Objective: To assess the performance of Acute Physiology and Chronic Health Evaluation II (APACHE II) and Simplified Acute Physiology Score II (SAPS II) in Thai critically ill patients.
Material and Method: Prospective observational cohort study conducted between July 1, 2004 and October 31, 2005 in the Intensive Care Unit (ICU) of Songklanagarind Hospital, an 800-beds tertiary referral university teaching hospital.
Results: One thousand three hundred sixteen patients were enrolled. There were 310 deaths (23.6%) at hospital discharge. APACHE II and SAPS II predicted hospital mortality 30.5±28.2 and 30.5±29.8 respectively. Both models showed excellent discrimination. The discrimination of APACHE II was better than SAPS II (0.911 and 0.888, p<0.001). However, both systems presented a poor calibration. The Hosmer-Lemeshow goodness-of-fit H and C statistics were 66.59 and 66.65 of APACHE II (p<0.001) and 54.01 and 71.44 of SAPS II (p<0.001).
Conclusion: APACHE II provided better discrimination than SAPS II, but both models showed poor calibration in over predicting mortality in our ICU patients. Customized or new severity scoring systems should be developed for critically ill patients in Thailand.

Keywords: Severity of illness, Intensive care, Mortality prediction, Acute Physiology and Chronic Health Evaluation (APACHE II), Simplified Acute Physiology Score (SAPS II)