Association between trajectories of urea nitrogen-to-creatinine ratio and all-cause mortality in patients with acute pancreatitis: analysis of the MIMIC-IV database

Acute pancreatitis (AP) progression to severe acute pancreatitis (SAP) carries high mortality and healthcare burden, with current prognostic tools limited by delayed results or single-time-point biomarkers. The blood urea nitrogen/creatinine ratio (BCR) is a promising prognostic indicator, but its dynamic trajectory’s link to mortality in critically ill AP patients remains unclear. This study aimed to investigate the association between BCR trajectories and all-cause mortality at 28 and 90 days in AP patients. Data were extracted from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database (Version 3.0). Group-based trajectory modeling (GBTM) was applied to identify distinct BCR trajectory patterns using the first valid paired BUN–creatinine measurement within each of three predefined time windows (0–24 h, 24–48 h, and 48–72 h post-ICU admission). Kaplan–Meier curves with log-rank tests were used to compare survival outcomes across trajectory groups. Cox proportional hazards models—Crude (unadjusted), Model I (adjusted for age, sex, Sequential Organ Failure Assessment [SOFA] score, intensive care unit length of stay [LOS], and Continuous Renal Replacement Therapy [CRRT]), and Parsimonious Primary Model (additionally adjusted for albumin, norepinephrine, and ventilation)—were constructed to examine the association between BCR trajectories and mortality risk. Subgroup analyses with interaction tests were performed to assess the robustness of the findings. Among 606 eligible patients, four BCR trajectories were identified: moderate-stable (44.7%), low-stable (39.8%), descend-ascend (2.1%), and high-stable (13.4%; Class 4). Patients in Class 4 had significantly higher 28- and 90-day mortality. After multivariable adjustment, Class 4 remained independently associated with 28-day mortality (HR 3.05, 95% CI 1.58–5.89, P < 0.001) and 90-day mortality (HR 1.95, 95% CI 1.21–3.13, P = 0.006). Subgroup analyses revealed no significant interactions (all P for interaction > 0.05). Dynamic BCR trajectories are independently associated with 28- and 90-day all-cause mortality in acute pancreatitis, with the high-stable trajectory (Class 4) conferring the greatest mortality risk. BCR trajectory assessment may inform early risk stratification, although its clinical utility and impact on outcomes require prospective validation. Not applicable.