Diagnostic testing H&S

Question 1

Define sensitivity & equation

Answer 1

The proportion of affected individuals CORRECTLY identified by the test

Sn = True positives / all affected (as determined by the gold standard)

• Read DOWN the table

Question 2

Define specificity & equation

Answer 2

Specificity = the proportion of unaffected individuals CORRECTLY identified by the test

Sp = True negatives / all unaffected (as determined by the gold standard)

• Read down the table

Question 3

How would you maximise sensitivity?

Answer 3

• Decrease the cut off score
• More people would be classed as positive (for depression etc)
• ADV: more people with depression will be flagged
• DISADV: more people without depression would also be flagged

Question 4

How would you maximise specificity?

Answer 4

• Increase the cut off score
• Fewer people will be identified as depressed
• ADV: fewer people without depression would be flagged for therapy
• DISADV: fewer people with depression would also be flagged

Question 5

Equation for true prevalence

Answer 5

(TP + FN) / n

Question 6

Definition for positive predictive value and negative predictive value & equation

Answer 6

PPV = Proportion of positive tests that are correct (true positives / test positives)

NPV = Proportion of negative tests that are correct (true negatives / test negatives)

Question 7

Definition and equation for likelihood ratio

Answer 7

Likelihood ratio is an expression of risk
- LR +ve result = true positive rate / false positive rate = sensitivity / (1 - specificity)

• LR -ve = false negative rate / true negative rate

Question 8

What is the significance of a larger LR+ve and smaller LR-ve

Answer 8

• The larger the LR+ve - greater chance of having disease if your test is positive
• The smaller the LR-ve - less chance of having disease if your test is negative