong

Question 30

What is Likelihood ratio for a negative test (LR-)?

Answer 30

Likelihood of testing negatively rightly/wrongly

Question 31

Calculation for LR+

Answer 31

(A/A+C)/(B/B+D)

Question 32

Calculation for LR-

Answer 32

(C/A+C)/(D/B+D)

Question 33

Name another calculation for LR+

Answer 33

Sens/1-spec

Question 34

Name another calculation for LR-

Answer 34

1-sens/spec

Question 35

What is LR+ for most clinically available tests?

Answer 35

>1

Question 36

What value of LR is used to use a diagnostic tool

Answer 36

>10

Question 37

What LR value is used to avoid a diagnostic tool

Answer 37

<0.1

Question 38

What does likelihood of a diagnosis depend on?

Answer 38

Prevalence or prior probability of disease before applying the test

Question 39

What is the pretest probability

Answer 39

Prevalence in studied population

Question 40

Calculation for pretest probability

Answer 40

(A+C)/(A+B+C+D)

Question 41

How to calculate post-test probability

Answer 41

Convert pretest probability to prest odds Convert pretest odds to post test odds Convert posttest odds to post-test probability

Question 42

Calculation to convert pretest probability to pretest odds

Answer 42

Prob/1-prob (A+C)/(B+D) i.e. diseased/not diseased

Question 43

Calculation to convert pretest odds to post-test odds

Answer 43

LR x Pretest odds

Question 44

Calculation to convert post-test odds to post-test probability

Answer 44

Post-test odds / 1+ posttest odds

Question 45

What is the Bayesian nomogram to calculate post-test probability

Answer 45

Draw a straight line across pretest probability and likelihood ratio Post-test probability s the value obtained on the other side of the nomogram

Question 46

Who created the Bayesian Nomogram

Answer 46

Fagan

Question 47

What does positive predictive value (PPV) tell us?

Answer 47

The chance of having the disease if test +ve

Question 48

Calculation for PPV

Answer 48

A/A+B

Question 49

What does PPV increase with

Answer 49

Prevalence

Question 50

What does the negative predictive value (NPV) tell us?

Answer 50

Chance of not having the disease if test -ve

Question 51

Calculation for NPV

Answer 51

D/C+D

Question 52

What happens to NPV as prevalence increases?

Answer 52

NPV decreases