Abstract
OBJECTIVE: To establish a rapid and concise prognosis scoring system for pancreatitis in the emergency department based on bedside arterial blood gas analysis (ABG). METHODS: A single-center, retrospective cohort study was used to establish the new scoring system, and a validation group was used to verify it. The primary endpoint was 60-day death, and secondary endpoints were 28-day death, admission to the intensive care unit (AICU), requirement for mechanical ventilation (MV) and persistent organ failure (POF). Receiver operating characteristic (ROC) curves was drawn to validate the predictive value of the new scoring system. The performance of the new scoring system was compared with that of conventional predictive scoring. RESULTS: 443 patients were in the derivation group and 217 patients in the validation group, of which 27 and 25 died during follow-up. A total of 443 patients in the derivation group, 27 of whom died during the follow-up period. Multivariate regression analysis showed that mental status, hematocrit (HCT), base excess (BE) and Serum ionic calcium (Ca(2+)) were independent risk factors for 60-day mortality of pancreatitis, and they were used to create a new scoring system (MHBC). In the derivation and validation, the ability of MHBC (AUC= 0.922, 0.773, respectively) to predict 60-day mortality from pancreatitis was no less than that of APACHE II (AUC= 0.838, 0.748, respectively) and BISAP (AUC= 0.791, 0.750, respectively), while, MHBC is more quickly and concisely than APACHE II and BISAP. Compared with MHBC less than or equal to 2, when MHBC is greater than 2, the 28-day mortality, 60-day mortality and the incidence of AICU, MV and POF increased significantly (P <0.001). CONCLUSION: The MHBC can quickly and concisely evaluate the 60-day mortality, 28-day mortality, and the incidence of AICU, MV and POF of patients with acute pancreatitis in the emergency department.