EUS for suspected choledocholithiasis: Do benefits outweigh costs?

A prospective, controlled study

Elisabetta Buscarini, MD, Paolo Tansini, MD, Daniele Vallisa, MD, Alessandro Zambelli, MD, Luigi Buscarini, MD
Crema and Piacenza, Italy

Background: This prospective study of the use of EUS to prevent unnecessary endoscopic retrograde cholangiography in patients with suspected choledocholithiasis has two aims: tO evaluate the effectiveness, based on patient outcome, and the potential clinical and economic benefits of EUS. Methods: A prospective series of 485 patients (202 men, 283 women; mean age 66.2 years, range 20-94 years) suspected to have choledocholithiasis based on clinical, biochemical, and crosssectional imaging (US or CT) data underwent EUS. Positive EUS findings were confirmed by endoscopic retrograde cholangiography with sphincterotomy and/or by surgery; negative findings were confirmed by clinical follow-up. An EUS result was considered a true negative if the patient was confirmed symptom-free with normal tests on follow-up of at least 6 months.The costs of EUS (procedure, days of hospitalization, any morbidity) were compared with the estimated costs of the endoscopic retrograde cholangiography avoided in patients with true-negative EUS findings. Results: EUS findings were verified in 463 patients: EUS diagnosed choledocholithiasis in 239 (51.6%) and the absence of stones in 220 patients. In 4 patients (0.8%), EUS was incomplete. By sphincterotomy (209 patients), surgical bile duct exploration (39), percutaneous cholangiography (1 case in which EUS demonstrated a bile duct tumor), and follow-up (214), EUS diagnoses were confirmed as follows: 237 true-positive, 216 true-negative, 2 false-positive, 4 false-negative, 4 incomplete (sensitivity 98%, specificity 99%, positive predictive value 99%, negative predictive value 98%, accuracy 97%). In 214 (46%) patients, more invasive investigations of the bile duct were avoided. The mean cost for patients managed by the EUS-based strategy was E374.50 (:1:284.72), which was significantly less than the theoretical mean cost of C443.80 (p < 0.001) for patients undergoing endoscopic retrograde cholangiography. Conclusions: The results of this study confirm that EUS is highly reliable for the diagnosis of choledocholithiasis. Its use offers considerable clinical and economic advantages by preventing inappropriate and more invasive evaluation of the bile duct. (Gastrointest Endosc 2003;57:510-8.)

Endoscopic retrograde cholangiography (ERC) is considered the standard nonsurgical technique for diagnosis of bile duct stones. ERC allows stone removal at the s a m e endoscopic session by endoscopic sphincterotomy (EST), but it is an invasive method associated with s u b s t a n t i a l morbidity. Noninvasive diagnostic tests are performed to select 9 patients with suspected choledocholithiasis for ERC. Based on clinical and biochemical criteria together with transabdominal US findings, patients can be grouped into risk classes, ranging from low to high, for duct stones. 1 If this initial triage identifies a patient as being at risk of having bile duct stones,
Received July 17, 2002. For revision September 17, 2002. Accepted October 30, 2002. Current affiliations: Gastroenterology Department, Ospedale Maggiore, Crema, Gastroenterology Department, Ospedale Civile, Piacenza, and Internal Medicine Department, Ospedale Civile, Piacenza, Italy. Reprint requests: Elisabetta Buscarini, MD, Gastroenterology Department, Ospedale Maggiore, via Macall~-1,26013 Crema, Italy. Copyright 9 2003 by the American ~ociety for Gastrointestinal Endoscopy 0016-5107/2003/$30.00 + 0 doi :10.l O67 / mge.2003.14 9
510

ERC is performed, generally with EST if stones are actually found. However, in series of patients defined as being at high risk for bile duct stones based on preliminary screening tests, the proportion shown by ERC to actually have choledocholithiasis has ranged b e t w e e n 27% and 50%. 1-7 Even with stricter use of preoperative criteria, the presence of stones was detected in only 66% of the patients, s Canto et al. 9 found that 70% of patients classified as being at high risk had choledocholithiasis; when classified as being at moderate, indeterminate, or low risk, bile duct stones were observed in 28%, 4%, and 0% of the patients. Therefore, the results of this triage strategy are unsatisfactory, and it increases the risk of overuse of ERC. 1~ The ideal alternative should have the same diagnostic accuracy as ERC but minimal or no invasiveness; available data indicate that EUS has these characteristics.6,s, 11-14 In fact, EUS has been shown to have a diagnostic accuracy for bile duct stones of about 95%, which compares favorably with ERC. 6,7,11,13 Because of the high US frequencies used (7.5, 12 MHz), EUS has a resolution of less t h a n 1 mm, m a k i n g it the best imaging technique available at present for imaging
VOLUME 57, NO. 4, 2003

GASTROINTESTINAL ENDOSCOPY

E U S for suspected choledocholithiasis: benefits vs. costs

E Buscarini, P Tansini, D Vallisa, et al.

the e x t r a h e p a t i c biliary tract. O t h e r a d v a n t a g e s of E U S over E R C are m i n i m a l i n v a s i v e n e s s or semii n v a s i v e n e s s with a l m o s t no p r o c e d u r e - r e l a t e d complications a n d a negligible failure rate.12,13,15 I n particular, no complications were r e p o r t e d in the several series of biliary E U S w h i c h comprise m o r e t h a n 1000 patients.6,8,9,11-14 M a n y i n v e s t i g a t o r s h a v e p o i n t e d out t h a t E U S m a y therefore p r e v e n t i n a p p r o p r i a t e invasive e v a l u a t i o n of the bile duct.S, 13 T h e s e d a t a p r o m p t e d us to a p p l y E U S as a firststep t e c h n i q u e after t r a n s a b d o m i n a l U S in a l a r g e series of consecutive p a t i e n t s w i t h s u s p e c t e d choledocholithiasis; p a t i e n t s w i t h bile d u c t stones at E U S w e r e e v e n t u a l l y r e f e r r e d for t r e a t m e n t w h e r e a s t h o s e w i t h n e g a t i v e E U S f i n d i n g s w e r e followed. T h e a i m of this p r o s p e c t i v e s t u d y w a s to a n a l y z e p a t i e n t o u t c o m e in t e r m s of effectiveness as well as t h e p o t e n t i a l clinical a n d economic benefits of E U S implicit in the avoidance of u n n e c e s s a r y ERC.

PATIENTS AND METHODS

All patients were enrolled in a prospective, ongoing protocol for evaluation of the clinical utility of EUS for the diagnosis of choledocholithiasis. Patients were enrolled if they met the following criteria for suspected choledocholithiasis: (1) history of biliary-type colicky pain or recent cholangitis, and a history of jaundice; (2) recent acute pancreatitis; (3) serum bilirubin and/or alkaline phosphatase or 7-glutamyl transpeptidase, and/or aminotransferases more than twice the upper normal limit; (4) dilatation of the intrahepatic and/or extrahepatic bile ducts (>7 mm) and/or a suspicion of choledocholithiasis on transabdominal US or CT. Patients who had a dilated bile duct at US and at least one clinical criterion were classified as being at intermediate risk; those with 3 or more criteria were classified as high risk. Exclusion criteria were as follows: refusal to participate, any factor that rendered the patient unsuitable for treatment of choledocholithiasis, and previous gastrectomy. Patients with a definite transabdominal US diagnosis of choledocholithiasis were also excluded. The US criterion for this diagnosis was an echo-rich image with acoustic shadowing within the bile duct. Patients with acute cholangitis and/or severe biliary pancreatitis with obstructive jaundice underwent ERC with EST. The protocol for the study was approved by the review board of our hospital. After US, each patient underwent EUS to identify either bile duct stones (echo-rich structures, possibly moving within the bile duct, with or without acoustic shadowing) or biliary sludge. In case of a positive EUS, the patient was referred for treatment of bile duct stones by EST or surgery depending on whether the gallbladder was present, clinical condition, and preference of the referring physician. EST or surgical intervention was performed within 5 days after EUS. The EUS result was considered a true-positive if stones were confirmed by EST or surgical exploration. In case of a negative EUS, patients with cholelithiasis underwent laparoscopic or open cholecystecVOLUME 57, NO. 4, 2003

tomy and were then followed; patients with a prior cholecystectomy and a negative EUS result were followed. Two clinical, biochemical (bilirubin, alkaline phosphatase, 7glutamyltranspeptidase, aminotransferases, serum amylase) and US evaluations were scheduled during the 3 months after EUS. Thereafter, at least 2 additional clinical and/or US evaluations were scheduled every 4 months. Any further available evaluation, including information obtained by telephone interview, was also recorded. A negative EUS result was considered true if the patient was symptom-free with normal biochemical tests and US imaging at a follow-up of at least 6 months. Patients were informed as to choices of management depending on EUS results; potential morbidity of endoscopic and surgical treatment of choledocholithiasis was explained. Written informed consent was obtained from all patients for the procedures performed and participation in the study. The study had 2 endpoints: (1) clinical outcome, based on the ability of EUS to prevent inappropriate invasive evaluation of the bile duct; (2) the economic implications of the EUS-strategy for suspected choledocholithiasis.

Procedures
EUS was performed with a radial scanning echoendoscope (GF-UM20, Olympus Optical Co., Ltd., Tokyo, Japan) at 7.5 and 12 MHz frequencies. All examinations were performed by one endoscopist (3 years experience with biliopancreatic EUS) with the assistance of one nurse. Patients were sedated by intravenous administration of diazepam (5-10 mg). The transducer was inserted to the second portion of the duodenum and gradually withdrawn to visualize the biliopancreatic region, in particular the main duodenal papilla, extrahepatic bile duct, cystic and hepatic ducts, and gallbladder. The number and size of stones and diameter of the bile duct were recorded. The mean duration of the examinations was 22 minutes. Each examination was recorded on videotape and hardcopy photographs were taken of the most significant findings. The mean cost of an EUS in our center was C 159. ERC with EST was performed with a standard video duodenoscope with the patient sedated as for EUS. All procedures were performed by a single endoscopist, different than the one who performed EUS, 2 nurse assistants, and a radiology technician. Antibiotics were administered prophylactically before the procedure and for 48 hours thereafter. After opacification of the bile duct, EST was performed and the findings recorded. Stones were cleared from the duct with Dormia basket and/or balloon catheter. The mean cost of ERC in our center was ~263.3. In patients undergoing surgical bile duct exploration, stones were cleared with a stone retrieval basket inserted by means of the cystic duct or by choledochotomy after intraoperative cholangiography to confirm the presence of stone(s).

Complications
Complications of endoscopic procedures were classified as follows: mild (unplanned hospital admission, prolongation of hospital stay for up to 3 days), moderate (4-10 days
GASTROINTESTINAL ENDOSCOPY

511

E Buscarini, P Tansini, D Vallisa, et al.

E U S for suspected choledocholithiasis: benefits vs. costs

ocha
,umo,

Percuta~eous i . . . . . . . . cholangiography :~, 2 persistent p ; n / ] ~

] r--o---n

2 negJ

i
3co

I !?=~

Figure 2. Endosonographic image of 4-mm stone within slightly dilated common bile duct. compared with the estimated cost of the ERC procedures that were avoided based on the number of patients with true negative EUS findings. The latter cost estimate was calculated by adding the following to the cost of an ERC: (1) the cost of hospitalization related to ERC (based on data from our department; mean hospitalization is 2 days for patients undergoing diagnostic ERC) and (2) costs associated with procedure-related morbidity calculated on the basis of a 2% major complication rate for diagnostic ERC with minimum prolongation of hospitalization of 11 days.17,!8

cky pa n/

'

I "4n.gat,v s I

la"o"o, u01

Figure 1. Outcomes for 463 patients with s pected choledocholithiasis who underwent EUS. ~, :i~
4

. . . . 3' . ~; hosp]tahzat]on), severe (>10 days hospltahzatlon or th~ need for surgical intervention or intensive c ~ ) , and fatal. 16

Statistical methods
Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of EUS were calculated. The mean cost (SD) for patients undergoing the EUSbased strategy was calculated taking into account the cost of the procedure, hospitalization, and complications. A one-sample t test was then used to compare this mean value with the hypothetical cost sustained by a patient undergoing ERC. A one-way sensitivity analysis was also performed, plotting mean costs of ERC (estimated cost) and EUS (actual cost) against the estimated risk of bile duct stones to determine the least costly diagnostic strategy as a function of the risk of bile duct stones (range 0% to 100%). This established the threshold (in terms of the risk of bile duct stones) for which the costs of the EUS and ERC strategies are identical. Statistical software was used for all calculations (SPSS 8.0, SPSS, Inc., Chicago, Ill.).

Cost analysis
The cost of making the diagnosis of choledocholithiasis was analyzed, excluding all costs of treatment. Without EUS, all patients with suspected choledocholithiasis would undergo ERC before EST; with use of EUS, ERC would be avoided in patients with negative findings. Because the benefit of EUS is expected to arise from the ability to avoid unnecessary ERC, the economic implications of its use are evaluated by weighing the costs of EUS performed on all patients against the costs of ERC for patients in the same population who avoided this procedure (corresponding to the true-negative rate for EUS). The cost for all EUS examinations was compared with the cost-savings realized by avoidance of ERC procedures. Three indicators were chosen to evaluate cost: (1) mean costs for EUS and ERC based on actual costs in our center and calculations suggested by the Italian Societies of Gastroenterology and Endoscopy (including costs for medical staff, nurse(s), disposable materials and drugs, and equipment amortization and maintenance); (2) hospitalization cost (one day in the'gastroenterology unit costs C309.8); and (3) costs associated with morbidity induced by the procedures calculated by adding charges for additional hospital days. In the analysis, the actual cost of EUS (procedure, days hospitalization directly linked to EUS, any morbidity) was
512

RESULTS
A t o t a l of 485 p a t i e n t s (202 m e n , 283 w o m e n ; m e a n a g e 66.2 y e a r s , r a n g e 20-94 y e a r s ) w e r e p r o s p e c t i v e l y s t u d i e d b e t w e e n J a n u a r y 1995 a n d D e c e m b e r 1998. O f these, 194 h a d u n d e r g o n e cholec y s t e c t o m y (2 d a y s to 48 y e a r s e a r l i e r ) . I n 25 patients, US demonstrated a normal gallbladder; t h e r e m a i n i n g 266 h a d g a l l s t o n e s b y US.
V O L U M E 57, NO. 4, 2003

GASTROINTESTINAL ENDOSCOPY

E U S for suspected choledocholithiasis: benefits vs. costs

E Buscarini, P Tansini, D Vallisa, et al.

Patient o u t c o m e s

The E U S findings were confirmed in 463 patients (Fig. 1). Twenty-two patients with negative E U S findings were lost to follow-up. In telephone interviews, these patients reported no symptoms of choledocholithiasis but had not undergone any followup evaluation for various reasons (absence of symptoms, difficulty in r e t u r n i n g to our center). Risk stratification for the likelihood of choledocholithiasis was as follows: i n t e r m e d i a t e risk, 287 patients (61.9%); high risk, 176 patients (38.1%). An EUS diagnosis of choledocholithiasis was made in 239 of the 463 patients (51.6%) (Fig. 2); absence of stones was found in 220 patients (87 with previous cholecystectomy). In 4 p a t i e n t s (0.8%), the E U S examination was incomplete (uncooperative patient, 1; pyloric stenosis, 2; duodenal bulb stenosis, 1). Bile duct stones were solitary in 87 cases. Mean stone d i a m e t e r was 7.04 mm (range 2-23 mm); stones were 1 cm or less in diameter in 182 cases (76%) and 5 mm or less in 97 cases (40%) (Fig. 3). The bile duct diameter was less than 7 mm in 58 cases (24%). E U S demonstrated cholelithiasis or gallbladder sludge in 18 of 25 patients in whom transabdominal US had found a normal gallbladder (Fig. 3). ERC with EST was performed in 212 p a t i e n t s and surgery with bile duct stone extraction in 27. Two hundred sixty-two patients underwent cholecystectomy; 113 of these cholecystectomies were performed in the 220 patients with negative E U S findings. Among patients who u n d e r w e n t EST, 156 (73%) had undergone EUS within the preceding 24 hours. Altogether the mean interval between E U S and EST was 1.6 days (range 0-4 days). EST and EUS results agreed in 204 cases. In 36 of these cases, stones were missed at ERC and found only after EST; in all of these cases the stones were less than 5 mm in diameter. In 2 patients, bile duct stones were not found after EST despite repeated passages of a Dormia basket and balloon catheter. In 6 patients (2.8%), EST was unsuccessful and the bile duct of these patients (all with gallstones) was explored at surgery. Thirty-nine patients u n d e r w e n t surgical exploration of the bile duct (7 laparoscopic cholecystectomy, 32 open surgery). The mean interval b e t w e e n EUS and surgery was 4.9 days (range 3-9 days). For 33 patients (including the 6 in whom EST was unsuccessful) the results of surgical exploration agreed with the EUS findings; in 2 patients with a negative EUS but persistent colicky pain or cholestasis, surgical exploration demonstrated stones. The 4 patients in whom EUS was unsuccessful underwent surgical exploration at which bile duct stones were found in 3. One patient in whom EUS revealed a choledochal
VOLUME 57, NO. 4, 2003

sludge

Figure 3. A, EUS image showing small quantity of biliary (arrowhead)within gallbladder, B, EUS image in same patient showing three 2-mm stones (arrowheads) within common bile duct. Transabdominal US in this patient who recently sustained an episode of acute pancreatitis was normal.

tumor underwent percutaneous transhepatic biliary drainage, which confirmed the EUS diagnosis. Two hundred seventeen patients with a negative EUS were followed: 3 underwent ERC/EST 12, 32, and 58 days, respectively, after EUS for episodic colicky pain associated with cholestasis in 2 cases and j a u n d i c e in the other. EST d e m o n s t r a t e d choledocholithiasis in 2 of these patients. For the remaining 214 patients, mean follow-up was 13.4 months (range 7-29 months). In 78 patients, biochemical tests of liver function (alkaline phosphatase, y-glutamyltransferase) were still elevated to 2 to 3 times normal values at the first follow-up evaluation. In all patients, the liver function tests returned to normal at 3 months' follow-up. None complained of symptoms or manifested any clinical, imaging, or biochemical abnormality during further follow-up.
GASTROINTESTINAL ENDOSCOPY
513

E Buscarini, P Tansini, D Vallisa, et al.

E U S for suspected choledocholithiasis: benefits vs. costs

Table 1. Cost analysis of the EUS-based strategy for the diagnosis of choledocholithiasis
EUS-based strategy No. of examinations Cost per procedure Procedure costs Hospitalization costs Complication cost Final cost Final cost/patientw Average saving per patient with EUS 463 EUS EUS (159) 73,650 99,780* 0 173,430 374.511 443.8 - 374.5 = 69.3 ERC spared 216 ERC ERC (263.3) 56,890 133,870t 14,7255 205,480 443.8t

Costs are expressed in Euros. *322 days of hospitalization for 463 patients who underwent EUS on different days (see Results). t2 days of hospitalization/patient undergoing ERC. 52% complication rate = 4.32 complications/216 patients, for a cost of 11 days of hospitalization for each complication. w costs divided by 463 patients. ]]Difference is statistically significant (p < 0.001).

Altogether choledocholithiasis was demonstrated in 244 (52%) of the 463 patients enrolled. Choledocholithiasis was found in 117 of 176 (66.4%) patients at high risk and 127 of the 287 (44.2%) at intermediate risk for choledocholithiasis. The EUS results were verified by EST (209 patients), surgical bile duct exploration (39), percutaneous cholangiography (1 patient in whom EUS demonstrated a choledochal tumor), and follow-up (214). EUS diagnoses were as follows: 237 true-positive, 216 true-negative, 2 false-positive, 4 false-negative, and 4 incomplete (sensitivity 98%, specificity 99%, positive predictive value 99%, negative predictive value 98%, accuracy 97% for the diagnosis of choledocholithiasis). In 214 patients (46%), more invasive evaluation of the bile duct was avoided. The mean duration of hospitalization for patients who underwent ERC/EST was 6.2 days (range 2-182).

negative. Severe pancreatitis complicated by abdominal fluid collections and multiorgan failure developed that resulted in extended hospitalization (3 days, gastroenterology department; 134 days, intensive care unit; 45 days, surgery department). The patient recovered, but was permanently unable to work. The patient eventually brought suit for recovery of damages and was awarded the sum of 129,114. The morbidity costs of ERC/EST in our study resulted from a total of 118 hospitalization days in the Gastroenterology Unit, 134 hospitalization days in the intensive care unit, and 45 in the surgery unit for a total cost of C 147,500, corresponding to a mean morbidity cost per ERC/EST of C686.05. If the monetary award of C 129,114 to compensate the patient who sustained the single severe complication is included in the calculation, the mean morbidity cost per ERC/EST is C 1286.5.

Complications
Complications were observed in patients who underwent EUS and EST on separate days, appearing after EST and attributable to that procedure. There was no procedure-related death. Among the 215 patients who underwent ERC/ EST, mild complications were noted in 25 (11.6%; 2 self-limiting episodes of bleeding, 23 instances of elevation of amylase level) with a mean prolongation of hospitalization of 1.8 days. Complications of moderate severity occurred in 7 patients (3.2%; 1 episode of bleeding requiring blood transfusion, 6 pancreatitis) with prolonged hospitalization a mean of 8.1 days. One (0.4%) severe complication occurred in a patient with a negative EUS who underwent ERC during follow-up, which proved
514 GASTROINTESTINAL ENDOSCOPY

Costs
The EUS-based strategy for the diagnosis of bile duct stones implied the avoidance of 216 ERC procedures, corresponding to the true negative EUS. The cost analysis therefore involved comparing the cost of the 463 EUS (~73,650) with the cost of the ERC procedures that were avoided (~56,890). Hospitalization costs were calculated in the EUS group as ~99,780 and estimated in the ERC group (on the basis of data from our department: 2 days hospitalization for a diagnostic ERC) as ~ 133,870. The cost of potential complications of ERC, taken as 14,725, was added to this sum, for a final total cost of ~ 205,480 (with a final cost per patient [total cost divided by 216] of ~951.2). Therefore, because our EUS strategy cost ~173,430, this management
V O L U M E 57, NO. 4, 2003

E U S for suspected choledocholithiasis: benefits vs. costs

E Buscarini, P Tansini, D Vallisa, et al.

implies a total saving of ~ 3 2 , 0 5 5 and, for 463 patients, an average saving of ~69.23 per patient (Table 1). The mean cost per patients for the EUSbased strategy was ~374.50 (+284.72). The theoretical mean cost per patient for the ERC-based strategy was ~443.80 (p < 0.001). One-way sensitivity analysis suggests t h a t the least costly diagnostic strategy changes as the risk of bile duct stones increases from 0% to 100%, and from 0% to 60.6% EUS is less costly than ERC (Fig. 4); the threshold (in terms of the risk of bile duct stones) for which the costs of the EUS and ERC strategies are identical is therefore 60.6%; and for all levels of risk greater than 60.6% ERC is least costly.
DISCUSSION

1400

1200

~
I

EUS

951.2

ERC

IO O 800

6OO

4OO
374.5

The selection of patients for ERC/EST by using commonly available, noninvasive diagnostic tests (i.e., clinical, biochemical, sonographic) has limited reliability, as confirmed by the present study. In a group of patients defined p r e i n t e r v e n t i o n as at intermediate or high risk, the proportion with choledocholithiasis was 52%. The current approach to patients with suspected bile duct stones is characterized by a high number of unnecessary ERC procedures. This procedure, until now, has been considered the s t a n d a r d s t u d y for imaging the bile duct. However, the complication rate associated with ERC is significant; there is an appreciable mortality rate as well. Large prospective case series have found overall complication rates of 5% to 6% and mortality rates of 0.07% to 0.1% after diagnostic ERC.18,19 A large multicenter study from Italy found a major complication rate of 1.3% and a mortality rate of 0.2% for diagnostic ERC. 2o Moreover, to arrive at a diagnosis of bile duct stones, a certain number of ERC procedures m u s t be completed with EST, which has a complication rate ranging b e t w e e n 7% and 10% 21,22 and a reported mortality rate of 0.2% to 2.2%. 22,23 Reported rates ofpancreatitis after ERC and EST ranged from 0% to 40%, although a figure of about 5% is typical. 21 Pancreatitis is an extremely undesirable complication t h a t can be severe and prolonged; this was the worst ERC-related complication in the present study, and it occurred in a patient who proved to be free of stones. Complications of EST usually develop immediately or shortly after the procedure, although long-term sequelae m a y also occur. In one long-term follow-up study, stenosis, new stones and nonobstructive cholangitis were found in 13% of patients 6 to 11 years after E S T . 24 In another study in which patients who had EST were followed for 15 to 17 years, a substantial proportion had bacterial contamination of the biliary
VOL'UME 57, NO. 4, 2003

200 .00

eo.5 RISK

100.00

Figure 4. One-way sensitivity analysis. Mean cost (expressed in euros) of EUS and ERC strategies versus risk from 0% to 100% of choledocholithiasis.

tree attributed to the permanent loss of the biliary sphincter. 25 These early and long-term sequelae reinforce the need to avoid inappropriate EST. Based on a s t u d y of claims for compensation of severe and fatal complications after diagnostic and therapeutic ERC, including 9 fatalities, Trap et al. 26 emphasized the need to better define the indications for ERC, especially because ERC was normal in 6 of the 9 fatal c a s e s . 26 An alternative to the c u r r e n t approach to the diagnosis of choledocholithiasis is thus needed, ideally one t h a t r e s u l t s in referral of only those patients with bile duct stones for ERC/EST. Studies comparing EUS and ERC have invariably concluded that a better method is needed to reduce uncertainty in the diagnosis of choledocholithiasis and that EUS has this potential and should therefore replace ERC for diagnostic purposes.6,S,10,12-14,27 To our knowledge, the present study is unique in that it is the first large study to use EUS in an algorithm for selection of patients for ERC/EST. Transabdominal US, the noninvasive procedure commonly used first in patients with extrahepatic cholestasis, has a low sensitivity for choledocholithiasis. 28 New imaging techniques, such as spiral CT and magnetic resonance cholangiography (MRC), offer a higher degree of safety than ERC when choledocholithiasis is suspected. Studies of the use of spiral CT reported t h u s far include only limited n u m b e r s of patients with results inferior to EUS (sensitivity 85%-88%, specificity 88%-97%, diagnosGASTROINTESTINAL ENDOSCOPY 515

E Buscarini, P Tansini, D Vallisa, et al.

E U S for suspected choledocholithiasis: benefits vs. costs

tic accuracy 8 6 % - 9 4 % ) . 29,30 A further drawback of CT is the need for a contrast medium. MRC, despite being completely noninvasive, has been compared with EUS in relatively few patients with suspected choledocholithiasis. Its methodology is not yet fully s t a n d a r d i z e d and various examination protocols exist, m a k i n g the interpretation of results somewhat difficult. EUS offers higher resolution (0.1 mm vs. 1.5 mm), which explains the lower sensitivity of MRC for small stones as compared with EUS. 31 Moreover, in contrast to EUS, some blind zones exist for MRC, notably the papillary and peripapillary region. Indeed, the diagnosis of choledocholithiasis by MRC is less satisfactory than by EUS. Sensitivity was only 40% in o n e s t u d y , 32 although in others the sensitivity has ranged from 81% to 92%, specificity from 94% to 98%, and accuracy is about 94%. 33-35 A comparison study of EUS versus MRC for the diagnosis of choledocholithiasis found better results for EUS (for EUS vs. MRC, respectively, specificity, 95% vs. 72%; positive predictive, 90% vs. 62%; negative predictive, 100% vs. 96%). 36 These investigators concluded that MRC can be used when there is a contraindication to EUS. EUS is thus the single best imaging technique for bile duct stones. It has important additional advantages. EUS is the only imaging technique capable of easily detecting bile duct sludge. Moreover, it permits identification of gallstones as well as microlithiasis missed by transabdominal US, findings of clinical relevance for management of unexplained acute pancreatitis. 37-39 E U S can detect pathologic disorders that are poorly evident (if at all) at ERCP. E U S before ERC also enhances the efficacy of the l a t t e r by encouraging the use of aggressive techniques when appropriate (e.g., precut papillotomy). The p r e s e n t s t u d y confirms the high diagnostic accuracy of EUS for detecting chole9 docholithiasis; accuracy is independent of bile duct stone size, being highly sensitive even for stones less than 5 mm in diameter. Few EUS procedures were unsuccessful. The safety of the procedure is also demonstrated in this large series, in which no EUSrelated complications were encountered. The major aim of the present study, a reduction in inappropriate invasive evaluation of the bile duct, was achieved in 46% of the patients. The study group included patients at intermediate to high risk for bile duct stones based on clinical parameters. The main clinical impact of EUS for the diagnosis of bile duct stones was, therefore, a substantial reduction in the use of ERC and consequently f e w e r ERC-related complications in patients with suspected bile duct stones. Thus, E U S should be applied not only in patients at low or intermediate risk (as previously
516

suggested) 13 because the percentage of high-risk patients who have stones is not always substantial. The EUS-based strategy in the present study was significantly more cost-effective t h a n a s t r a t e g y based on the use of ERC alone. In the study of Prat et al., s the cost of diagnostic options based on the preliminary use of EUS for the diagnosis of bile duct stones (as in the present study) was calculated in a hypothetical population. These investigators concluded that the EUS strategy was more expensive t h a n the ERC approach, even with a decrease in hospital stay for patients who h a d EUS. In t h a t study, the cost of the EUS procedure was considered to be greater than that for ERC. This discrepancy m a y reflect variables peculiar to a given country (e.g., duration of the examination, type of sedation, presence of an anesthetist). In our experience EUS for choledocholithiasis required an average of 22 minutes; mild sedation was generally sufficient. Moreover, in the present study calculations for ERCrelated costs were conservative: the estimate of costs related to ERC complications applies only to the major untoward events, and takes into account only the cost of minimal prolongation of hospital stay (11 days) and not the additional costs of possible surgical treatment, intensive care, blood transfusion, and other forms of support. Potential mortality and legal expenses were not factored into the cost calculations. Moreover, an intangible cost, impossible to quantify, yet arguably the most important, is the significant loss in quality of life arising from complications, usually transient, although in some cases resulting in permanent disability. 4~ If, compared with the conservative estimate of ERC costs, the E U S - b a s e d s t r a t e g y proved even moderately cost-effective, the economic advantage of EUS would be consistently greater in most clinical scenarios. According to our sensitivity analysis the cost of the EUS and ERC strategies would be identical when the risk of choledocholithiasis is 60.6%, which is greater than proportion of patients with stones generally found by use of clinical predictors. 4-7 However, the sensitivity analysis results depend on the cost data used. These costs pertain to our center; actual values will likely differ in other countries and different health systems. Other critical factors that influence cost are the ERC complication rate for the individual center and performance of ERC on an outpatient basis. The figures used in the p r e s e n t study are believed to reasonably represent Italian and European settings. Optimally, E U S and ERC, as indicated by the EUS findings, should be performed on the same day. Thus, with proper organization EUS need not prolong hospital stay; moreover, in some cases EUS can
VOLUME 57, NO. 4, 2003

GASTROINTESTINAL ENDOSCOPY

E U S for suspected choledocholithiasis: benefits vs. costs

E Buscarini, P Tansini, D Vallisa, et al.

be performed on an outpatient basis. From a technical standpoint, patients with suspected bile duct stones should undergo EUS in specialized centers where there is adequate experience with both EUS and ERC/EST. The accuracy of EUS clearly depends on operator skill, b u t such procedural skills are equally crucial for performance of ERC. 41 Use of EUS as a preliminary diagnostic imaging s t u d y in patients with suspected bile duct stones avoids a considerable number of unnecessary invasive procedures. This substantial clinical advantage is achieved at no additional cost, but, on the contrary, with cost savings.
REFERENCES
1. Abboud PA, Malet PF, Berlin JA, Staroscik R, Cabana M, Clarke JR, et al. Predictors of bile duct stones prior to cholecystectomy: a meta-analysis. Gastrointest Endosc 1996:44:450-9. 2. Onken JE, Brazer SE Eisen GM, Williams DM, Bouras EP, DeLong ER, et al. Predicting the presence of choledocholithiasis in patients with symptomatic choledocholithiasis. Am J Gastroenterol 1996;91:762-7. 3. Roston AD, Jacobson IM. Evaluation of the pattern of liver tests and yield of cholangiography in symptomatic choledocholithiasis: a prospective study. Gastrointest Endosc 1997;45:394-9. 4. Erikson RA, Carlson B. The role of endoscopic retrograde cholangiopancreatography in patients with laparoscopic cholecystectomies. Gastroenterology 1995;109:252-63. 5. Tham TCK, Lichtenstein DR, Vandevoort J, Wong RCK, Brooks D, Van Dam J, et al. Role of endoscopic retrograde cholangiopancreatography for suspected choledocholithiasis in patients undergoing laparoscopic cholecystectomy. Gastrointest Endosc 1998;47:50-6. 6. Amouyal P, Amouyal G, Levy P, Tuzet P, Palazzo L, Vilgrain V, et al. Diagnosis of choledocholithiasis by endoscopic ultrasonography. Gastroenterology 1994;106:1062-7. 7. Sugiyama M, Atomi Y. Endoscopic ultrasonography for diagnosing choledocholithiasis: a prospective comparative study with ultrasonography and computed tomography. Gastrointest Endosc 1997;45:143-6. 8. Prat F, Amouyal G, Amouyal P, Pelletier G, Fritsch J, Choury AD, et al. Prospective controlled study of endoscopic ultrasonography and endoscopic retrograde cholangiography in patients with suspected bile duct lithiasis. Lancet 1996;347:75-9. 9. Canto MIF, Chak A, Stellato T, Sivak MV Jr. Endoscopic ultrasonography versus cholangiography for the diagnosis of choledocholithiasis. Gastrointest Endosc 1998;47:439-48. 10. Sahai AV, Mauldin PD, Marsi V, Hawes RH, Hoffman BJ. Bile duct stones and laparoscopic cholecystectomy: a decision analysis to assess the roles of intraoperative cholangiography, EUS and ERCP. Gastrointest Endosc 1999;49:334-43. 11. Buscarini E, Buscarini L. The role of endosonography in the diagnosis of choledocholithiasis. Eur J Ultrasound 1999;10:117-26. 12. Shim CS, Hoo JH, Park CW, Kim YS, Lee JS, Lee MS, et al. Effectiveness of endoscopic ultrasonography in the diagnosis of choledocholithiasis prior to laparoscopic cholecystectomy. Endoscopy 1995;27:428-32. 13. Palazzo L, Girollet PP, Salmeron M, Silvain C, Roseau G, Canard JM, et al. Value of endoscopic ultrasonography in the diagnosis of bile duct stones: comparison with surgical exploration and ERCP. Gastrointest Endosc 1995;42:225-31. 14. Aubertin JM, Levoir D, Bouillot JL, Becheur H, Bloch F, Aouad K, et al. Endoscopic ultrasonography immediately
VOLUME 57, NO. 4, 2003

prior to laparoscopic cholecystectomy: a prospective evaluation. Endoscopy 1996;28:667-73. 15. RSsch T, Dittler HJ, Fockens P, Yasuda K, Lightdale C. Major complications of endoscopic ultrasonography: results of a survey of 42,105 cases [abstract]. Gastrointest Endosc 1993;39:AB370. 16. Cotton PB. Outcomes of endoscopy procedures: struggling towards definition. Gastrointest Endosc 1994;40:514-8. 17. Erickson RA, Carlson B. The role of endoscopic retrograde cholangiopancreatography in patients with laparoscopic cholecystectomies. Gastroenterology 1995;109:252-63. 18. Cotton PB, Jowell PS, Baillie J. Spectrum of complications after diagnostic ERCP and effect of comorbidities [abstract]. Gastrointest Endosc 1994;40:18. 19. Davis WZ, Cotton PB, Arias A. ERCP and sphincterotomy in the context of laparoscopic cholecystectomy: academic and community practice patterns and results. Am J Gastroenterol 1997;92:597-601. 20. Loperfido S, Angelini G, Benedetti G. Major early complications from diagnostic and therapeutic ERCP: a prospective multicenter study. Endoscopy 1999;31:125-30. 21. Freeman ML, Nelson DB, Sherman S, Haber GB, Herman ME, Dorsher PJ, et al. Complications of endoscopic biliary sphincterotomy. N Engl J Med 1996;335:908-18. 22. Sherman S, Ruffolo TA, Hawes RH, Lehman GA. Complications of endoscopic sphincterotomy. Gastroenterology 1991; 101:1068-72. 23. Lambert ME, Betts CD, Hill J, Faragher EB, Martin DF, Tweedle DEF. Endoscopic sphincterotomy: the whole truth. Br J Surg 1991;78:473-6. 24. Hawes RH, Cotton PB, Vallon AG. Follow up 6 to 11 years after duodenoscopic sphincterotomy for stones in patients with prior cholecystectomy. Gastroenterology 1990;98:1008-12. 25. Bergman JG, vanBerkel AM, Groen AK, Schoeman MN, Offerhaus J, Tytgat GNJ, et al. Biliary manometry, bacterial characteristics, bile composition, and histologic changes fifteen to seventeen years after endoscopic sphincterotomy. Gastrointest Endosc 1997;45:400-5. 26. Trap R, Adamsen S, Hart-Hansen O, Henriksen M. Severe and fatal complications after diagnostic and therapeutic ERCP: a prospective series of claims to insurance covering public hospitals. Endoscopy 1999;31:125-30. 27. Canto M. Endoscopic ultrasonography and gallstone disease. Gastrointest Endosc 1996;43:$37-43. 28. Vilgrain V, Palazzo L. Choledocholithiasis: role of US and endoscopic ultrasound. Abdom Imaging 2001;26:7-14. 29. Neitlich JD, Topazian M, Smith RC, Gupta A, Burrell MI, Rosenfield AT. Detection of choledocholithiasis: comparison of unenhanced helical CT and endoscopic retrograde cholangiopancreatography. Radiology 1997;203:753-7. 30. Polkowski M, Palucki J, Regula J, Tilszer A, Butruk E. Helical computed tomographic cholangiography versus endosonography for suspected bile duct stones: a prospective blinded study in non-jaundiced patients. Gut 1999;45:744-9. 31. Lambert R. Clinical outcome of EUS in biliary diseases. Endoscopy 2000;32:558-61. 32. Scheiman JM, Carlos RC, Barnett JL, Elta GH, Nostrant TT, Chey WD, et al. Can endoscopic ultrasound or magnetic resonance cholangiopancreatography replace ERCP in patients with suspected biliary disease? A prospective trial and cost analysis. Am J Gastroenterol 2001;96:2900-4. 33. Guibaud L, Bret PM, Reinhold C, Atri M, Barkun AN. Bile duct obstruction and choledocholithiasis: diagnosis with MR cholangiopancreatography. Radiology 1995;197:109-15. 34. Holzknecht N, Gauger J, Sackmann M, Thoeni RF, Schurig J, Holl J, et al. Breath-hold MR cholangiography with snapshot
GASTROINTESTINAL ENDOSCOPY
517

E Buscarini, P Tansini, D VaUisa, et al.

EUS for suspected choledocholithiasis: benefits vs. costs

techniques: prospective comparison with endoscopic retrograde cholangiography. Radiology 1998;206:657-64. 35. Materne R, van Beers BE, Gigot JF, Jamart J, Geubel A, Pringot J, et al. Extrahepatic biliary obstruction: magnetic resonance imaging compared with endoscopic ultrasonography. Endoscopy 2000;32:3-9. 36. deLedinghen V, Lecesne R, Raymond JM, Gense V, Amouretti M, Drouillard J, et al. Diagnosis of choledocholithiasis: EUS or magnetic resonance cholangiography? A prospective controlled study. Gastrointest Endosc 1999;49:26-31. 37. Dill JE, CaUls J, Berkhouse L, Evans P, Martin D, Palmer ST. Combined endoscopic ultrasound and stimulated biliary drainage in cholecystitis and microlithiasis--diagnoses and outcomes. Endoscopy 1995;27:424-7.

38. Chak A, Hawes RH, Cooper GS, Hoffman B, Catalano MF, Wong RCK, et al. Prospective assessment of the utility of EUS in the evaluation of gallstone pancreatitis. Gastrointest Endosc 1999;49:599-604. 39. Liu CL, Lo CM, Chan JK, Poon RT, Lain CM, Fan ST, et al. Detection of choledocholithiasis by EUS in acute pancreatitis: a prospective evaluation in 100 consecutive patients. Gastrointest Endosc 2001;54:325-30. 40. Savader SJ, Lillemoe KD, Prescott CA, Winick AB, Vanbrux AC, Lund GB, et al. Laparoscopic cholecystectomy-related bile duct injuries. A health and financial disaster. Ann Surg 1997;225:268-73. 41. Baillie J. ERCP training for the few, not for all. Gut 1999;45: 9-10.

A v a i l a b i l i t y of J o u r n a l b a c k i s s u e s
As a service to our subscribers, copies of back issues of Gastrointestinal Endoscopy for the preceding 5 y e a r s are m a i n t a i n e d a n d a r e available for p u r c h a s e from Mosby until i n v e n t o r y is depleted. P l e a s e write to Mosby, Subscription C u s t o m e r Service, 6277 S e a H a r b o r Dr., Orlando, F L 32887 or call 800-654-2452 or 407-345-4000 for i n f o r m a t i o n on a v a i l a b i l i t y of p a r t i c u l a r issues a n d prices.

518 GASTROINTESTINAL ENDOSCOPY

VOLUME 57, NO. 4, 2003