To the Editor.
—In a recent issue of JAMA, Dr Le Gall and colleagues1 contributed two intensive care risk-adjustment models, newly calibrated for severe sepsis, to the growing list of tools available for critical care outcome prediction.The fact that the original MPM and SAPS II models required recalibration to fit patients with severe sepsis raises a question about their use in generating risk-adjusted outcomes figures for critical care audit. This need for recalibration indicates that the original models inadequately account for the severity of illness of patients with severe sepsis. Can ICUs with high adjusted mortality rates based on the original versions of MPM and SAPS II defend their standard of care by pointing to their case mix of severely septic patients?On page 649, the authors ask, "[H]ow are the probability models useful for a therapeutic trial?" Presumably, the authors would have investigators report treatment efficacy from
Rubenfeld GD. Customized Probability Models for Severe Sepsis: Correction and Clarification. JAMA. 1995;274(11):872. doi:10.1001/jama.1995.03530110034025