W5, 6

Question 1

Likelihood Ratios formula?

Answer 1

Probability that you’d see some evidence if your hypothesis were true/ Probability that you’d see the exact same evidence if your hypothesis were false

Question 2

Define Likelihood ratio?

Answer 2

A number representing the diagnostic usefulness of a test

Question 3

LR=1?

Answer 3

Useless (doesn’t count as an evidence)

Question 4

LR>1?

Answer 4

Increases the probability (the higher the better)

Question 5

LR<1?

Answer 5

Decreases the probability (the lower the better)

Question 6

LR+?

Answer 6

Finding was present (not that it necessarily increases the probability)

Question 7

LR-?

Answer 7

Finding was absent (not a negative number or decreased the probability, necessarily)

Question 8

The more extreme the initial probability, the less it will change in the light of evidence.
True or False

Answer 8

True

Question 9

Why is it important to avoid using more than one of the correlated findings?

Answer 9

Avoid using more than one of the correlated findings, otherwise your probability estimate will be too extreme

Question 10

How do you calculate likelihood ratios if you can’t find them in diagnostic literature?

Answer 10

Calculate from sensitivity and specificity

Question 11

Explain sensitivity?

Answer 11

• in patients who have the disease, the probability that the test will be positive
• numerator on the likelihood ratio
• true +ive
• the probability that you’d see certain evidence if your hypothesis were true

Question 12

Explain Specificity?

Answer 12

• in patients who don’t have the disease, the probability that the test will be negative
• true -ive
• the complement of the probability that you’d see the same evidence if your hypothesis were false

Question 13

False +ive?

Answer 13

1 minus true -ive (denominator)

Question 14

LR ratio?

Answer 14

LR= sensitivity/ 1-specificity
(finding is present)

Question 15

LR -ive ratio?

Answer 15

LR-ive = 1-sensitivity/ specificity
(finding is absent)

Question 16

The more uncertain you initially are, generally the less evidence you need.
Why?

Answer 16

You have not firmly set your mind on one diagnosis and are flexible in considering other evidence.
it takes more/better evidence to go from, say 1% to 5% than from 51% to 55% certainty (though both increase the probability by 4%)

Question 17

LR values and there ~ change in probability (%)

Answer 17

LR 10= +45%
LR 5= +30%
LR 2= +15%
LR 1= 0%
LR 0.5 (1/2)= -15%
LR 0.2 (1/5)= -30%
LR 0.1 (1/10)- -45%

Question 18

What is Positive Predictive Value (PPV)?

Answer 18

Probability that the disease is present given that a test was positive

Question 19

What is Negative Predictive Value (NPV)?

Answer 19

Probability that a disease is absent given that a test was negative

Question 20

What are pathognomonic findings?

Answer 20

• pathos= disease; gnomon= indicator
• findings that, if present, strongly increases the probability of a condition (High LR+)
• not necessarily particularly sensitive but highly specific

Question 21

What are “Sine qua non” findings?

Answer 21

• without which it could not be
• findings which, if absent, strongly decreases the probability of a condition (Low LR-, close to zero)