Evaluation of prognostic factors in patients with acute pancreatitis

The severity of acute pancreatitis is variable and does not always correlate with structural and functional changes in the pancreas. More precise predictors of severity are necessary to enable intensive therapy to be targeted at patients with severe attacks, and to judge efficacy of treatment, to help in the early detection of complications, and to facilitate comparison of patients from different centers.

On admission, clinical criteria, biochemical and hematological parameters and multiple prognostic scores (Ranson, Imrie and APACHE-II scores) were collected from patients with acute pancreatitis.

Two hundred and twenty-seven patients were seen during the study period. The overall mortality rate was 11.4%. A significantly higher mortality was found in patients with severe pancreatitis (25.8%) than in those with mild disease (1.5%, p=0.00001). Mortality was related to the presence of fever at admission (16.5% vs. 5.0%, p=0.006), and not to age, sex, etiological associations, or concomitant medical or surgical diseases. Of the biochemical and hematological parameters tested, ten factors (plasma glucose, BUN, serum creatinine, serum calcium, serum lactate dehydrogenase, serum albumin, red cell count, white cell count, hematocrit, and lymphocytes count) were found to be statistically significant, and four factors (serum potassium, alkaline phosphatase, total bilirubin, and hemoglobin) were marginally significant. In high-risk patients (Ranson and Imrie scores > or = 3, APACHE-II score > or = 10) mortality rates were higher (22.5%, 23.0%, and 22.5%, respectively) than in low-risk patients (2.4%, 2.3%, and 0.8%, respectively, p=0.00001). APACHE-II scores may be especially useful for monitoring the progress of patients with pancreatic necrosis and secondary pancreatic necrosis.

Acute pancreatitis still represents a condition of variable severity. The adoption of multiparametric criteria proposed together with morphological evaluation consents the formulation of a discreetly reliable prognosis on the evolution of the disease a few days from onset, even though this still appears insufficient to plan a varied and timely therapeutic program.