Min Woo Kang1, Navdeep Tangri2, Soie Kwon1, Lilin Li1,3, Hyeseung Lee1, Seung Seok Han1, Jung Nam An4, Jeonghwan Lee5, Dong Ki Kim1,6, Chun Soo Lim5,6, Yon Su Kim1,6, Sejoong Kim7, Jung Pyo Lee5,6. 1. Department of Internal Medicine, Seoul National University Hospital, Republic of Korea, Seoul. 2. Department of Internal Medicine, University of Manitoba, Manitoba, Canada. 3. Department of Intensive Care Unit, Yanbian University Hospital, Yanji, China. 4. Department of Internal Medicine, Hallym University Sacred Heart Hospital, Anyang, Republic of Korea. 5. Department of Internal Medicine, Seoul National University Boramae Medical Center, Seoul, Republic of Korea. 6. Department of Internal Medicine, Seoul National University College of Medicine, Seoul, Republic of Korea. 7. Department of Internal Medicine, Seoul National University Bundang Hospital, Seongnam, Republic of Korea.
Abstract
Background: Predicting the risk of death in patients admitted to the critical care unit facilitates appropriate management. In particular, among patients who are critically ill, patients with continuous RRT (CRRT) have high mortality, and predicting the mortality risk of these patients is difficult. The purpose of this study was to develop models for predicting the mortality risk of patients on CRRT and to validate the models externally. Methods: A total of 699 adult patients with CRRT who participated in the VolumE maNagement Under body composition monitoring in critically ill patientS on CRRT (VENUS) trial and 1515 adult patients with CRRT in Seoul National University Hospital were selected as the development and validation cohorts, respectively. Using 11 predictor variables selected by the Cox proportional hazards model and clinical importance, equations predicting mortality within 7, 14, and 28 days were developed with development cohort data. Results: The equation using 11 variables had area under the time-dependent receiver operating characteristic curve (AUROC) values of 0.75, 0.74, and 0.73 for predicting 7-, 14-, and 28-day mortality, respectively. All equations had significantly higher AUROCs than the Sequential Organ Failure Assessment (SOFA) and Acute Physiology and Chronic Health Evaluation II (APACHE II) scores. The 11-variable equation was superior to the SOFA and APACHE II scores in the integrated discrimination index and net reclassification improvement analyses. Conclusions: The newly developed equations for predicting CRRT patient mortality showed superior performance to the previous scoring systems, and they can help physicians manage patients.
Background: Predicting the risk of death in patients admitted to the critical care unit facilitates appropriate management. In particular, among patients who are critically ill, patients with continuous RRT (CRRT) have high mortality, and predicting the mortality risk of these patients is difficult. The purpose of this study was to develop models for predicting the mortality risk of patients on CRRT and to validate the models externally. Methods: A total of 699 adult patients with CRRT who participated in the VolumE maNagement Under body composition monitoring in critically ill patientS on CRRT (VENUS) trial and 1515 adult patients with CRRT in Seoul National University Hospital were selected as the development and validation cohorts, respectively. Using 11 predictor variables selected by the Cox proportional hazards model and clinical importance, equations predicting mortality within 7, 14, and 28 days were developed with development cohort data. Results: The equation using 11 variables had area under the time-dependent receiver operating characteristic curve (AUROC) values of 0.75, 0.74, and 0.73 for predicting 7-, 14-, and 28-day mortality, respectively. All equations had significantly higher AUROCs than the Sequential Organ Failure Assessment (SOFA) and Acute Physiology and Chronic Health Evaluation II (APACHE II) scores. The 11-variable equation was superior to the SOFA and APACHE II scores in the integrated discrimination index and net reclassification improvement analyses. Conclusions: The newly developed equations for predicting CRRT patient mortality showed superior performance to the previous scoring systems, and they can help physicians manage patients.
Authors: M Capuzzo; V Valpondi; A Sgarbi; S Bortolazzi; V Pavoni; G Gilli; G Candini; G Gritti; R Alvisi Journal: Intensive Care Med Date: 2000-12 Impact factor: 17.440
Authors: Marylou Cárdenas-Turanzas; Joe Ensor; Chris Wakefield; Karen Zhang; Susannah Kish Wallace; Kristen J Price; Joseph L Nates Journal: J Crit Care Date: 2012-07-02 Impact factor: 3.425
Authors: Min Woo Kang; Jayoun Kim; Dong Ki Kim; Kook-Hwan Oh; Kwon Wook Joo; Yon Su Kim; Seung Seok Han Journal: Crit Care Date: 2020-02-06 Impact factor: 9.097
Authors: Andrew S Allegretti; David J R Steele; Jo Ann David-Kasdan; Ednan Bajwa; John L Niles; Ishir Bhan Journal: Crit Care Date: 2013-06-20 Impact factor: 9.097