Math

Question 1

Sensitivity

Answer 1

-Probability that a test result will be positive when the disease is present
-TP/(TP+FN)
-A negative super sensitive test essentially rules out disease

Question 1

Specificity

Answer 1

-Probability that a test result will be negative when the disease is not present
-TN/(FP+TN)
-A positive super specific test essentially rules in disease

Question 2

Confusion matrix

Answer 2

2x2 table in which disease is columns and test is rows

Question 3

Positive predictive value

Answer 3

-Probability disease is present when test is positive
-TP/(TP+FP)

Question 4

Negative predictive value

Answer 4

-Probability disease is not present when test is negative
-TN/(TN+FN)
-decreased by increased prevalence

Question 5

Negative likelihood ratio

Answer 5

-LR(-)
-Ratio between probability of negative test result given present of disease, and the probability of a negative test result given the absence of the disease
-False negative rate / True negative rate
-(1-sensitivity)/specificity

Question 6

Positive likelihood ratio

Answer 6

-LR(+)
-Ratio between probability of a positive test in the presence of disease, and the probability of a positive test in the absence of disease
-True positive rate / False positive rate
-sensitivity/(1-specificity)

Question 7

Type 1 error

Answer 7

-Incorrect rejection of a true null hypothesis
-False positive

Question 8

Type II error

Answer 8

-Failure to reject a false null hypothesis
-False negative

Question 9

ROC curve

Answer 9

-False positive rate on x axis
-True positive rate on y axis
-Sensitivity on x axis
-1-specificity on y axis
-point on upper left is 100% sensitivity and 100% specificity: perfect classification
-Diagonal line is random

Question 10

Bayes’s theorem

Answer 10

P(A|B)=P(B|A)*P(A)/P(B)

Question 11

Odds of X

Answer 11

P(X)/(1-P(X))

Question 12

Fagan Nomogram

Answer 12

Tool for Bayes theorem
-To calculate the post-test probability from the pre-test probability and likelihood ratio

Question 13

Precision

Answer 13

Fraction of retrieved instances that are relevant
-AKA positive predictive value
-Measure of exactness or quality

Question 14

Recall

Answer 14

Fraction of relevant instances that are retrieved
-AKA sensitivity
-Measure of completeness or quantity

Question 15

Relative Risk (RR)

Answer 15

-Risk in study group divided by risk in control group
-i.e. 3/10 DIVIDED BY 5/11

Question 16

Absolute Risk Reduction (ARR)

Answer 16

-Risk in control group MINUS risk in study group

Question 17

Relative Risk Reduction (RRR)

Answer 17

-ARR divided by control risk
-(Risk in control group MINUS risk in study group)/Risk in control group
-Usually more dramatic a number than ARR

Question 18

Number Needed to Treat (NNT)

Answer 18

-1 / ARR

Question 19

Number Needed to Harm (NNH)

Answer 19

When the outcome is harm

Question 20

Fall-out

Answer 20

False positive rate

Question 21

F1 score

Answer 21

2(precisionrecall/precision+recall)