Original data information for participants.

Background
Type 2 diabetes mellitus (T2DM) complicated with ischemic stroke is a major challenge to global public health and is related to poor prognosis. However, the role of blood urea nitrogen(BUN)to serum albumin ratio (BAR) in predicting in-hospital mortality of T2DM patients with ischemic stroke has not been fully explored. This study was carried out to investigate the relationship between BAR level and in-hospital mortality of T2DM patients with ischemic stroke.
Methods
The MIMIC-IV database was searched for data on T2DM patients with ischemic stroke. The primary outcome was in-hospital mortality. The BAR was calculated as follows: BUN (mg/dl)/ serum albumin (g/dl). Logistic regression was employed to investigate the relationship between BAR and in-hospital mortality of T2DM patients with ischemic stroke. The restricted cubic spline (RCS) was leveraged to examine the dose-response relationship of BAR with the outcome. The receiver operating characteristic (ROC) curve was utilized to measure the ability of BAR to predict the outcome. In addition, the decision curve analysis (DCA) was employed to explore the value of BAR in clinical practice. The consistency and robustness of the research results were assessed by subgroup analysis and the presence of interactions using a likelihood ratio test.
Results
Finally, 1136 patients were included for evaluation in this study. As BAR levels increased, the in-hospital mortality of T2DM patients with ischemic stroke also increased (OR:1.06; 95% CI:1.01–1.11; P < 0.01). The RCS analysis suggested that there was a linear relationship between BAR and in-hospital mortality of T2DM patients with ischemic stroke (p = 0.276). The ROC curve indicated that BAR was superior to BUN, Sepsis-related Organ Failure Assessment (SOFA), and Glasgow Coma Scale (GCS) in predicting the in-hospital mortality of T2DM patients with ischemic stroke. The DCA curve indicated that the net benefit of BAR was better than BUN, SOFA, and GCS. Subgroup analysis showed that there was no interaction between BAR and each subgroup (all p-value > 0.05).
Conclusion
The in-hospital mortality of T2DM patients with ischemic stroke increased with elevated BAR levels.