Predictors, severity and associate factors of acute pancreatitis: A tertiary hospital’s experience

Abstract

Aim: Acute pancreatitis is a common disorder that occurs following an acute response to a pancreatic injury. The aim of this study was to assess predictors of severity and associated factors, as well as the association of different classification systems of severity among pa-tients with acute pancreatitis (AP).
Methods: A retrospective case series study was conducted in Albania including 150 patients with AP between March 2021 and March 2022. Variables such as baseline characteristics, la-boratory findings, and calculated scores of known severity classifications were analyzed. Pa-tients were graded as having mild, moderate, or severe acute pancreatitis based on the Revi-sion of the Atlanta Classification (RAC). Ordinal logistic regression was used to model the relationship between the ordinal variable (RAC categories) and the explanatory variables men-tioned above.
Results: Women with AP had a higher average age than men with AP (62.5 vs. 57.5 years old, respectively, p<0.05). Additionally, the alcoholic etiology in males prevailed in 100% of cases, while the biliary etiology was more common in females (64.2% compared to 35.8% in males, p<0.001). Ordinal logistic regression showed that a one unit increase in the CT Severi-ty Index (CTSI) and Bedside Index for Severity in Acute Pancreatitis resulted in a 0.968 and 0.430 times increase, respectively, in the ordered log-odds of being in a higher RAC classifi-cation category. The presence of Systemic Inflammatory Response Syndrome (SIRS) (vs. non-present) resulted in a 2.98 higher ordered logit. Conversely, a one unit increase in satura-tion level decreased the ordered log-odds by approximately 0.4 times.
Conclusion: The severity of acute pancreatitis is a medical event that requires accurate pre-diction, for which many classification systems have been compiled, with the RAC being the most recent consensus. CTSI, the presence of SIRS, and saturation levels are significantly as-sociated with RAC, without excluding the discussion on the predictive value of laboratory findings, such as glycemia, azotemia, and creatinine.