This article is only available in the PDF format. Download the PDF to view the article, as well as its associated figures and tables.
To the Editor.—In his letter, Dr Merrill points out that there are various arbitrary ways to look at data. The use of circle curves, as done by Van De Water et al (Arch Surg 112:679-683, 1977), is just as arbitrary as the hyperbolic curves suggested by Dr Merrill. Both present boundary conditions that group the patients in reference to the data and the clinical condition that are correlated. I think that neither is an adequate description for the data. When setting up boundary conditions, it is necessary to look at the extremes and see whether they are reasonable. For instance, for the circle boundaries, in the normal domain it is possible to have a point with a very high systolic pressure and very low pulsatile flow, or conversely, of very low systolic pressure and a very high pulsatile flow, both conditions being absurd. In the hyperbolic curve model, a
FINSETH F. Noninvasive Assessment of the Peripheral Vascular System. Arch Surg. 1978;113(1):111. doi:10.1001/archsurg.1978.01370130113023