Scholarlink

Background <p>Acute carbon monoxide poisoning (ACOP) commonly results in delayed encephalopathy (DE). This study aimed to develop a predictive model to estimate the odds of DE in patients with ACOP.</p> Methods <p>Clinical data were retrospectively collected from 102 ACOP patients admitted to a single institution between January 1, 2020, and June 30, 2024. The least absolute shrinkage and selection operator regression was used to identify influencing factors, and then a prediction model based on logistic regression was constructed. The discrimination, calibration and clinical practicality of the model were respectively evaluated according to the area under the receiver operating characteristic (ROC) curve (AUC), calibration curve, decision curve analysis (DCA) and clinical impact curve (CIC).</p> Results <p>A total of 102 ACOP patients were included in the study, of who 15 developed DE. Length of stay, comorbidities (e.g., hypertension, diabetes), carboxyhemoglobin, and hyperbaric oxygen therapy after discharge were influencing factors. The four-variable model displayed high prediction ability with the AUC of 0.933 [95% confidence interval (CI): 0.877–0.991]. The calibration curve (bootstraps = 1,000) showed good calibration with a brier score of 0.066 (95%CI: 0.030–0.102). In addition, the DCA and CIC also showed good clinical practicality.</p> Conclusions <p>The prediction model constructed in this study showed good discrimination, calibration and clinical practicality, which could help medical personnel identify DE in patients with ACOP.</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

Study on the construction of a predictive model for delayed encephalopathy after acute carbon monoxide poisoning

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年2月7日

查看全文
  • Huankai Gong,
  • Chengjin You,
  • Xin Chen,
  • Yao Zhu,
  • Yanchun Wang

摘要

Background

Acute carbon monoxide poisoning (ACOP) commonly results in delayed encephalopathy (DE). This study aimed to develop a predictive model to estimate the odds of DE in patients with ACOP.

Methods

Clinical data were retrospectively collected from 102 ACOP patients admitted to a single institution between January 1, 2020, and June 30, 2024. The least absolute shrinkage and selection operator regression was used to identify influencing factors, and then a prediction model based on logistic regression was constructed. The discrimination, calibration and clinical practicality of the model were respectively evaluated according to the area under the receiver operating characteristic (ROC) curve (AUC), calibration curve, decision curve analysis (DCA) and clinical impact curve (CIC).

Results

A total of 102 ACOP patients were included in the study, of who 15 developed DE. Length of stay, comorbidities (e.g., hypertension, diabetes), carboxyhemoglobin, and hyperbaric oxygen therapy after discharge were influencing factors. The four-variable model displayed high prediction ability with the AUC of 0.933 [95% confidence interval (CI): 0.877–0.991]. The calibration curve (bootstraps = 1,000) showed good calibration with a brier score of 0.066 (95%CI: 0.030–0.102). In addition, the DCA and CIC also showed good clinical practicality.

Conclusions

The prediction model constructed in this study showed good discrimination, calibration and clinical practicality, which could help medical personnel identify DE in patients with ACOP.