[Skip to Content]
[Skip to Content Landing]
November 11, 1983

Construction of Clinical Algorithms

Author Affiliations

Providence Hospital Southfield, Mich
Kupat Cholim Canaan Safed, Israel
Systems Development EKS, Inc Detroit

JAMA. 1983;250(18):2471. doi:10.1001/jama.1983.03340180033017

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.—  While not fully sharing the enthusiasm for clinical algorithms of Dr Margolis in his review of their uses (1983;249:627), we would like to suggest a novel approach to their construction. For many years, one of us (F.N.B.) has been developing a computer program that "learns" the approach of the clinical experts with which it interacts. The format of the interaction is structured by the mentor physician in the language of the usual clinical data and summary interpretation of individual cases, a process at which, as noted in Dr Levinson's accompanying COMMENTARY (1983;249:607), expert physicians tend to excel. After the program has had sufficient experience, it acquires independent decision-making capability for even the most complex problems. The eventual clinical importance of this decision-making capability is not known, but to students of medical practice, the resultant net of logical relationships between data points and learned conclusion states (the "judgment