Simple Predictors of Mortality in Acute Pancreatitis
Scott AM, Baidoun M, Farook N, Acevedo HG, Bhatti SK, Dueweke J, Nair V, Mohammed H, Adhami A, Ahluwalia G, Hebbar S, and Miller J. Simple Predictors of Mortality in Acute Pancreatitis. J Gen Intern Med 2019; 34(2):S353.
J Gen Intern Med
Background: Various historical studies have found that hypoxia portends a poor prognosis in patients hospitalized with acute pancreatitis. The objective of this study is to test early respiratory compromise on hospital presentation as a mortality predictor in acute pancreatitis. Secondary aims of the study were assessing prognostication abilities of the quick sequential organ failure assessment (qSOFA) and hypoalbuminemia in acute pancreatitis. Methods: This is a retrospective observational study based on adult patients hospitalized for acute pancreatitis. The data were collected from eight standalone emergency departments and five hospitals within Henry Ford Health System of Southeast Michigan. Patients with lipase levels less than three times the upper limit of the normal laboratory range were excluded. The primary outcome was inpatient mortality. Univariate and multivariate logistic regression was performed to determine predictors of adverse outcomes. The main variable of interest was respiratory compromise, defined as an initial SpO2 of 92% on presentation. We combined other key variables on presentation to derive an area under the curve (AUC) for predicting mortality. Results: The study included 2,090 patients, mean age 55.5 years (SD 17.5) and 51.5% female. The median presenting lipase was 976 U/L (IQR 281-2500). Death was uncommon (n=34, 1.6%). Via univariate analysis, highly significant predictors of mortality were SpO2 92%, qSOFA score 2, and low albumin. Patients with a presenting SpO2 92% had 9.2% mortality, in contrast to a 1.3% mortality rate among those with an SpO2 above 92% on presentation (OR 7.5, 95% CI 3.1-17.7). Those with a qSOFA score 2 suffered 9.8% mortality, compared to an average 1.3% mortality rate among those with a qSOFA of 0 or 1 (OR 8.6, 95% CI 3.9-18.9). Adjusting for age, gender, race, leukocytosis, hematocrit, and major comorbidities, qSOFA 2 remained a significant predictor of mortality (OR 2.6, 95% CI 1.7-4.1, p < 0.001). Low albumin (OR 2.7, 95% CI 1.2-6.2, p 0.02) and SpO2 92% (OR 2.6, 95% CI 1.0-7.1, p 0.05) also remained significant predictors. By combining qSOFA with the presence of low albumin and SpO2 92% into a novel early acute pancreatitis score (EAPS), the EAPS score had good accuracy for predicting mortality (AUC 0.80, 95% CI 0.72-0.89). Conclusions: Independent predictors of mortality on initial presentation include low albumin, SpO2 92%, and a qSOFA score 2. Perhaps these simple prognosticators can guide clinical practice by more precisely identifying those with acute pancreatitis who are more critically ill and could benefit from closer monitoring.