This Article  • References  • Full text PDF  • Send to a friend  • Save in My Folder  • Save to citation manager  • Permissions  Citing Articles  • Citing articles on Web of Science (2)  • Contact me when this article is cited  Related Content  • Similar articles in JAMA  Social Bookmarking   What's this?

Robert L. Obenchain, PhD
Lilly Research Laboratories Indianapolis, Ind

Jose A. Sacristán, MD
Lilly SA Madrid, Spain

JAMA. 1997;277(24):1932-1933.

	Since this article does not have an abstract, we have provided the first 150 words of the full text PDF and any section headings.

In Reply.

—While we sympathize with many of the points raised by Drs Stinnett and Mullahy, we cannot agree that the magnitude of a negative C/E ratio "conveys no useful information." This is an important part of what you "see" when you plot the outcome of a health economics study as a point (C,E on the cost-effectiveness plane.¹

Polar coordinates (radius and angle) are helpful to understand the technical issues here (Figure). Stinnett and Mullahy point out that the C/E ratio is an imperfect measure of angle; any given ratio corresponds to 2 angles that differ by exactly 180°. Outcomes in quadrants II and IV generate negative C/E ratios. An outcome in quadrant IV means that the experimental treatment dominates the standard treatment on long-range average, whereas an outcome in quadrant II indicates the opposite (ie, the standard treatment dominates the experimental treatment).

Quadrant II (0,0) Quadrant III AC Quadrant I . . . [Full Text PDF of this Article]

CiteULike    Connotea    Del.icio.us    Digg    Reddit    Technorati    Twitter     What's this?