The role of red blood cell distribution width in prognosis prediction among critically ill patients with obstructive sleep apnea: Insights from the MIMIC-IV database

Obstructive sleep apnea (OSA) is prevalent in critically ill patients and associated with increased mortality. This study aimed to investigate the association between red blood cell distribution width (RDW) at admission and one-year mortality in critically ill OSA patients. A retrospective cohort study involving 1950 critically ill patients with OSA from the MIMIC-IV (v3.1) database. Kaplan-Meier survival analysis, multivariable Cox regression, and restricted cubic spline (RCS) analyses were performed to evaluate the association between RDW and all-cause mortality. Subgroup analyses were performed to assess the consistency of the RDW-mortality association and identify potential interaction effects. The incidence of mortality increased significantly across RDW quartiles (Q1: 4.6% vs. Q2: 12.1% vs. Q3: 18.4% vs. Q4: 34.7%, P < 0.001). Kaplan-Meier analysis demonstrated significantly higher 30-day and one-year all-cause mortality among patients in the higher RDW quartiles (log-rank P< 0.001). Multivariate Cox regression analysis revealed that RDW levels were independently correlated with one-year mortality after adjusting for multiple variables for both continuous (hazard ratio [HR] 1.15, 95% confidence interval [CI]: 1.10-1.20; P < 0.001) and categorical comparisons (Q4 vs. Q1: HR 4.42, 95% CI: 2.67-7.31; P< 0.001). RCS regression confirmed a significant positive relationship between RDW and the risk of short- or long-term mortality (P_{for overall} <0.001). Subgroup analysis confirmed the robustness of the results and identified interactions between RDW and White race, and BMI status (P for interaction = 0.011 and 0.052, respectively). This study established a positive association between RDW and all-cause mortality in OSA patients admitted to the ICU. Higher RDW levels were associated with an increased risk of short- and long-term all-cause mortalities.