Scholarlink

<p>To construct a nomogram risk prediction model for disorders associated with sepsis comorbid with chronic obstructive pulmonary disease (COPD) and to test the calibration and discrimination of the model. Patients with sepsis and COPD in the MIMIC-III database were divided into a model group and a validation group. Certain risk factors of high importance were obtained via univariate and multivariate logistic regression analysis in each group. These risk factors of clinical importance were selected for constructing the 30-day mortality prediction nomogram for the population. 1077 sepsis patients with COPD were extracted from the MIMIC-III database and then retrospectively analyzed. Six independent risk factors for critical illness (age, MAP, respiratory rate, SAPS-II score, INR, and BUN) could be obtained from optimal subset regression analysis. The optimal subset regression coefficients for the 6 factors were 1.03 and <i>P</i> &lt; 0.05; 0.98 and <i>P</i> &lt; 0.05; 1.09 and <i>P</i> &lt; 0.05; 1.03 and <i>P</i> &lt; 0.05; 1.20 and <i>P</i> &lt; 0.01; and 1.01 and <i>P</i> &lt; 0.05, respectively. In the model cohort, the AUC was 0.742 (95% CI 0.704–0.780), and in the validation cohort, the AUC was 0.758 (95% CI 0.676–0.840). The internal and external calibration curves basically overlapped with the standard curve, which meant that the model was good at calibration. The nomogram had good discriminating ability and calibration, which is important for providing a reliable prediction of patients with sepsis and COPD and provides a basis for clinical decisions and timely diagnosis and treatment.</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

Prognostic value of a mortality risk nomogram model for sepsis combined with COPD patients based on the MIMIC-III database

  • 期刊论文

  • 发表日期：

    2026年2月21日

查看全文
  • Fenmei Shi,
  • Xuelian Huang,
  • Jun Peng,
  • Kelin Zhang

摘要

To construct a nomogram risk prediction model for disorders associated with sepsis comorbid with chronic obstructive pulmonary disease (COPD) and to test the calibration and discrimination of the model. Patients with sepsis and COPD in the MIMIC-III database were divided into a model group and a validation group. Certain risk factors of high importance were obtained via univariate and multivariate logistic regression analysis in each group. These risk factors of clinical importance were selected for constructing the 30-day mortality prediction nomogram for the population. 1077 sepsis patients with COPD were extracted from the MIMIC-III database and then retrospectively analyzed. Six independent risk factors for critical illness (age, MAP, respiratory rate, SAPS-II score, INR, and BUN) could be obtained from optimal subset regression analysis. The optimal subset regression coefficients for the 6 factors were 1.03 and P < 0.05; 0.98 and P < 0.05; 1.09 and P < 0.05; 1.03 and P < 0.05; 1.20 and P < 0.01; and 1.01 and P < 0.05, respectively. In the model cohort, the AUC was 0.742 (95% CI 0.704–0.780), and in the validation cohort, the AUC was 0.758 (95% CI 0.676–0.840). The internal and external calibration curves basically overlapped with the standard curve, which meant that the model was good at calibration. The nomogram had good discriminating ability and calibration, which is important for providing a reliable prediction of patients with sepsis and COPD and provides a basis for clinical decisions and timely diagnosis and treatment.