However, it remains unclear whether the additional data needed to compute ICU scores improves mortality prediction for critically ill patients compared to the simpler EM scores. Clinicians require firm models for mortality prediction in critically ill patients, and multiple scoring systems have been developed in both the ICU 7-10 and the EM 11-13. In general, EM-based scoring systems employ a handful of variables that are readily available on all patients, while ICU scoring systems employ a larger number of variables that are frequently available only in those patients that are critically ill (e.g., arterial blood gas measurements). It remains unknown which of these scores perform best in patients of ICU. While it seems intuitive that scores using a larger number of data inputs would perform better than more niggardly scoring systems, simpler scores may actually outperform more complex scores when the population has been well-defined. For example, the original Acute Physiology and Chronic Health Evaluation (APACHE) score had 34 variables and when reduced to 12 for APACHE II, it performed better in aggregate than did its predecessor
However, it remains unclear whether the additional data needed to compute ICU scores improves mortality prediction for critically ill patients compared to the simpler EM scores. Clinicians require firm models for mortality prediction in critically ill patients, and multiple scoring systems have been developed in both the ICU 7-10 and the EM 11-13. In general, EM-based scoring systems employ a handful of variables that are readily available on all patients, while ICU scoring systems employ a larger number of variables that are frequently available only in those patients that are critically ill (e.g., arterial blood gas measurements). It remains unknown which of these scores perform best in patients of ICU. While it seems intuitive that scores using a larger number of data inputs would perform better than more niggardly scoring systems, simpler scores may actually outperform more complex scores when the population has been well-defined. For example, the original Acute Physiology and Chronic Health Evaluation (APACHE) score had 34 variables and when reduced to 12 for APACHE II, it performed better in aggregate than did its predecessor