Department of President's Office, Youjiang Medical University for Nationalities, Baise, China.
Department of Comprehensive Internal Medicine, The Affiliated Tumor Hospital of Guangxi Medical University, Nanning, China.
Intern Emerg Med. 2024 Jan;19(1):49-58. doi: 10.1007/s11739-023-03436-z. Epub 2023 Oct 5.
This study aims to develop and validate a prognostic nomogram that accurately predicts the short-term survival rate of cirrhotic patients with acute kidney damage (AKI) upon ICU admission. For this purpose, we examined the admission data of 3060 cirrhosis patients with AKI from 2008 to 2019 in the MIMIC-IV database. All included patients were randomly assigned to derivation and validation cohorts in a 7:3 ratio. The derivation cohort used the least absolute shrinkage and selection operator (LASSO) regression model to identify independent predictors of AKI. A prognostic nomogram was constructed via multivariate logistic regression analysis in the derivation cohort and subsequently verified in the validation cohort. Nomogram's discrimination, calibration, and clinical utility were evaluated using the C-index, calibration plot, and decision curve analysis (DCA). A total of 2138 patients were enrolled in the derivation cohort, with a median follow-up period of 15 days, a median survival time of 41 days, and a death rate of 568 patients (26.6%). The cumulative survival rates at 15 and 30 days were 75.8% and 57.5%, respectively. The results of the multivariate analysis indicated that advanced AKI stage, use of vasoactive drugs, advanced age, lower levels of ALB, lower mean sBp, longer INR, and longer PT were all independent risk factors that significantly influenced the all-cause mortality of cirrhosis patients with AKI (all p < 0.01). The C-indices for the derivation and the validation cohorts were 0.821 (95% CI 0.800-0.842) and 0.831 (95% CI 0.810-0.852), respectively. The model's calibration plot demonstrated high consistency between predicted and actual probabilities. Furthermore, the DCA showed that the nomogram was clinically valuable. Therefore, the developed and internally validated prognostic nomogram exhibited favorable discrimination, calibration, and clinical utility in forecasting the 15-day and 30-day survival rates of cirrhosis patients with AKI upon admission to the ICU.
本研究旨在开发和验证一个预测模型，该模型可以准确预测 ICU 收治的肝硬化合并急性肾损伤（AKI）患者的短期生存率。为此，我们检查了 2008 年至 2019 年 MIMIC-IV 数据库中 3060 例肝硬化合并 AKI 患者的入院数据。所有纳入的患者按 7:3 的比例随机分配到推导和验证队列中。推导队列使用最小绝对值收缩和选择算子（LASSO）回归模型来确定 AKI 的独立预测因素。通过推导队列中的多变量逻辑回归分析构建预后列线图，并在验证队列中进行验证。使用 C 指数、校准图和决策曲线分析（DCA）评估列线图的区分度、校准和临床实用性。共有 2138 例患者纳入推导队列，中位随访时间为 15 天，中位生存时间为 41 天，死亡率为 568 例（26.6%）。15 天和 30 天的累积生存率分别为 75.8%和 57.5%。多因素分析结果表明，晚期 AKI 分期、血管活性药物的使用、年龄较大、ALB 水平较低、平均 sBp 较低、INR 较长和 PT 较长均为肝硬化合并 AKI 患者全因死亡率的独立危险因素（均 P < 0.01）。推导队列和验证队列的 C 指数分别为 0.821（95%CI 0.800-0.842）和 0.831（95%CI 0.810-0.852）。模型校准图显示预测概率与实际概率具有高度一致性。此外，DCA 表明该列线图具有临床价值。因此，开发并内部验证的预后列线图在预测肝硬化合并 AKI 患者 ICU 收治后 15 天和 30 天的生存率方面具有良好的区分度、校准度和临床实用性。
