The plain radiographic features of biliary tract disease usually are nonspecific [1]. The most common bowel gas finding seen in patients with acute biliary disease is a generalized ileus. Gallstones rarely are detected on plain radiographs, because only 20% of stones have a sufficient calcium concentration to make them radiopaque. Air in the biliary tree may result from a biliary-enteric fistula, prior sphincterotomy, or a biliary-enteric surgical anastomosis. Infections with gas-producing organisms rarely present with gas in the biliary tree or gallbladder wall.

Ultrasonography of the biliary tree is an extremely useful diagnostic test in the ICU setting and can be performed at bedside with good results [2]. It is a sensitive test for determining biliary ductal dilatation, and the accuracy of ultrasonography in detecting cholelithiasis exceeds 95%. However, its accuracy in detecting choledocholithiasis may be limited, as gas in the duodenum can obscure visualization of the distal bile duct. In the presence of cholelithiasis or gallbladder sludge, the findings of ductal dilatation, elevated liver enzymes, abdominal pain, and fever are strongly suggestive of cholangitis. Findings on ultrasonography that may indicate acute gallbladder disease include focal tenderness over the gallbladder, thickening of the gallbladder wall, and pericholecystitic fluid collections, but none is specific for cholecystitis. The technique also may detect other abnormalities, including liver lesions, pancreatic masses, abscesses, and ascites.

Scanning the abdomen after an intravenous injection of technetium-99m iminodiacetic acid yields physiologic and structural information regarding the biliary tract. Filling the gallbladder with radionuclide confirms the patency of the cystic duct and virtually excludes the diagnosis of acute cholecystitis. False-positive examinations can be seen in patients with chronic cholecystitis, on long-term parenteral hyperalimentation, or after prolonged fasting. Delayed views and routine pretreatment with cholecystokinin increase the accuracy of technetium-99m iminodiacetic acid scanning for acute cholecystitis to greater than 93% [3]. Hepatobiliary scanning is also useful in identifying structural abnormalities of the biliary tree such as significant bile duct leaks, which can be identified in almost all patients with this problem. Scanning has a limited role in patients with poor hepatocellular function, complete biliary obstruction, and cholangitis because these defects often prevent adequate uptake or excretion of the radiopharmaceutical into the biliary tree.

CT is highly accurate for the detection of biliary tract disease [4]. The sensitivity for detecting choledocholithiasis is as high as 88% using helical CT. Unlike ultrasonography or radionuclide scanning, however, CT cannot be used portably in the ICU. Findings on CT for gallbladder disease include thickening of the gallbladder wall, pericholecystitic fluid, and adjacent abscesses. In addition, CT is highly accurate for the detection of biliary tract obstruction (i.e., the level and the cause). It also allows detailed visualization of the pancreas, vessels, and surrounding organs and can be used to assess the severity of pancreatitis including complications such as necrotizing pancreatitis or the formation of pseudocysts.

Advances in magnetic resonance imaging (MRI) technology have greatly improved the resolution of biliary imaging [5]. The use of magnetic resonance cholangiogram images can be manipulated to display highly accurate representations of the biliary tree. The sensitivity of this technique rivals that of direct cholangiography. In one study, the overall sensitivity and specificity of magnetic resonance cholangiograms for diagnosis of bile duct stones were 100% and 95.6%, respectively, with corresponding positive and negative predictive values of 92.6% and 100% [6]. This technique also permits visualization of any
mass contiguous to a bile duct stricture. In the intensive care setting, its use may be limited as patients must be transported to the scanner, and problems with magnetic compatibility of support equipment must be overcome. The noninvasive images of the biliary tree obtained, however, can be highly diagnostic and preclude other more invasive studies.

When evaluating the ICU patient with suspected biliary tract disease, ultrasonography should be the initial procedure of choice, followed by hepatobiliary scanning if cystic duct obstruction or bile leakage is suspected. Both are portable and noninvasive. Ultrasonography is highly accurate for the detection of gallstones and structural pathology. Hepatobiliary scanning, on the contrary, provides physiologic information, primarily regarding patency of the cystic duct. Such physiologic data are especially important for patients with suspected calculous or acalculous cholecystitis. CT or MRI should be reserved for those patients in whom sonographic or radionuclide findings are equivocal, if other intra-abdominal pathology needs to be excluded, or if ductal dilatation is seen on ultrasonography without a clearly defined etiology.

The technique of ERCP is used for both diagnostic and therapeutic purposes as described in Chapter 13. In brief, a side-viewing endoscope is passed through the mouth into the second duodenum, where the major ampulla is identified and cannulated. The biliary tree is then opacified with contrast injected through a catheter, allowing a retrograde cholangiogram to be obtained. Fluoroscopy and standard radiographs are used to examine the biliary tree and define such abnormalities as stones, strictures, leaks, and obstruction. Endoscopic therapy, including stone removal, biliary drainage, or stricture dilatation, can be accomplished in the same setting. ERCP can be used in the evaluation and therapy of the ICU patient, especially if the patient can be stabilized for endoscopy and transported to a fluoroscopy room. Rarely is it necessary to perform emergent biliary decompression at the bedside using portable fluoroscopy. Coagulopathies should be corrected before the procedure, especially if an endoscopic sphincterotomy (electrocautery incision of the sphincter of Oddi in the duodenal wall for stone removal or drainage) is anticipated. If coagulopathies cannot be satisfactorily corrected, a stent can be placed into the bile duct to ensure drainage without performing a sphincterotomy. Major morbidity from the diagnostic procedure includes pancreatitis, cholangitis, perforation, and hemorrhage. The complication rates of ERCP in a recent review noted reduced rates, compared with prior reporting, of pancreatitis at 3.5%, infection at 1.4%, and perforation at 0.6% under standard conditions [7]. The value of ERCP is largely operator dependent and can be highly successful in the delineation and treatment of biliary disease in the ICU patient [8].

EUS involves the transoral passage of an endoscope with an ultrasonic transducer at the tip. The limitations of transabdominal ultrasonography are overcome with this modality because all areas of the biliary tree, including the intrapancreatic portion of the bile duct as well as the pancreas, can be imaged without interference from gas in the intestines. EUS can reliably identify cholelithiasis and is more sensitive than transabdominal ultrasonography in detecting choledocholithiasis in patients with biliary pancreatitis. Although EUS is typically an elective procedure and uncommonly used in the ICU, the test may be useful in identifying those patients who would benefit from endoscopic stone extraction by ERCP [9].

Percutaneous transhepatic cholangiography (PTC) also may be used in evaluating the ICU patient. The technique requires fluoroscopy to guide passage of a needle into the intrahepatic bile ducts. The biliary tree is then filled with contrast and images are taken. The use of PTC for the diagnosis of biliary tract pathology has been supplanted by ERCP and noninvasive examinations, such as ultrasonography, CT, and MRI. Currently, PTC is used primarily as an initial step in percutaneous transhepatic biliary drainage. Decompression of the biliary tree via a percutaneous catheter is a highly effective method for rapid nonoperative and nonendoscopic biliary decompression. This procedure is indicated when a patient requires emergent biliary drainage but is not stable enough to undergo ERCP under conscious sedation, if the major papilla cannot be reached endoscopically because of postsurgical anatomy or a technical failure in cannulating the bile duct. The technique involves an initial PTC to delineate the biliary anatomy, followed by selective cannulation of an appropriate intrahepatic bile duct with an 18-gauge needle. A guidewire is then passed into the biliary tree, the tract is dilated, and a drainage catheter placed. Successful drainage can be established in almost all patients. Percutaneous biliary drainage is an invasive procedure, and acute complications, including hemorrhage, sepsis, and bile leakage, occur in 1% to 5% of patients [10].