Scholarlink

Background <p>Exacerbations of chronic obstructive pulmonary disease (COPD) deteriorate patient outcomes and impose substantial burdens on healthcare systems and families. Given this, the present study aimed to develop a prediction model for in-hospital mortality in patients with acute exacerbation of COPD (AECOPD) using readily accessible clinical variables, with the goal of guiding clinical interventions.</p> Methods <p>A total of 878 consecutive AECOPD patients (807 non-death cases, 71 in-hospital death cases) were prospectively enrolled. Patients were randomly divided into a training set (<i>n</i> = 616) and a validation set (<i>n</i> = 262) at a 7:3 ratio. Logistic regression analysis was performed on the training set to identify factors influencing in-hospital mortality, followed by construction of a nomogram. The model’s discrimination, calibration, and clinical applicability were evaluated using the area under the receiver-operating characteristic curve (AUC), calibration curves, Hosmer–Lemeshow test, decision curve analysis (DCA), and clinical impact curve (CIC).</p> Results <p>The in-hospital mortality rate of AECOPD patients was 8.1%. Independent risk factors for in-hospital mortality included non-invasive mechanical ventilation, invasive mechanical ventilation, disease duration &gt; 10&#xa0;years, severe condition at admission, and comorbid heart failure (all <i>P</i> &lt; 0.05). The nomogram showed an AUC of 0.909 (95% CI: 0.875–0.942) in the training set and 0.841 (95% CI: 0.742–0.939) in the validation set. Calibration curves and Hosmer–Lemeshow test (training set: χ<sup>2</sup> = 14.13, <i>P</i> = 0.12; validation set: χ<sup>2</sup> = 7.83, <i>P</i> = 0.55) confirmed good fit. DCA demonstrated higher net benefits of the model than “treat-all” or “treat-none” strategies within threshold probabilities of 5.0%–42.5% (training set) and 5.0%–42.0% (validation set). At the optimal threshold (0.082), the model’s sensitivity, specificity, and accuracy were 0.885, 0.789, 0.797 (training set) and 0.632, 0.753, 0.744 (validation set), respectively.</p> Conclusion <p>The nomogram based on mechanical ventilation, disease duration, admission condition, cor pulmonale, and heart failure exhibits excellent predictive performance for in-hospital mortality in AECOPD patients, providing a valuable tool for clinical decision-making.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Construction of a risk prediction model for in-hospital mortality in patients with acute exacerbation of chronic obstructive pulmonary disease

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月21日

查看全文
  • Jieyun Zhu,
  • Qiuyun Song,
  • Yin Shen,
  • Chunli Huang,
  • Dongzan Pan,
  • QiaoyanWang,
  • Zhaoqiang Cai,
  • Changguang Ye,
  • Zhao Lu

摘要

Background

Exacerbations of chronic obstructive pulmonary disease (COPD) deteriorate patient outcomes and impose substantial burdens on healthcare systems and families. Given this, the present study aimed to develop a prediction model for in-hospital mortality in patients with acute exacerbation of COPD (AECOPD) using readily accessible clinical variables, with the goal of guiding clinical interventions.

Methods

A total of 878 consecutive AECOPD patients (807 non-death cases, 71 in-hospital death cases) were prospectively enrolled. Patients were randomly divided into a training set (n = 616) and a validation set (n = 262) at a 7:3 ratio. Logistic regression analysis was performed on the training set to identify factors influencing in-hospital mortality, followed by construction of a nomogram. The model’s discrimination, calibration, and clinical applicability were evaluated using the area under the receiver-operating characteristic curve (AUC), calibration curves, Hosmer–Lemeshow test, decision curve analysis (DCA), and clinical impact curve (CIC).

Results

The in-hospital mortality rate of AECOPD patients was 8.1%. Independent risk factors for in-hospital mortality included non-invasive mechanical ventilation, invasive mechanical ventilation, disease duration > 10 years, severe condition at admission, and comorbid heart failure (all P < 0.05). The nomogram showed an AUC of 0.909 (95% CI: 0.875–0.942) in the training set and 0.841 (95% CI: 0.742–0.939) in the validation set. Calibration curves and Hosmer–Lemeshow test (training set: χ2 = 14.13, P = 0.12; validation set: χ2 = 7.83, P = 0.55) confirmed good fit. DCA demonstrated higher net benefits of the model than “treat-all” or “treat-none” strategies within threshold probabilities of 5.0%–42.5% (training set) and 5.0%–42.0% (validation set). At the optimal threshold (0.082), the model’s sensitivity, specificity, and accuracy were 0.885, 0.789, 0.797 (training set) and 0.632, 0.753, 0.744 (validation set), respectively.

Conclusion

The nomogram based on mechanical ventilation, disease duration, admission condition, cor pulmonale, and heart failure exhibits excellent predictive performance for in-hospital mortality in AECOPD patients, providing a valuable tool for clinical decision-making.