To the Editor.—The article "Gauging the Severity of Surgical Sepsis"1 deserves several comments. First, any article that deals with a linear function of values to which a "severity score" is attached, even if these values are the "squares" of other values, should quote the work of J. P. Krischer.2-4 He is, in my opinion, the world authority on these scoring systems and would, I am sure, point out the numerous faults in this article as well.
In the "Patients and Methods" section, Stevens gave an example from which it was difficult to determine exactly how the score was calculated. First, he stated that the patient was in septic shock and required vasopressors, but he did not mention the dosages and assigned the patient a score of 4. According to Stevens' own schedule, vasopressors are required for patients with scores of 3 through 5, and he seems to
Koss N. The Correct Use of Statistical Techniques. Arch Surg. 1984;119(7):866. doi:10.1001/archsurg.1984.01390190102024