6T: EBM5: Screening and Diagnostic Tests

Question 1

3 types of prevention

Answer 1

Primary: Prevent disease from occurring
Secondary: Identify disease process early, to prevent sequelae or initiate treatment
Tertiary: Minimize mortality and morbidity

Question 2

What is the goal of screening?

Answer 2

To detect disease early in order to prevent mortality and morbidity

Question 3

Lead Time Bias

Answer 3

when extra follow up time is added to the screened group. In reality, the screening test has no impact on the course of the disease, but it appears to be protective because of the increased follow up.

Question 4

Length Bias

Answer 4

Diseases can progress slow or fast. Diseases that progress slower are going to be easier to detect in their pre-clinical phase, because their pre-clinical phase is longer.

Question 5

Overdiagnosis

Answer 5

A type of length bias pertaining to slow growing cancers. Overdiagnosed cancers can be cured, but don’t need to be because the cancer will not go on to cause symptoms/death. If mortality rates remain stable despite increased screening then this is a sign of overdiagnosis.

Question 6

Essential features of a “screenable” disease

Answer 6

• disease is common or of unusual severity
• disease has a known natural history and lag phase between possible detection and irreversible outcome
• disease is treatable

Question 7

Essential features of a screening test

Answer 7

• acceptable to the public
• risk/benefit profile is favorable
• economically and technically feasible
• accurate and repeatable
• acceptable levels of sensitivity and specificity

Question 8

test validity

Answer 8

ability to distinguish between who has a disease and who does not

Question 9

test sensitivity

Answer 9

the ability of a test to correctly identify those who have the disease
- pro: few false negatives
- con: worry about false positives
- used to rule out disease
= TP/(TP + FN)

Question 10

test specificity

Answer 10

the ability of a test to correctly identify those who do not have a disease
- pro: few false positives
- con: worry about false negatives
- used to rule in disease
= TN/(TN + FP)

Question 11

unimodal vs. bimodal test distribution

Answer 11

unimodal curves have one peak

bimodal curves have two peaks

Question 12

Positive predictive value (PPV)

Answer 12

The probability that disease is present given that the test is positive
= TP/(TP + FP)

Question 13

Negative predictive value (NPV)

Answer 13

Probability that disease is not present given that the test is negative
= TN/(TN + FN)

Question 14

What does prevalence effect? (sensitivity, specificity, NPV and PPV)

Answer 14

Prevalence effects PPV and NPV, but not specificity or sensitivity. For diseased with higher prevalence, we would use a test with a higher PPV

Question 15

define:

a. pretest probability

b. posttest probability

Answer 15

a. prevalence is also called pretest probability. It is the probability of disease before the test result is known
b. also called the predictive value. It is the probability of disease after test result is known

Question 16

Likelihood Ratio

Answer 16

expresses how many times more (or less) likely test result is to be found in diseased (compared to non-diseased) people.
If LR = 1, then test has no discriminatory value
If LR+ > 10 (or equal to), then test has high diagnostic value
If LR-

Question 17

Sequential Testing

Answer 17

If test 1 = negative, stop
if test 1 = positive, continue to next test
- low sensitivity
- high specificity

Question 18

Simultaneous Testing

Answer 18

• for screening an large group quickly
• use 2 tests at the same time.
  If both are negative, then total result is negative
  If either test is positive, then overall test is positive
• high sensitivity
• low specificity