Hourly Blood Urea Nitrogen Trajectories and ICU Mortality Risk in Critically Ill Patients with Gastrointestinal Bleeding: A Multicenter Retrospective Cohort Study.

INTRODUCTION

Elevated blood urea nitrogen (BUN) is associated with adverse outcomes in gastrointestinal bleeding (GIB), but isolated measurements may not capture early temporal patterns.

METHODS

A multicenter retrospective cohort study included 5847 adult patients with GIB from four databases (2008-2022). Serial BUN measurements during the first 24 h after ICU admission were mapped to hourly intervals, with at least 12 observed values required. Generalized additive models assessed concentration-duration patterns, adjusted logistic regression evaluated 15 BUN ranges, and restricted cubic splines examined time-weighted BUN. Meta-analysis, measurement-density sensitivity analyses, and static-dynamic model comparisons were performed.

RESULTS

Male and female patients comprised 56.5-58.7% and 41.3-43.5% of each cohort,, respectively. ICU mortality ranged from 18.0% to 19.7%. BUN < 30 mg/dL sustained for > 12 h lowered mortality odds (OR: 0.40-0.62), while BUN ≥ 50 mg/dL for >18 h markedly increased them (OR: 1.55-3.39). Prolonged time within 5-30 mg/dL was associated with lower mortality odds, while greater exposure to 40-70 mg/dL increased risk, with no detectable between-database heterogeneity. Associations were nonlinear and remained broadly consistent in denser-measurement subsets. Dynamic models improved AUC from 0.678-0.701 to 0.718-0.748 (all DeLong P ≤ 0.05), with improved reclassification and lower Brier scores.

CONCLUSION

Early dynamic BUN trajectories were associated with ICU mortality and provided incremental prognostic information beyond admission BUN alone. Incorporating both the magnitude and persistence of BUN abnormalities may support early risk stratification in critically ill patients with GIB.