Relationship between the mean of 24-h venous blood glucose and in-hospital mortality among patients with subarachnoid hemorrhage: A matched cohort study

Abstract

Objective: The aim of this study was to explore the correlation between the mean of 24-h venous blood glucose (BG) and in-hospital mortality and all-cause mortality (ACM) in patients with subarachnoid hemorrhage (SAH). Methods: Detailed clinical information was acquired from the Medical Information Mart for Intensive IV (MIMIC-IV) database. The best cutoff value of mean BG was calculated using the X-tile program. Univariate and multivariate logistic regressive analyses were utilized to analyze the prognosis significance of mean BG, and survival curves were drawn using the Kaplan-Meier (K-M) approach. To improve the reliability of results and balance the impact of underlying confounders, the 1:1 propensity score matching (PSM) approach was utilized. Results: An overall of 1,230 subjects were selected herein. The optimal cutoff value of the mean BG for in-hospital mortality was 152.25. In addition, 367 pairs of score-matched subjects were acquired after PSM analysis, and nearly all variables' differences were balanced. K-M analysis showed that patients with mean BG ≥ 152.25 mg/dl had significantly higher in-hospital, 3-month, and 6-month mortalities compared with patients with mean BG < 152.25 mg/dl (p < 0.001). The multivariable logistic regressive analyses revealed that patients with mean BG ≥ 152.25 mg/dl had significantly increased in-hospital mortality compared with patients with mean BG < 152.25 mg/dl after the adjustment for possible confounders (OR = 1.994, 95% CI: 1.321–3.012, p = 0.001). Similar outcomes were discovered in the PSM cohort. Conclusion: Our data suggested that mean BG was related to ACM of patients with SAH. More studies are needed to further analyze the role of the mean of 24-h venous BG in patients with SAH.