Impact of Sepsis Onset Timing on All-Cause Mortality in Acute Pancreatitis: A Multicenter Retrospective Cohort Study

Background Sepsis complicates acute pancreatitis (AP), increasing mortality risk. Few studies have examined how sepsis and its onset timing affect mortality in AP. This study evaluates the association between sepsis occurrence and all-cause mortality in AP, focusing specifically on the impact of sepsis onset timing. Methods This multicenter retrospective cohort study included 494 ICU-admitted AP patients from the MIMIC-IV database and 91 from our center. Patients were grouped by sepsis occurrence and onset timing. Clinical outcomes were in-hospital and 90-day all-cause mortality. Machine learning identified key variables associated with mortality. Multivariable regression analyzed the impact of sepsis and its onset timing on mortality. To reduce baseline differences, propensity score matching (PSM) based on time to sepsis was conducted. After PSM, Kaplan-Meier survival analyses incorporated data from our center for validation. Restricted cubic spline analysis examined any nonlinear relationship between sepsis onset timing and mortality. Results Patients with sepsis had significantly higher in-hospital and 90-day mortality rates than those without sepsis ( p < 0.05). Sepsis was identified as a significant risk factor for in-hospital mortality and remained significantly associated after adjusting for key variables ( p < 0.05). However, sepsis onset timing did not significantly impact in-hospital or 90-day mortality. These findings were validated after PSM and with our center's data. No nonlinear relationship between sepsis onset timing and mortality was found. Conclusion Sepsis significantly increases all-cause mortality in AP patients, but the timing of its onset has limited impact. Continuous monitoring and intervention for sepsis during hospitalization are recommended to improve prognosis.