AUTHOR=Cui Chen , Mu Fei , Tang Meng , Lin Rui , Wang Mingming , Zhao Xian , Guan Yue , Wang Jingwen 

TITLE=A prediction and interpretation machine learning framework of mortality risk among severe infection patients with pseudomonas aeruginosa

JOURNAL=Frontiers in Medicine

VOLUME=Volume 9 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/medicine/articles/10.3389/fmed.2022.942356

DOI=10.3389/fmed.2022.942356

ISSN=2296-858X

ABSTRACT=Pseudomonas aeruginosa is an important opportunistic pathogen that causes a wide range of acute and chronic infections. Early identification of the risk factors is urgently needed for severe infection patients with Pseudomonas Aeruginosa. There has been no detailed relevant machine learning based research, and no known research has focused on exploring relationships between key risk clinical variables and their tendency to affect clinical outcome of patients. In this study, we collected 571 severe infection with Pseudomonas Aeruginosa patients admitted to the Xijing Hospital of the Fourth Military Medical University from January 2010, to July 2021. Basic clinical information, clinical signs and symptoms, laboratory indicators, bacterial culture and drug related were recorded. A model based on machine learning algorithm of XGBoost was applied to predict mortality risk prediction of pseudomonas aeruginosa infection in severe patients. The performance of XGBoost model (AUROC=0.94±0.01, AUPRC=0.94±0.03) was greater than the performance of support vector machine (AUROC=0.90±0.03, AUPRC=0.91±0.02) and random forest (AUROC=0.93±0.03, AUPRC=0.89±0.04). This study also aimed to interpret the model and explore the impact of clinical variables and their interaction relationship. The interpretation analysis highlighted the effects of age, high-alert drugs and the number of drug varieties. Further stratification clarified the necessity of different treatment for severe infection for different populations.