Scholarlink

Background <p>The association between the stress hyperglycemia ratio (SHR), which represents the degree of acute stress hyperglycemic status, and the risk of mortality in cerebrovascular disease patients in the intensive care unit (ICU) remains unclear. This study aims to investigate the predictive ability of SHR for in-hospital mortality in critically ill cerebrovascular disease patients and to assess its potential to enhance existing predictive models.</p> Methods <p>We extracted data from the Medical Information Mart for Intensive Care (MIMIC-IV) database for patients diagnosed with cerebrovascular disease and used Cox regression to assess the association between SHR and mortality. To investigate the nature of this association, we applied restricted cubic spline analysis to determine if it is linear. The predictive ability of SHR for mortality risk was evaluated using receiver operating characteristic (ROC) curves and the C-index.</p> Results <p>We included a total of 2,461 patients, with a mean age of 70.55 ± 14.59 years, and 1,221 (49.61%) being female. Cox regression analysis revealed that SHR was independently associated with both in-hospital mortality (per standard deviation (SD) increase: hazard ratio (HR) 1.35, 95% confidence interval (CI) 1.23–1.48) and ICU mortality (per SD increase: HR 1.37, 95% CI 1.21–1.54). The risk of death increased in an approximately linear fashion when SHR exceeded 0.77–0.79. Subgroup analysis indicated the association was more pronounced in non-diabetic individuals. Additionally, incorporating SHR into existing models improved both discrimination and reclassification performance.</p> Conclusion <p>SHR serves as an independent risk factor for in-hospital mortality in cerebrovascular disease patients in the ICU. Adding SHR to existing models enhances their predictive performance, offering clinical value in the identification of high-risk patients.</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

Assessment of stress hyperglycemia ratio to predict all-cause mortality in patients with critical cerebrovascular disease: a retrospective cohort study from the MIMIC-IV database

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年2月7日

查看全文
  • Yuwen Chen,
  • Jian Xu,
  • Fan He,
  • An’an Huang,
  • Jie Wang,
  • Bingchen Liu,
  • Qucheng Wei

摘要

Background

The association between the stress hyperglycemia ratio (SHR), which represents the degree of acute stress hyperglycemic status, and the risk of mortality in cerebrovascular disease patients in the intensive care unit (ICU) remains unclear. This study aims to investigate the predictive ability of SHR for in-hospital mortality in critically ill cerebrovascular disease patients and to assess its potential to enhance existing predictive models.

Methods

We extracted data from the Medical Information Mart for Intensive Care (MIMIC-IV) database for patients diagnosed with cerebrovascular disease and used Cox regression to assess the association between SHR and mortality. To investigate the nature of this association, we applied restricted cubic spline analysis to determine if it is linear. The predictive ability of SHR for mortality risk was evaluated using receiver operating characteristic (ROC) curves and the C-index.

Results

We included a total of 2,461 patients, with a mean age of 70.55 ± 14.59 years, and 1,221 (49.61%) being female. Cox regression analysis revealed that SHR was independently associated with both in-hospital mortality (per standard deviation (SD) increase: hazard ratio (HR) 1.35, 95% confidence interval (CI) 1.23–1.48) and ICU mortality (per SD increase: HR 1.37, 95% CI 1.21–1.54). The risk of death increased in an approximately linear fashion when SHR exceeded 0.77–0.79. Subgroup analysis indicated the association was more pronounced in non-diabetic individuals. Additionally, incorporating SHR into existing models improved both discrimination and reclassification performance.

Conclusion

SHR serves as an independent risk factor for in-hospital mortality in cerebrovascular disease patients in the ICU. Adding SHR to existing models enhances their predictive performance, offering clinical value in the identification of high-risk patients.