AUTHOR=Hu Linwang , Yu Jie , Deng Jian , Zhou Hong , Yang Feng , Lu Xiaohang 

TITLE=Development of nomogram to predict in-hospital death for patients with intracerebral hemorrhage: A retrospective cohort study

JOURNAL=Frontiers in Neurology

VOLUME=Volume 13 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2022.968623

DOI=10.3389/fneur.2022.968623

ISSN=1664-2295

ABSTRACT=Aim: To investigate the association between systemic immune-inflammation (SII) and risk of in-hospital death for intracerebral hemorrhage (ICH) patients in the intensive care units (ICU), and to further develop a prediction model related to SII in predicting the risk of in-hospital death for patients with ICH.
Methods: In this retrospective cohort study, we included 1,176 ICH patients from the Medical Information Mart for Intensive Care III (MIMIC-III) database. All patients were randomly assigned to the training group for the construction of nomogram and testing group for the validation of nomogram based on a ratio of 8:2. Predictors were screened by the least absolute shrinkage and selection operator (LASSO) regression analysis. Multivariate Cox regression analysis was used to investigate the association between SII and in-hospital death of ICH patients in the ICU, and develop a predictive model in predicting the in-hospital death risk for ICU patients with ICH. Receiver operator characteristic curve was used to assess the predicting performance of constructed nomogram. 
Results: 232 ICH patients died and 708 survived in the training group. LASSO regression showed some predictors, including white blood cell count, glucose, blood urea nitrogen, SII, Glasgow Coma Scale, age, heart rate, mean artery pressure, red blood cell, bicarbonate, red blood cell distribution width, liver cirrhosis, respiratory failure, renal failure, malignant cancer, vasopressor and mechanical ventilation. A prediction model integrating these predictors was established. The area under the curve of the nomogram was 0.810 in the training group and 0.822 in the testing group, indicating that this nomogram might have a good performance.
Conclusion: SII was associated with death risk of in-hospital for ICH patients in the ICU. A nomogram for death risk of in-hospital for ICH patients in the ICU was developed and validated.