What question did this study set out to answer?

This study aims to identify risk factors associated with short-term mortality in patients with acute respiratory failure and develop a nomogram for predictions.

March 26, 2026

708: Construction of a Nomogram for Predicting Short-Term Mortality in Acute Respiratory Failure Patients

Key Points

This study aims to identify risk factors associated with short-term mortality in patients with acute respiratory failure and develop a nomogram for predictions.
Retrospective analysis of clinical data from the MIMIC-IV database
Patients diagnosed with acute respiratory failure were divided into training and validation groups
Risk factors selected using backward stepwise selection method of Cox regression
Nomogram constructed and compared to existing scoring systems using various statistical indices.
Identified nine independent risk factors for short-term mortality including age, WBC, glucose, temperature, metastatic solid tumor, and others
Nomogram showed strong predictive performance with C-indices of 0.782 and 0.749 for training and validation sets respectively
Area under the curve (AUC) for nomogram was 0.811 in training and 0.790 in validation cohort
Nomogram significantly outperformed traditional scoring systems indicated by positive NRI and IDI values.

Abstract

Introduction: This study aimed to identify short-term mortality risk factors for patients with acute respiratory failure (ARF) the MIMIC-IV database, construct a prognostic nomogram and evaluate its predictive performance compared to conventional scoring systems. Methods: Clinical data from patients diagnosed with ARF were retrospectively collected from the MIMIC-IV database and randomly divided into training and validation groups. The variables were selected via the backwards stepwise selection method of Cox regression, and a nomogram was constructed. The nomogram was compared with acute physiology score III (APSIII), simplified acute physiology scores II (SAPS II) and oxford acute severity of illness score (OASIS) model via the C-index, area under the receiver operating characteristic curve (ROC), net reclassification index (NRI), integrated discrimination improvement index (IDI), decision curve analysis (DCA). Results: A total of 559 patients were included. The study identified nine independent risk factors: age (HR: 1.022,95% CI: 1.008-1.036,P=0.002), WBC (HR: 1.060,95% CI:1.033-1.086,P< 0.001), glucose levels (HR:1.002,95% CI: 1.001-1.004,P=0.003), temperature (HR: 0.544, 95% CI: 0.430-0.689,P< 0.001), metastatic solid tumor (HR: 2.138, 95% CI: 1.045-4.372,P=0.037), malignant cancer (HR: 2.455, 95% CI: 1.456-4.138,P< 0.001), diabetes without chronic complications (HR: 0.288, 95% CI: 0.157-0.807,P< 0.001), cerebrovascular disease (HR: 2.156, 95% CI: 1.180-3.940,P=0.012), dementia (HR: 2.23, 95% CI: 1.132-4.392,P=0.020). The nomogram demonstrated strong discriminative performance with C-indices of 0.782 and 0.749 in the training and validation sets, respectively. The AUC for the Training and Validation cohorts were 0.811 (APS III: 0.652; SAPS II: 0.672; OASIS: 0.624) and 0.790 (APS III: 0.634; SAPS II: 0.652; OASIS: 0.609), respectively. The nomogram also significantly outperformed traditional scoring systems, as evidenced by positive NRI and IDI values. Conclusions: The newly developed nomogram exhibits superior predictive capability to traditional scoring systems (APS III, SAPS II and OASIS scores), offering clinicians a practical and reliable tool for accurately assessing short-term mortality risks in ICU patients with ARF.

Bookmark

Cite This Study

Wang et al. (Sun,) studied this question.

synapsesocial.com/papers/69c4cd3efdc3bde4489194f9 https://doi.org/https://doi.org/10.1097/01.ccm.0001184828.48722.de

Bookmark