血管活性药物评分对ICU合并心房颤动患者 预后的预测价值：基于MIMIC-IV数据库 的回顾性研究

Predictive Value of the Vasoactive-Inotropic Score for Prognosis in ICU Patients with Atrial Fibrillation: A Retrospective Study Based on the MIMIC-IV Database

摘要: 目的：探讨重症监护室(ICU)合并心房颤动(AF)患者入ICU早期血管活性–正性肌力药物评分(VIS)与短期及中长期全因病死率的独立关联及剂量–反应关系。方法：采用多中心回顾性队列研究设计，提取大样本数据库MIMIC-IV中合并AF的ICU患者。暴露变量为入ICU前24 h内最大VIS，按四分位数(Q1~Q4)进行分组。主要结局为30 d、90 d及365 d全因病死率。采用Kaplan-Meier法绘制生存曲线，构建多因素Cox比例风险回归模型评估VIS与病死率的关联，应用限制性立方样条(RCS)拟合非线性关系，并进行亚组及交互作用分析。结果：研究共纳入MIMIC-IV队列2687例患者。K-M生存分析显示，随着VIS分位等级的升高，患者30 d、90 d及365 d累积全因病死率均显著增加(Log-rank P均<0.001)。多因素Cox回归分析表明，在充分校正混杂因素后，VIS升高是各随访节点病死率增加的独立危险因素，且呈极其显著的剂量递增趋势(P for trend均<0.001)。RCS分析进一步揭示VIS与病死率之间存在显著的非线性关系，死亡风险在低分值区间(VIS < 15分)随剂量陡增，随后进入平缓的平台期。亚组分析表明，VIS在未合并心力衰竭或慢性肾脏病的患者中，对病死率的预测效能更为显著(交互作用P < 0.05)。结论：入ICU早期VIS升高是合并AF的ICU患者短、中、长期死亡风险增加的独立危险因素，二者呈显著的非线性(阈值)关联。VIS在既往无严重心肾基础疾病的患者中预后分层价值更优，可作为该特殊高危人群早期临床评估与医疗决策的简便量化工具。

Abstract: Objective: To investigate the independent association and dose-response relationship between the early vasoactive-inotropic score (VIS) upon intensive care unit (ICU) admission and short- to long-term all-cause mortality among ICU patients complicated by atrial fibrillation (AF). Methods: In this retrospective cohort study, data of ICU patients with AF were extracted from the MIMIC-IV database. The primary exposure was the maximum VIS within the first 24 hours of ICU admission, categorized into quartiles (Q1~Q4). The primary outcomes were 30-, 90-, and 365-day all-cause mortality. Survival curves were plotted using the Kaplan-Meier method. Multivariable Cox proportional hazards regression models were constructed to evaluate the association between VIS and mortality. Restricted cubic splines (RCS) were employed to fit nonlinear relationships, followed by subgroup and interaction analyses. Results: A total of 2687 patients from the MIMIC-IV database were included. Kaplan-Meier analysis demonstrated a stepwise increase in the 30-, 90-, and 365-day cumulative all-cause mortality across ascending VIS quartiles (all log-rank P < 0.001). After fully adjusting for confounders, multivariable Cox regression confirmed that an elevated VIS was an independent risk factor for increased mortality at all follow-up endpoints, exhibiting a highly significant dose-response trend (all P for trend < 0.001). RCS analysis further revealed a distinct nonlinear relationship between VIS and mortality, where the mortality risk rose steeply in the lower score range (VIS < 15) before reaching a plateau. Subgroup analyses indicated that the prognostic value of VIS was more pronounced in patients without concurrent heart failure or chronic kidney disease (P for interaction < 0.05). Conclusions: An elevated early VIS is an independent risk factor for increased short- to long-term mortality among ICU patients complicated by AF, characterized by a distinct nonlinear (threshold) relationship. VIS offers superior prognostic stratification value for individuals without severe baseline cardiorenal diseases, serving as an accessible and quantitative tool to guide early clinical assessment and medical decision-making in this specific high-risk population.