In Reply.—The comments by Drs Boutwell and Brown are most appropriate. I consulted with Dr Murray A. Katz, the author of the reference in question, and he concurs that (1) his proposed formula was intended for use when the sensitivity of a test is, or approaches, 1.00, and (2) I had misplaced the parentheses in the denominator of the formula.
I am comforted only by their agreement that the conclusions reached by Evans et al and us remain appropriate. I have certainly learned from this overly enthusiastic excursion into statistical territory and hope that our readers will also benefit from the perspicacity of Drs Boutwell and Brown.
Dr Boutwell also points out quite well some problems previously highlighted by Doubilet et al,1 with respect to the various interpretations of "costeffectiveness" in medicine. As a minimum, cost savings should be coupled with a demonstrable equal (or better) health outcome.