Unveiling the role of stress hyperglycemia in predicting mortality for critically ill hemorrhagic stroke patients: insights from MIMIC-IV

Abstract

Background: Hemorrhagic stroke (HS), including intracerebral hemorrhage (ICH) and subarachnoid hemorrhage (SAH), is associated with high mortality and morbidity. Stress hyperglycemia ratio (SHR), reflecting acute glycemic responses relative to baseline glucose levels, has been linked to poor outcomes in critical illnesses. However, research on its prognostic significance in HS patients admitted to the intensive care unit (ICU) is limited. This study aims to assess the association between SHR and all-cause mortality (ACM) in critically ill HS patients.

Methods: Patients diagnosed with HS were extracted from the Medical Information Mart for Intensive Care-IV (MIMIC-IV) database using ICD-9/10 codes. SHR was calculated as [admission glucose (mg/dL)/(28.7 × HbA1c (%) - 46.7)]. Patients were stratified into tertiles. Primary outcomes were ICU, in-hospital, 30-day, 90-day, 180-day, and 1-year mortality. Cox regression and restricted cubic splines (RCS) evaluated the dose-response relationship between SHR and ACM. Kaplan-Meier (K-M) analysis assessed survival across tertiles, with subgroup analysis and interaction tests for effect modification.

Results: The study included 1,749 patients, with a median age of 68 years (IQR: 57-79), and 53.2% were male. The observed mortality rates were 10.6% in the ICU, 15.2% in-hospital, 19.6% at 30 days, 24.2% at 90 days, 27.8% at 180 days, and 31.7% at 1 year. Multivariate Cox regression analysis indicated that elevated SHR was independently associated with increased ACM at 30 days (adjusted hazard ratio [aHR]: 1.41; 95% confidence interval [CI]: 1.10-1.81; P = 0.006), 90 days (aHR: 1.33; 95% CI: 1.08-1.65; P = 0.008), and 1 year (aHR: 1.27; 95% CI: 1.05-1.54; P = 0.014). RCS analysis demonstrated a linear association between SHR and ACM, with no evidence of non-linearity. Subgroup analysis revealed consistent associations across various patient characteristics.

Conclusion: SHR is significantly associated with ACM in critically ill patients with HS, supporting its potential role as a prognostic marker for risk stratification and guiding clinical management. Incorporating SHR into routine risk assessment may facilitate early identification of high-risk patients, enabling timely interventions and improved outcomes.

Keywords: MIMIC-IV database; all-cause mortality; hemorrhagic stroke; prognosis; stress hyperglycemia ratio.

Figure 1

Flowchart illustrating the selection process for patients diagnosed with hemorrhagic stroke (HS) from the MIMIC-IV database.

Figure 2

LASSO penalized regression analysis for identifying key variables associated with all-cause mortality in patients with hemorrhagic stroke. (A) Coefficient profiles of variables plotted against the log(lambda) sequence. (B) Ten-fold cross-validation for tuning parameter selection in the LASSO model. The vertical dashed line indicates the optimal lambda value that minimizes the partial likelihood deviance.

Figure 3

Kaplan-Meier survival curves for all-cause mortality (ACM) across tertiles of stress hyperglycemia ratio (SHR) in patients with hemorrhagic stroke. Survival probabilities are depicted for ICU mortality (A), in-hospital mortality (B), 30-day mortality (C), 90-day mortality (D), 180-day mortality (E), and 1-year mortality (F). Log-rank tests were used to compare survival differences across tertiles.

Figure 4

Restricted cubic spline (RCS) curves illustrating the association between stress hyperglycemia ratio (SHR) and all-cause mortality (ACM). The analysis models ICU mortality (A), in-hospital mortality (B), 30-day mortality (C), 90-day mortality (D), 180-day mortality (E), and 1-year mortality (F). Solid lines represent hazard ratios (HRs), and shaded areas denote 95% confidence intervals. Adjusted for age, gender, race, sodium, hemoglobin, WBC, surgical procedure, GCS, CR, ventilation, oxygen delivery, AMI, peripheral vascular disease, ventilator-associated pneumonia, INR, SIRS, malignancy, hypertension and diabetes.

Figure 5

Subgroup analyses evaluating the association between stress hyperglycemia ratio (SHR) and ICU mortality (A), in-hospital mortality (B), 30-day mortality (C), 90-day mortality (D), 180-day mortality (E), and 1-year mortality (F). Each subgroup analysis was adjusted for age, gender, race, sodium, hemoglobin, WBC, surgical procedure, GCS, creatinine, ventilation, oxygen delivery, AMI, peripheral vascular disease, ventilator-associated pneumonia, INR, SIRS, malignancy, hypertension, and diabetes, excluding the subgroup variable itself.