Screenings in Medicine

Question 1

What are the two questions to a physician?

Answer 1

How accurate is the screening test you’re recommending?

How confident are you in the results when they come back?

Question 2

What are the four possible outcomes for screenings?

Answer 2

1. True Positive (A) Truth and accuracy
2. True Negative
3. False Negative
4. False Positive (D) Truth and accuracy

Question 3

What is a true positive?

Answer 3

Test correctly reports a positive result in a patient who truly has disease. (Box A)

Question 4

What is a True negative?

Answer 4

Test correctly reports a negative result in a patient who does not have disease. (Box D)

Question 5

What is a false postivie?

Answer 5

Test incorrectly reports a positive result when the patient does not have disease. (Box B positive test row negative disease column)

Question 6

What is a false negative?

Answer 6

Test incorrectly reports a negative result when the patient does have disease. (Box C Negative row Yes disease column)

Question 7

What tests determine how accurate the screening tests are?

Answer 7

Sensitivity and specificity

Based on KNOWN disease status

Question 8

What is sensitivity?

Answer 8

% of time in a known positive sample that the test indicates a positive test. How accurate can a test correctly detect presence of disease when disease is actually present. Highly sensitive test has a low false negative rate. 100% sensitive = 0% False negatives
TP/(TP+FN)x100%
A / (A+C)

Question 9

What is specificity?

Answer 9

% of time that a test is negative in a patient that is known to not have disease. How accurately a test can correctly detect the absence of disease when the disease is known to be absent.
Box D/ (B+D)
TN/(FP+TN)x100%
High specific test has a low false positive

Question 10

What question does sensitivity and specificity anwer?

Answer 10

How accurate is the screening test?

Question 11

Can a test have great specificity and bad sensitivity?

Answer 11

Yes

Question 12

Interpret a sensitivity of 38% and a specificity of 62%.

Answer 12

Only 38% of the time the test comes back accurate (positive) on patients who truly have disease. (sensitivity)
Only 62% of the tests come back negative in patients who do not have disease in those who are known to not have disease. This indicates that the test is better in the absence of disease because the specificity is higher.

Question 13

What tests are used to determine how confident you can be in the results of a test?

Answer 13

Positive Predictive Value and Negative Predictive Value

Question 14

What is PPV?

Answer 14

Percentage of time that a test comes back positive that is right for the patient. Correct prediction. High PPV has a low false positive rate. PPV=TP/(TP+FP)x100%
A/(A+B) row totals

Question 15

What is NPV?

Answer 15

Percentage of time that a test comes back negative correctly predicts that the patient does not have disease.
NPV=TN/(FN+TN)x100%
D/(C+D)

Question 16

How does PP and NPV change with prevalence?

Answer 16

It is easier to find disease with higher prevalence because its easier to find things that exist together.
However prevalence will not change sensitivity and specificity because those are about the test accuracy.

Question 17

Validity?

Answer 17

Ability to distinguish between those with and without disease

Question 18

Reliability?

Answer 18

Ability to reproduce results

Question 19

T or F: A valid test is always reliable.

Answer 19

Yes

Question 20

Likelihood ratio (-)?

Answer 20

Probability of negative test in those with true disease and without. False negatives true negatives

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

Question 21

LR (+)?

Answer 21

Likelihood of positive test in those with true disease vs positive without disease.

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

Question 22

Lt (+) should be <0.1 True or False?

Answer 22

False LR+ is > than 10

Question 23

LR (-) needs to be less than 0.1?

Answer 23

True