1. Screening Tests

Serial

Screening strategy is a special usage of (parallel/serial) testing.

- Sensitivity

- FN

A screening test should have high (sensitivity/specificity) in order to minimize (FP/FN).

Prevalence (improves PPV)

With serial testing, the first testing is used to increase the _______ and uses the second test on a smaller group.

Only the positives

Using serial testing, who is retested using a second, confirmatory test?

F ( ideally, screening programs focus on the populations of animals at high risk, but not yet sick (pre-clinical, apparently healthy animals).)

T/F: With screening programs, your target population is clinical affected animals.

Pre-clinical phase

What phase of disease is best to use screening tests

Early (secondary) detection of disease.

(REMEMBER: There are 3 types of prevention. Screening is an example of secondary prevention).

Why screen a "healthy" (no clinical signs yet) population?

- false negatives

- disease has no good tx

- too many FP (resulting in invasive diagnostics to tx)

- overdiagnosis bias

- lead-time bias

5 examples of when screening can be harmful

overdiagnosis bias

What is the term that is defined as - identification of cases that would never have gotten sick or died of the disease, yet causes extra diagnostics with needless expense.

lead-time bias

What is the term that is defined as - when early treatment has no benefit on the expected time to death of the animal, yet false looks like a survival benefit

(31.5 / (31.5 + 18.5) x 100=) 63%

Case:

- In your clinic, you are working on screening dogs for cranial cruciate ligament rupture. Test A is a physical exam, which has 70% sensitivity and 90% specificity. Test B is taking radiographs, which has 90% sensitivity and 80% specificity. Assuming that the responses to tests A and B are independent, and assuming a pre-test probability of disease of 50%, calculate the net sensitivity.

(49 / (49 + 1) x 100 —>) 98%

Case:

- In your clinic, you are working on screening dogs for cranial cruciate ligament rupture. Test A is a physical exam, which has 70% sensitivity and 90% specificity. Test B is taking radiographs, which has 90% sensitivity and 80% specificity. Assuming that the responses to tests A and B are independent, and assuming a pre-test probability of disease of 50%, calculate the net specificity.

(31.5 / (31.5 + 1) x 100 —>) 97%

Case:

- In your clinic, you are working on screening dogs for cranial cruciate ligament rupture. Test A is a physical exam, which has 70% sensitivity and 90% specificity. Test B is taking radiographs, which has 90% sensitivity and 80% specificity. Assuming that the responses to tests A and B are independent, and assuming a pre-test probability of disease of 50%, calculate the resulting PPV.

(49 / (49 + 18.5) x 100 —> ) 73%

Case:

- In your clinic, you are working on screening dogs for cranial cruciate ligament rupture. Test A is a physical exam, which has 70% sensitivity and 90% specificity. Test B is taking radiographs, which has 90% sensitivity and 80% specificity. Assuming that the responses to tests A and B are independent, and assuming a pre-test probability of disease of 50%, calculate the resulting NPV.

The "lead time"

Screening should be timed in order to increase what?

Treatment

Ideally, the screening provides an opportunity for effective ___________.