Clinical Decision Making

Question 1

We will treat a patient if the probability is above the

Answer 1

Treatment threshold

Question 2

We will perform diagnostic testing if the probability is above the

Answer 2

Testing threshold

Question 3

What is the PICO?

Answer 3

Patient, Intervention, Comparison, Outcome

Question 4

In therapeutic decision making, what are three questions we want to ask ourselves?

Answer 4

1. ) How great is the benefit?
2. ) How great is the risk?
3. ) How sure am I of these numbers?

Question 5

To asses how great is the benefit, we want to look at?

Answer 5

RRR vs ARR, and also NNT

Question 6

To asses how great the risk is, we wanted to compare the

Answer 6

RRI vs ARI, and also the NNH

Question 7

If a treatment reduces bad outcomes from 4% to 3%, what is the

1. ) ARR
2. ) RRR
3. ) NNT

Answer 7

1. ) 1%
2. ) 25%
3. ) 100

Question 8

If the treatment increases bad outcomes from 10% to 15%, what is the

1. ) ARI
2. ) RRI
3. ) NNH

Answer 8

1. ) 5%
2. ) 50%
3. ) 20

Question 9

What diseases are the most common causes of death in women ages 20-40?

Answer 9

Malignancy (Leukemia, Lymphoma, and brain), and HIV

Question 10

Ratio between the likelihood of a particular test result in those with the disease to the likelihood of the same test result in those without the disease

Answer 10

Likelihood Ratio

Question 11

How do we calculate the LR for a positive test (LR+)?

Answer 11

Likelihood of positive test in diseased / Likelihood of positive test in non-diseased

Question 12

Another way to calculate LR+ is?

Answer 12

LR+ = Sensitivity / 1 - specificity

Question 13

What is the LR (-)?

Answer 13

Likelihood of neg test in diseased / likelihood of neg test in non-diseased

Question 14

Another way to calculate LR (-) is?

Answer 14

LR (-) = 1 - sensitivity / specificity

Question 15

A useless LR+ or LR (-) is an LR ~ to

Answer 15

1

Question 16

A perfect LR+ is

Answer 16

LR+ = infinity. But anything > 10 is good

Question 17

What is a perfect LR(-)?

Answer 17

LR(-) = 0

But anything < 0.1 is good

Question 18

How do we calculate post-test probability from LR?

Answer 18

Pretest odds X LR = Post-test odds

Question 19

A positive test makes a BIG difference in

Answer 19

Post-test probability

Question 20

A negative test makes a modest difference in

Answer 20

Post-test probability

Question 21

There is a high potential for over-diagnosis of

Answer 21

Prostate cancer

Question 22

Typically the largest group in testing is

Answer 22

True negatives

Question 23

Typically, the smallest group in testing is

Answer 23

False negatives

Question 24

The only group with a potential medical benefit from mammogram screening is the

-Usually a small group

Answer 24

True positives

Question 25

An observational study shows that people with cancer detected by screening have many more low-stage (1-2) and many fewer high-stage (3-4) cancers than those presenting with symptoms. This is called

Answer 25

Improved stage distribution

Question 26

A randomized trial of screening shows that people randomized to the screening arm live much longer after the diagnosis of cancer than people randomized to usual care. This is called

Answer 26

Improved case survival rate