June 25, 1997

The Negative Side of Cost-effectiveness Analysis-Reply

Author Affiliations

University of Wisconsin Madison
Harvard University Boston, Mass
University of Minnesota Minneapolis
McMaster University Hamilton, Ontario
Rutgers University New Brunswick, NJ
University of Washington Seattle

JAMA. 1997;277(24):1933. doi:10.1001/jama.1997.03540480033027

In Reply.  —Drs Stinnett and Mullahy provide a helpful illustration of the problems that occur when standard cost-effectiveness calculations lead to a negative C/E ratio. Negative ratios occur in a cost-effectiveness analysis when 1 of the 2 alternatives being compared dominates the other, ie, 1 of the alternatives is both less costly and more effective than the other. Trouble arises when part of the legitimate range of uncertainty about the cost difference or the difference in effectiveness between the 2 alternatives includes cases in which 1 alternative dominates the other.Stinnett and Mullahy rightly point out that computing a CI for the C/E ratio does not circumvent the basic problem that it is inappropriate to compute a C/E ratio in this circumstance and advocate the alternative of using net benefit rather than a C/E ratio.From the terminology, it would seem that Stinnett and Mullahy are advocating the use of cost-benefit analysis,