Objective: Recent case control data suggested that a severe course of acute pancreatitis in HIV+ patients was 1) common (50% of cases), 2) poorly predicted by Ranson's criteria (sensitivity 41%), and 3) accurately predicted by a diagnosis of AIDS (positive predictive value 67%). However, the definition of severity included length of stay in hospital and excluded commonly accepted markers (local complications, systemic complications, and need for surgery). The aim of this study was to determine 1) the prevalence of severity and 2) the value of these predictors with regard to severity, as defined by commonly accepted standardized criteria in patients with AIDS and acute pancreatitis.
Methods: A retrospective review identified 50 patients with AIDS exhibiting clinical, laboratory, and/or radiological features of acute pancreatitis.
Results: Only five patients followed a severe course as defined by accepted markers. Of these patients, 29 had values available for at least nine of 11 of Ranson's criteria (sensitivity 80%, specificity 54%). Points were awarded most commonly for decreased serum Ca2+ (n = 14) and elevated serum LDH (n = 7).
Conclusions: In patients with AIDS and acute pancreatitis at our institutions, 1) the prevalence of severity and 2) the sensitivity of Ranson's criteria with regard to severity is comparable to that reported in large historical case series of immunocompetent patients. Pseudohypocalcemia and/or elevation in LDH are frequent, likely due to the catabolic infectious disease state.