Quiz 1

Question 1

Test threshold

Answer 1

the probability below which a diagnostic test will not be ordered or performed because the possibility of the diagnosis is so remote

Question 2

Treatment threshold

Answer 2

the probability above which a diagnostic test will not be ordered or performed because the possibility of the diagnosis is so great that immediate treatment is indicated (open/compound fracture, ECG, MI, smoker, obese)

Question 3

Critique on Dx Tests

Answer 3

• Why do these studies?
• What are the best study designs?
• Is the new test better than others available?
• Are studies on the diagnosis valid?
• Does the test help distinguish patients with or without the disorder?
• Would the test benefit my patient?

Question 4

Sensitivity

Answer 4

Ability to detect patients who actually have the condition

Question 5

SnOUT

Answer 5

higher Sensitivity = more likely that a Negative finding rules out the disorder.

Question 6

Sensitivity calculation

Answer 6

Sensitivity = d / (b + d)

Question 7

Specificity

Answer 7

ability to detect those who do not have the condition

Question 8

SpIN

Answer 8

higher Specificity = more likely that a Positive finding rules in the disorder.

Question 9

Specificity calculation

Answer 9

Specificity = a / (a + c)

Question 10

Positive predictive value

Answer 10

• Ability of the test to correctly determine % of people with disorder from all of the people with positive test results
• Patients w disorder that tests positive/patients who test positive.

Question 11

Negative predictive value

Answer 11

• Ability of the test to correctly determine % of people without the disorder from all of the people with a negative test result
• Patients’ w/o disorder that test negative/patients who test negative.

Question 12

Positive likelihood ratio

Answer 12

Likelihood that a positive test was observed in a person with the disorder vs a person without the disorder (>10)
- > 10 is large and conclusive change, 1-2 negligible change pre to post test
- LR+ = sensitivity / (1 – specificity)

Question 13

Negative likelihood ratio

Answer 13

Likelihood that a negative test result is observed in a person with the disorder vs a person without the disorder (<1)
a. <0.1 large and conclusive, 0.5-1 negligible change pre to post test
b. LR- = (1 – specificity) / sensitivity

Question 14

Likelihood ratios

Answer 14

i. Likelihood ratios have values > 0
ii. Likelihood ratios = 1 is no better than chance that your Dx indicates the presence of the disorder

Question 15

Validity (2x2 square)

Answer 15

x axis (disease present, disease absent)
y axis (positive test, negative test)

Question 16

Limitations to sensitivity and specificity

Answer 16

• Only indicate test performance in individuals whose status are known (have condition or not).
• Reduce the choice information about a test to 2 options based on threshold to be either positive or negative.
  o Difficult when a test has a range of scores may indicate a range of severity of disorder. Inefficient to continuously calculate for each level of severity.
• Why can’t you have both 100% sensitivity and specificity? They are inversely related. As one increases the other tends to decrease. (True positives vs negatives).

Question 17

Advantage of likelihood ratios

Answer 17

• Ratios can be calculated for all levels of condition.
• Not dependent on prevalence of condition in the population.
• Ratios can be applied by an individual patient whereas sensitivity/specificity or +/- predictive values refer to groups of people.