Nomogram for predicting mandatory icu admission after gastrectomy for gastric cancer

Abstract

Background: This study aimed to identify risk factors that were associated with mandatory intensive care unit (ICU) admission after gastrectomy for gastric cancer. We then employed these risk factors to construct and validate a nomogram for predicting mandatory ICU admission after gastrectomy, which may identify those who require ICU indeed and improve ICU utilization. Methods: A number of 999 gastric cancer patients undergoing gastrectomy from January 2010 to June 2019 were included in the retrospective study. Forty-three patients were classified into mandatory ICU admission groups, and the remaining 956 patients were allocated into the no need for ICU admission group. The candidate variables, including patient demographic characteristics, preoperative laboratory tests and surgical variables, were compared between the two groups. We then carried out univariate and multivariate logistic regression analyses to find out risk factors for mandatory ICU admission. In order to develop the predictive model, we used Akaike information criterion (AIC) to select risk factors via a step-down backward process from the multivariate regression model. Results: A number of risk factors for mandatory ICU admission were identified and subsequently used to build the nomogram: age [odds ratio (OR), 1.03; 95% CI, 1.00–1.07; P=0.031], ASA status (III–IV vs. I– II: OR,1.74; 95% CI, 0.88–3.46; P=0.114), tumor size (OR, 1.28; 95% CI, 1.08–1.51; P=0.004), estimated blood loss (OR, 1.001; 95% CI, 1.000–1.001; P=0.082) as well as intraoperative transfusion (Yes vs. No: OR, 3.82; 95% CI, 1.87–7.82; P<0.001). C-index of the nomogram was 0.800, indicating good discrimination. Both Calibration curve and Hosmer-Lemeshow goodness-of-fit tests (P=0.128) showed that there was a high degree of agreement between the prediction and actual outcome. Conclusions: A nomogram to predict mandatory ICU admission after gastrectomy for gastric cancer was constructed and validated. Clinicians could apply this predictive model to improve usage of limited ICU resources effectively.