Disease incidence and prediction.

Question 1

Tests that define an exposure or a disease status are subjected to what sort of error?

Answer 1

Systematic and random error.

Question 2

Is random or systematic error important in regards to bias?

Answer 2

Systematic.

Question 3

What two things can result in information bias?

Answer 3

1. Imperfect definitions of study variables.

2. Flawed data collection procedures.

Question 4

What can information bias result in?

Answer 4

Misclassification of exposure or disease.

Question 5

What two types of misclassification bias exist?

Answer 5

1. Non differential.

2. Differential.

Question 6

What sort of bias exhibits bias towards the null (in terms of the Odds Ratio)?

Answer 6

Non differential.

Question 7

Misclassification bias is similar but different to information bias. True or false?

Answer 7

False, misclassification bias is a type of information bias.

Question 8

Is the OR higher, lower, the same, or both in differential misclassification bias?

Answer 8

Both, it can be higher or lower.

Question 9

Define non differential misclassification bias.

Answer 9

The same amount of bias towards exposed/ non exposed state is the same in both the cases and control group.

Question 10

Define differential misclassification bias.

Answer 10

The same amount of bias towards exposed/ non exposed state is different in the cases and control group.

Question 11

What defines the ‘True status’ of a disease when testing a new diagnostic test?

Answer 11

The current gold standard test.

Question 12

Define sensitivity.

Answer 12

Proportion of individuals with a disease which test positive.

Question 13

Define specificity.

Answer 13

Proportion of those without a disease who tested negative.

Question 14

What is the equation for sensitivity?

Answer 14

TP/(TP+FN)

Question 15

What is the equation for specificty?

Answer 15

TN/(TN+FP)

Question 16

What is the equation for false positivity rate?

Answer 16

FP/(FP+TN) –> FP/ all non diseased

Question 17

What is the equation for false negative rate?

Answer 17

FN/(FN+TP) -> FN/ all diseased

Question 18

Define ‘Positive Predictive Value’ (PPV).

Answer 18

Proportion of positive tests that correctly identify diseased individuals.

Question 19

Define ‘Negative Predictive Value’ (NPV)

Answer 19

Proportion of negative tests that correctly identify non-diseased individuals.

Question 20

What is the equation for PPV?

Answer 20

TP/(TP+FP).

Question 21

What is the equation for NPV?

Answer 21

TN/(TN+FN)

Question 22

Which of the following vary with as disease prevalence varies?

1. Specificity.
2. Sensitivity.
3. PPV.
4. NPV.

Answer 22

PPV/NPV.

Question 23

Despite the fact that PPV and NPV are influenced by prevalence a good test will always show a good PPV. True or false?

Answer 23

False. A good test can result in a poor PPV if the prevalence is low.

Question 24

Is PPV or NPV more commonly used?

Answer 24

PPV (specificity is used more than sensitivity also).

Question 25

Disease prevalence does not modify sensitivity or specificity. What does?

Answer 25

The chosen cut-off value of a continuous variable.

Question 26

What does lowering the cut-off point of a test do? Why does it do this?

Answer 26

Improves sensitivity but lowers specificity.

More false positives.

Question 27

What does increasing the cut-off point of a test do? Why does it do this?

Answer 27

Lowered sensitivity but improved specificity.

More false negatives.

Question 28

When is cut point decision in disease status especially important?

Answer 28

When the distribution of characteristic is ‘unimodal’. This is because the ‘grey area’ is so large.

Question 29

Is a test with a sensitivity of 50%/ specificity of 50% better or worse than a test of sensitivity of 1%/ specificity of 1% better?

Answer 29

Sensitivity of 1%/ specificity of 1%. This is because if you swap the definitions over it becomes 99%/ 99%.

Question 30

What does ROC stand for and why is it used?

Answer 30

Receiver Operator Characteristics.

Used to analyse the trade of between sensitivity and specificity.

Question 31

Would a ROC curve closer to the top left or bottom left corner represent a better test?

Answer 31

Top left corner.

Question 32

What is the AUC for a ROC curve and what does it show?

Answer 32

Area under the curve.

When this is 1 the test is perfect.

Question 33

What does a diagonal line of 45 degrees on a ROC curve represent?

Answer 33

A test with no diagnostic value (same amount of true positives and false negatives).

Question 34

What can you correct for if sensitivity and specificity are known?

Answer 34

Misclassification, providing you have the data for correctly classified subjects.