APACHE II-ECG scoring system: A novel and strong predictor of in-hospital mortality for patients treated in intensive...

Some novel electrocardiographic (ECG) parameters, such as T-wave peak to T-wave end duration (Tp-Te) and Tp-Te/Q-T interval (QT) ratios, have recently been found to be associated with cardiac ischemia and effective in predicting ventricular arrhythmias and mortality. This study examined the association between ECG repolarization parameters and mortality in intensive care unit (ICU) patients. A total of 232 ICU patients were retrospectively categorized as survivors or nonsurvivors retrospectively. Laboratory, demographic, and ECG parameters were compared between the groups. A novel ECG score was measured using the QT interval, Tp-Te, and Tp-Te/QT ratio upon admission to the ICU. We compared the ECG score, Acute Physiologic and Chronic Health Evaluation II (APACHE II)-score, and APACHE II-ECG scores (the combination of APACHE II and ECG score) regarding mortality using a biostatistical program. The mean age of the 232 patients was 69.96 ± 18.01 years, and 49.1% were male. The nonsurvivor group was significantly older and had higher ECG, APACHE II, and APACHE II-ECG scores. Multivariate Cox regression analysis revealed that higher levels of all 3 scores were independent risk factors for mortality ([hazard ratio, HR (95% CI): 1.847 (1.305-2.615), P = .001], [HR (95%CI): 1.146 (1.071-1.225), P < .001], and [HR (95% CI): 1.181 (1.117-1.249), P < .001], respectively). Receiver operating curve analysis of these scoring systems for predicting mortality in the ICU revealed a stronger predictive value for the APACHE II-ECG score (AUC [95% CI]: 0.872 [0.824-0.919], P < .001, sensitivity: 88.7%, specificity: 73.3%). Kaplan-Meier survival analysis revealed the superiority of the APACHE II-ECG score in predicting the survival of ICU patients (log rank chi-square: 80.366, P < .001). Our study suggests combining ECG repolarization parameters with APACHE II score offers a new, more robust system for stronger mortality prediction in ICU patients.