S. Tenner et al., PREDICTING GALLSTONE PANCREATITIS WITH LABORATORY PARAMETERS - A METAANALYSIS, The American journal of gastroenterology, 89(10), 1994, pp. 1863-1866
Objectives: In order to define the usefulness of biochemical laborator
y values in distinguishing gallstone from non-gallstone acute pancreat
itis, we performed a meta-analysis. Methods: Studies that utilized the
bilirubin, alkaline phosphatase, alanine aminotransferase (ALT), and
aspartate transaminase in the evaluation of the gallstone pancreatitis
were included. Results: Using receiver operating characteristic curve
s for each of these four parameters, we determined that the ALT level
was the most clinically useful parameter. The higher the serum level o
f ALT, the greater its specificity and positive predictive value in di
agnosing gallstone pancreatitis. At ALT levels greater than or equal t
o 150 IU/L (approximately a 3-fold elevation), the probability of gall
stone pancreatitis is 95%. Our analyses of total bilirubin and alkalin
e phosphatase serum levels indicate that these tests are not useful in
the diagnosis of gallstone pancreatitis. Aspartate transaminase level
s are nearly as useful as ALT. Conclusions: A greater than or equal to
3-fold elevation of ALT in the presence of acute pancreatitis has a p
ositive predictive value of 95% in diagnosing acute gallstone pancreat
itis.