Clinical diagnostics versus a theoretic algorithm in diagnosing abdominal pain.

In the United States, approximately half a million people are affected annually by gallbladder disease. Accurate diagnosis of acute cholecystitis versus a bad case of biliary colic is not always obvious, but distinguishing between the two has important therapeutic implications. Mills et al have performed a sophisticated statistical study relating clinical and laboratory variables with ultrasound findings in right upper quadrant abdominal pain. Their retrospective study includes 177 patients seen in a busy inner-city teaching hospital. They found four statistically significant predictors of ultrasounddiagnosed acute cholecystitis versus ultrasound-diagnosed cholelithiasis. They found that elevated levels of alkaline phosphatase, total bilirubin, clinical Murphy sign, and ultrasonic Murphy sign were predictors of the ultrasonic diagnosis of acute cholecystitis, biliary pancreatitis, and/or choledocholithiasis versus cholelithiasis. It is interesting to note that Murphy sign (a sign taught to all medical students) does, even after multivariate analysis, predict positive ultrasound. This underscores the importance of a good clinical examination with any patient who has abdominal pain. I vividly remember the first time I encountered an individual with Murphy sign. While visiting my parents during undergraduate school, my father was admitted with rightsided chest pain to rule out a myocardial infarction, since he had a history of coronary artery bypass. An internist came to examine him, and a Murphy sign was elicited. Immediately, the internist told us that it appeared that the gallbladder was inflamed. Even back then, an ultrasound was ordered, which suggested acute cholecystitis. More than a decade later, our diagnostic ability has not changed much. Since, at many institutions, emergent ultrasound may be more difficult to obtain, predicting positive results of ultrasound for right upper quadrant pain may help reduce medical costs. Mills et al have given the practicing physician some variables that are more likely to be associated with positive ultrasound findings. In future studies, they may wish to further extend their statistical analysis to include pathologically proven acute cholecystitis. As suggested by the authors, a theoretic algorithm or model that could predict acute cholecystitis would be useful. Prospective application of this model and results (including outcome, surgical findings, and pathologic findings) would be ideal. Until then, physicians will rely on clinical acumen, with knowledge gained from studies such as Mills et al to determine which patients should undergo ultrasound and initiate appropriate therapy.