Jing Qi1, Yishu Tang1, Huaizheng Liu1, Zheren Dai1, Kefu Zhou1, Tianyi Zhang1, Jun Liu1, Chuanzheng Sun2. 1. Department of Emergency, The Third Xiangya Hospital of Central South University, 138 Tongzipo Road, Changsha, 410013, China. 2. Department of Emergency, The Third Xiangya Hospital of Central South University, 138 Tongzipo Road, Changsha, 410013, China. Electronic address: sunchuanzheng@csu.edu.cn.
Abstract
BACKGROUND: The global volume of gastrointestinal surgery has increased steadily. However, there is still a lack of studies focused on the risk factors for post-gastrointestinal resection surgery patients in the intensive care units. METHODS: Post gastrointestinal resection surgery patient data were collected from the Medical Information Mart for Intensive Care (MIMIC-III) database and divided into training set and validation set, then analyzed by Univariate and multiple logistic regression. RESULTS: 795 patients were finally enrolled in our cohort. Multiple logistic regression showed that age (1.029 [1.006-1.053]), temperature (0.337 [0.207-0.547]), respiratory rate (1.133 [1.053-1.218]), mean arterial pressure (1.204 [1.039-1.396]), lactate (1.288 [1.112-1.493]), BUN (1.025 [1.010-1.040]) and vasopressor use (4.777 [2.499-9.130]) were independent factors associated with in-hospital mortality. Our new predicted nomogram achieved a better accuracy than SOFA score, SAPS-Ⅱ score, APACHE-Ⅲ score, and Elixhauser score. CONCLUSION: Our nomogram model could well predict in-hospital mortality for post-GI resection surgery patients receiving intensive care.
BACKGROUND: The global volume of gastrointestinal surgery has increased steadily. However, there is still a lack of studies focused on the risk factors for post-gastrointestinal resection surgery patients in the intensive care units. METHODS: Post gastrointestinal resection surgery patient data were collected from the Medical Information Mart for Intensive Care (MIMIC-III) database and divided into training set and validation set, then analyzed by Univariate and multiple logistic regression. RESULTS: 795 patients were finally enrolled in our cohort. Multiple logistic regression showed that age (1.029 [1.006-1.053]), temperature (0.337 [0.207-0.547]), respiratory rate (1.133 [1.053-1.218]), mean arterial pressure (1.204 [1.039-1.396]), lactate (1.288 [1.112-1.493]), BUN (1.025 [1.010-1.040]) and vasopressor use (4.777 [2.499-9.130]) were independent factors associated with in-hospital mortality. Our new predicted nomogram achieved a better accuracy than SOFA score, SAPS-Ⅱ score, APACHE-Ⅲ score, and Elixhauser score. CONCLUSION: Our nomogram model could well predict in-hospital mortality for post-GI resection surgery patients receiving intensive care.