Question

Suppose there is an inexpensive screening test ($20) with a lot of false positives (individuals test positive but don’t have cancer), and a more expensive test ($500) that is always accurate. Consider two strategies for screening 1000 people. Strategy A. Everyone gets the inexpensive test, which is then repeated on everyone who tests positive the first time. Those who test positive both times get the expensive test. Strategy B. Everyone gets the inexpensive test, and all who test positive get the expensive test. You will need this additional information: 20% of the population tests positive the first time on the inexpensive test. Of those who test positive, 50% would test positive again if that test is repeated. Of those who test positive both times, 20% have cancer. Of those who test positive the first time but would not the second time, only 2% have cancer. [Note: this information is sufficient to determine how many cancers will be detected under each strategy. Strategy A will detect all of the cancers that exist among those who test positive twice by the inexpensive test. Strategy B will detect all of the cancers that exist among those who test positive the first time.] a. Using these numbers only, what are the average costs per cancer found under Strategy A and Strategy B? b. Compared with a third strategy which is no screening and no cancer found, what are the appropriately defined ICERs (Incremental Cost-Effectiveness Ratios) for Strategy A and Strategy B, measuring effectiveness by number of cancers found? c. Comment on the difference between your answers to part a and part b. d. Could Strategy B ever be considered cost-effective?

Answer 1