To the Editor.
—We write to correct a misunderstanding that appeared in the Editorial by Drs Abraham and Raffin1 discussing results form the recent Phase III rhIL-1ra Sepsis Syndrome Study Group.2 The Editorial states that the APACHE [Acute Physiology and Chronic Health Evaluation] III prediction system used prospectively in this trial2 to risk-stratify patients could not be used to identify those who might benefit because it "does not directly reflect the biochemical alterations present in a patient with sepsis." The original JAMA article3 describing the risk prediction method used in this interleukin-1 receptor antagonist trial indicates that the predicted risk of mortality model was specifically designed for use in sepsis trials to account for the physiological profile found in patients who met the sepsis syndrome entry criteria. This was accomplished by using an independent database and determining the predictive importance of a wide range of physiological,
Knaus WA, Harrell FE, Wagner DP. Defining Sepsis. JAMA. 1994;272(24):1901. doi:10.1001/jama.1994.03520240029035