Sepsis Indicators in Acute Pancreatitis

Publisher Website
Google Scholar
Abstract
Twenty-one routine clinical and laboratory data in 161 patients with necrotizing pancreatitis (NP) undergoing surgical treatment were analyzed. The necrotic tissue at operation was bacterially infected in 41% of the patients. The goal of the study was to evaluate whether there was any special clinical feature in cases of an infection. The parameters were recorded during 48 h after admission as well as during 48 h before operation, and the frequencies submitted to both a univariate and a multivariate analysis (logistic regression model). In the period after admission, patients with infected necrosis significantly more often had a rectal temperature > 38.5.degree. C (p = 0.001). Before operation (i.e., after maximum conservative treatment), four findings were significantly related to an infection: rectal temperature > 38.5.degree. C, base excess > - 4 mmol/L, hematocrit < 35% (all p = 0.0001), and paO2 < 60 mm Hg (p = 0.001). The multivariate analysis, which calculates and quantifies the mutual influence of factors, showed a combination of three findings (rectal temperature > 38.5.degree. C, base excess > - 4 mmol/L, and hematocrit < 35%) to be related to necrosis infection before operation. All three criteria in a patient imply a probability of infection of 83%. It is noteworthy that the sepsis indicators were equally distributed in patients with focal, extended, or subtotal/total infected necrosis, but correlated with the necrosis extent in sterile necrotizing pancreatitis. Moreover, all parameters not related to the pancreatic infection [e.g., hyperglycemia, hypocalcemia, rise of lactic dehydrogenase (LDH), and the white blood cell count] correlated with the three necrosis categories.

This publication has 0 references indexed in Scilit: