Try it
PSM · Screening: combining two tests
Sample itemTwo independent tests for the same disease have sensitivities of 80% and 70%. In a screening programme both tests are done on every person, and anyone positive on either test is called positive. What is the effect on the combined sensitivity and specificity?
Original practice item, not a recalled exam question. Nothing here is recorded.
The answer and the reasoning
C. Sensitivity 94%, and overall specificity falls below that of either test
Interpreting either positive result as positive is parallel testing. A diseased person is missed only if both tests are negative: 0.2 x 0.3 = 0.06, so the combined sensitivity is 94%. More false positives accrue, so specificity falls. Series testing (both must be positive) would give a sensitivity of 0.8 x 0.7 = 56% with higher specificity.
Option by option
A. Sensitivity 56%, and overall specificity rises above that of either test
A sensitivity of 56% and higher specificity describes series testing, where both tests must be positive.
B. Sensitivity 94%, and overall specificity rises above that of either test
Parallel testing increases the chance of a false-positive result from either test, so specificity cannot rise.
C. Sensitivity 94%, and overall specificity falls below that of either test (correct)
Parallel testing gains sensitivity (1 - 0.06 = 94%) at the cost of specificity.
D. Sensitivity 75%, and overall specificity is unchanged
Sensitivity is not the average of the two tests; independent tests in parallel combine multiplicatively for the misses.
Source note: Original NextPrep sample item — pending expert review. This item is original practice material awaiting independent expert review, so confirm important facts in a standard textbook.
More PSM questions
- BiostatisticsOf 200 people who truly have a disease, a screening test is positive in 180. What is the…
- Environmental healthAfter one hour of contact time, what residual free chlorine level is generally recommend…
- EpidemiologyWhich study design is the most efficient for investigating risk factors of a rare disease?
- BiostatisticsIn hypothesis testing, a Type I error is: