Predictive performance of the SOFA 2.0 score for in-hospital mortality in patients with heat stroke: a multicenter, data-driven subphenotype study

SOFA 2.0评分对中暑患者院内死亡率的预测性能:一项多中心、数据驱动的亚表型研究

阅读:1

Abstract

OBJECTIVE: Heat stroke can progress rapidly, and the risk of in-hospital mortality increases once multiple organ dysfunction develops. Early risk stratification is therefore clinically important, yet comparative evidence across commonly used severity scores in heat stroke remains limited. METHODS: We conducted a multicenter retrospective cohort study of patients admitted with a first diagnosis of heat stroke to two tertiary hospitals in China between 2013 and 2023. The recalibrated SOFA 2.0 score (SOFA2), original SOFA, Modified Early Warning Score, National Early Warning Score, and Heat Stroke Severity Score were calculated using the first available data within 24 h of admission. In-hospital death was the primary outcome, with discharge alive treated as a competing event. Cumulative incidence functions and Fine-Gray models were used to assess risk gradients, and unsupervised clustering based on early clinical and laboratory features was applied to identify clinical subtypes. RESULTS: Among 292 patients (mean age 29.8 ± 14.9 years), 24 (8.2%) died during hospitalization. The cumulative incidence of in-hospital death increased stepwise across SOFA2 quartiles (Gray test, P < 0.001), whereas separation across original SOFA quartiles was less distinct. Higher SOFA2 scores were associated with an increased risk of mortality risk, with spline analyses indicating a generally monotonic risk increase. Two major clinical subtypes were identified; in the higher-risk subtype identified by data-driven clustering, SOFA2 showed numerically consistent discrimination and stable net benefit trends; however, these subtype-specific findings should be interpreted cautiously. CONCLUSIONS: SOFA2 may provide an early, continuous representation of in-hospital mortality risk in patients with heat stroke, although external validation is required.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。