Red Blood Cell Distribution Width as a Risk Factor for 30/90-Day Mortality in Patients with Gastrointestinal Bleeding: Analysis of the MIMIC-IV Database

The purpose of this research was to assess the relationship between red blood cell distribution width (RDW) and mortality in patients with gastrointestinal (GIB) bleeding in the intensive care unit (ICU).

The information of the participants was obtained from the Medical Information Mart for Intensive Care IV database. The main outcome of this research was 30/90-day mortality, with ICU mortality and in-hospital mortality as secondary outcomes.

This research included 2924 patients with gastrointestinal bleeding in total. Patients with higher RDW had considerably higher 30/90-day and in-hospital mortality rates, as well as longer hospital stays and ICU stays. According to the Kaplan–Meier analysis, the 30/90-day mortality rate was remarkably higher among participants in the higher RDW group (P < 0.0001). In the adjusted multivariate Cox regression analysis, for 30-day mortality, the HR (95% CI) was 1.75 (1.37, 2.24) in comparison to Q1 in the reference group (P < 0.001). Analyses of 90-day mortality and in-hospital mortality both showed the same results. In the subgroup analysis, gender, myocardial infarction, chronic pulmonary disease, cerebrovascular disease and renal disease had no significant effect on the correlation between RDW values and mortality (all P > 0.05). The area under the ROC curve for RDW was 0.599 (95% CI 0.581–0.617) and 0.606 (95% CI 0.588–0.624) in 30/90-day ICU mortality.

The current research showed that RDW could be utilized as an independent indicator of short-term mortality in critically ill GIB patients at 30 and 90 days of hospital admission.