Personalized sepsis mortality prediction: An interpretable machine learning nomogram

Lulu Weng

doi:10.1016/j.clinsp.2026.100872

2026-02-01

DOI：10.1016/j.clinsp.2026.100872

原文链接

查看原文

摘要

En 中文

目的：脓毒症仍是ICU患者的主要死亡原因，需要精确的预后工具以实现最佳管理。本研究旨在开发和验证一个基于可解释机器学习的列线图，用于预测脓毒症患者的院内死亡率，以指导临床决策。方法：这项回顾性队列研究纳入了2019年1月至2024年12月间入住ICU的407例成人脓毒症患者。患者被随机分为训练组（n=284）和验证组（n=123）。采用基于LASSO的多变量logistic回归构建预测列线图。结果：院内死亡率的7个独立预测因素为：年龄、腹部感染、血管加压药需求、WBC、BNP、APACHE II评分和机械通气支持。列线图显示出良好的判别能力，训练组的AUC为0.900，验证组为0.796。校准曲线、决策曲线分析和SHAP解释证实了模型具有良好的性能和临床实用性。结论：该新型基于机器学习的列线图提供了一个实用的、可解释的风险评估工具，用于早期死亡预测，有望通过增强临床理解和及时干预改善重症脓毒症患者的预后。

Keyword:

Nomogram

Prediction model

Risk stratification

Intensive care unit

SHAP

Mortality prediction

AI总结

对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。

期刊

Clinics

IF:

2.4

论文数:

219

被引数:

5.1K

机构

shanghai jiao tong university

学者数:

15.7W

论文数: 11.7W

被引数: 159

huzhou university

学者数:

1.1K

论文数: 479

被引数: 0

universiti malaya

学者数:

4.2K

论文数: 1.8K

被引数: 0

查看更多机构

引用论文

Reporting and handling missing values in clinical studies in intensive care units

INTENSIVE CARE MEDICINE

IF21.2

2013-05-18

105

OAAI

Vesin, Aurelien; Azoulay, Elie; Ruckly, Stephane; Vignoud, Lucile; Rusinova, Katerina; Benoit, Dominique; Soares, Marcio; Azeivedo-Maia, Paulo; Abroug, Fekri; Benbenishty, Judith; Timsit, Jean Francois

Impact of the timing of invasive mechanical ventilation in patients with sepsis: a multicenter cohort study

CRITICAL CARE

IF9.3

2024-09-09

OAAI

Kim, Gyungah; Oh, Dong Kyu; Lee, Su Yeon; Park, Mi Hyeon; Lim, Chae-Man

Novel cell death program leads to neutrophil extracellular traps 新型细胞死亡程序导致中性粒细胞胞外陷阱

JOURNAL OF CELL BIOLOGY

IF6.4

2007-01-08

2.7K

OAAI

Fuchs, Tobias A.; Abed, Ulrike; Goosmann, Christian; Hurwitz, Robert; Schulze, Ilka; Wahn, Volker; Weinrauch, Yvette; Brinkmann, Volker; Zychlinsky, Arturo

The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure SOFA (脓毒症相关器官衰竭评估) 评分描述器官功能障碍/衰竭

Intensive Care Medicine

IF0

1996-07-01

PREAI

J. -L. Vincent; R. Moreno; J. Takala; S. Willatts; A. De Mendonça; H. Bruining; C. K. Reinhart; P. M. Suter; L. G. Thijs

Essential biochemistry and physiology of (NT-pro)BNP

EUROPEAN JOURNAL OF HEART FAILURE

IF10.8

2004-03-20

461

PREAI

Hall, C

Course and outcome of Staphylococcus aureus bacteraemia:: a retrospective analysis of 308 episodes in a Swiss tertiary-care centre

CLINICAL MICROBIOLOGY AND INFECTION

IF8.5

2006-04-01

132

OAAI

Kaech, C; Elzi, L; Sendi, P; Frei, R; Laifer, G; Bassetti, S; Fluckiger, U

Predicting 30-days mortality for MIMIC-III patients with sepsis-3: a machine learning approach using XGboost

JOURNAL OF TRANSLATIONAL MEDICINE

IF7.5

2020-12-07

254

OAAI

Hou, Nianzong; Li, Mingzhe; He, Lu; Xie, Bing; Wang, Lin; Zhang, Rumin; Yu, Yong; Sun, Xiaodong; Pan, Zhengsheng; Wang, Kai

Brain natriuretic peptide for prediction of mortality in patients with sepsis: a systematic review and meta-analysis

CRITICAL CARE

IF9.3

2012-05-06

OAAI

Wang, Fei; Wu, Youping; Tang, Lu; Zhu, Weimin; Chen, Feng; Xu, Tao; Bo, Lulong; Li, Jinbao; Deng, Xiaoming

Epidemiology of severe sepsis

VIRULENCE

IF5.4

2013-12-11

1.0K

OAAI

Mayr, Florian B.; Yende, Sachin; Angus, Derek C.

The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3)脓毒症和脓毒症休克定义的第三次国际共识 (Sepsis-3)

JAMA-JOURNAL OF THE AMERICAN MEDICAL ASSOCIATION

IF55

2016-02-23

1.8W

OAAI

Singer, Mervyn; Deutschman, Clifford S.; Seymour, Christopher Warren; Shankar-Hari, Manu; Annane, Djillali; Bauer, Michael; Bellomo, Rinaldo; Bernard, Gordon R.; Chiche, Jean-Daniel; Coopersmith, Craig M.; Hotchkiss, Richard S.; Levy, Mitchell M.; Marshall, John C.; Martin, Greg S.; Opal, Steven M.; Rubenfeld, Gordon D.; van der Poll, Tom; Vincent, Jean-Louis; Angus, Derek C.

查看更多内容