The Legs For Life Screening for Peripheral Vascular Disease: results of a prospective study designed to improve patient compliance with physician recommendations.

PURPOSE: To determine how compliance with recommendations made by physicians during the 2000 Legs For Life National Screening for Peripheral Vascular Disease (PVD) and Leg Pain is affected through the use of (i) simple and concise patient information and recommendation cards and (ii) a "targeted" postscreening follow-up plan. MATERIALS AND METHODS: Patients were initially screened for PVD by completion of the Legs For Life Risk Factor Assessment form and determination of bilateral ankle/brachial indexes (ABIs). Each patient then met with an interventional radiologist or vascular surgeon. Patients with normal ABIs (>1.0 bilaterally) or mildly abnormal ABIs (<1.0 but >0.90) were classified as having no risk and low risk, respectively. Patients with ABIs of 0.70-0.89 were classified as having moderate risk for PVD and patients with ABIs <0.69 were classified as having high risk for PVD. Physicians reviewed the Risk Factor Assessment form with each patient and made specific lifestyle improvement recommendations. For the year 2000 screening, patients classified at moderate and high risk for PVD received special instructions and a card containing clearly printed information on the purpose of the Legs For Life screening, their level of risk for PVD, specific recommendations for follow-up, and phone numbers to call to help arrange for that follow-up. Two weeks after the screening, a second copy of this card was mailed to each moderate- and high-risk assessed patient. Four months later, each of these patients was contacted by telephone to determine if they had pursued additional care or testing. RESULTS: A total of 185 patients were screened, 42 (23%) of whom were determined to be at moderate or high risk for PVD. Four months after the screening, 39 (93%) of these patients were available for follow-up. Twenty (51%) patients had received no further medical advice or treatment. Nineteen (49%) patients had pursued further medical care which included physician consultation (n = 19; 100%), noninvasive Doppler evaluation (n = 10; 26%), diagnostic arteriography (n = 2; 5%), initiation of pharmacologic therapy for claudication (n = 1; 3%), percutaneous intervention (n = 1; 3%), or vascular surgery (n = 1; 3%). Seventeen of 39 patients (44%) reported that claudication-type leg pain was still a concern and/or lifestyle-limiting problem. CONCLUSION: Patients can be provided with problem-focused information and succinct physician recommendations at and soon after a screening for PVD, which can contribute to enhanced patient compliance. However, a host of personal, social, health, and physician-related issues still prevent a large percentage of patients from achieving relief of PVD-associated leg pain.

Safety of conscious sedation in interventional radiology.

PURPOSE: To identify rates of adverse events associated with the use of conscious sedation in interventional radiology. METHODS: In a 5-month period, prospective data were collected on patients undergoing conscious sedation for interventional radiology procedures (n = 594). Adverse events were categorized as respiratory, sedative, or major adverse events. Respiratory adverse events were those that required oral airway placement, ambu bag, or jaw thrust. Sedation adverse events were unresponsiveness, oxygen saturation less than 90%, use of flumazenil/naloxone, or agitation. Major adverse events were hypotension, intubation, CPR, or cardiac arrest. The frequency of adverse events for the five most common radiology procedures were determined. RESULTS: The five most common procedures (total n = 541) were biliary tube placement/exchange (n = 182), tunneled catheter placement (n = 135), diagnostic arteriography (n = 125), vascular interventions (n = 52), and other catheter insertions (n = 46). Rates for respiratory, sedation, and major adverse events were 4.7%, 4.2%, and 2.0%, respectively. The most frequent major adverse event was hypotension (2.0%). Biliary procedures had the highest rate of total adverse events (p < .05) and respiratory adverse events (p < .05). CONCLUSION: The frequency of adverse events is low with the use of conscious sedation during interventional procedures. The highest rates occurred during biliary interventions.

Treatment of hemodialysis catheter-associated fibrin sheaths by rt-PA infusion: critical analysis of 124 procedures.

PURPOSE: To prospectively evaluate the efficacy of a low-dose, 3-hour infusion of recombinant tissue plasminogen activator (rt-PA) for the treatment of hemodialysis catheter (HDC)-associated fibrin sheaths. This report expands the authors' experience with this technique over that previously reported. MATERIALS AND METHODS: Fifty-five patients with end-stage renal disease (38 women, 17 men) undergoing catheter-directed hemodialysis treatment were evaluated for 124 episodes of HDC dysfunction. This patient group had a mean age of 57 years and an age range of 23-92 years. Radiographic contrast studies and/or clinical evaluation were consistent with the presence of a fibrin sheath on the arterial and/or venous port in all cases. Each patient underwent a thrombolytic infusion consisting of 2.5 mg rt-PA in 50 mL normal saline at 17 mL/h (3-hour infusion) per port. All infusions were performed in the interventional radiology recovery room on an outpatient basis. Patients were followed prospectively for technical success, complications, catheter patency, and long-term outcome. RESULTS: The technical success rate, defined as return of effortless manual aspiration and infusion capability from both ports followed by at least one successful dialysis session, was 91%. No patient was excluded from rt-PA therapy because of contraindications, and the procedure-related complication rate was zero percent. A Kaplan-Meier survival analysis yielded primary patency rates at 30, 60, 90, and 120 days of 0.55, 0.36, 0.25, and 0.15 (SE <.10), respectively; secondary patency rates at 60, 120, 180, and 240 days were 0.70, 0.46, 0.30, and 0.27 (SE <.10), respectively (P < 001). At the end of the study period, all 52 surviving patients continued to undergo catheter-directed hemodialysis and 34 (65%) were using the same catheter present at the time of entrance into the study. Of the 18 patients (35%) requiring catheter exchange, 16 (89%) did for persistent malfunction after rt-PA therapy, one (5.5%) for infection, and one (5.5%) for a fractured hub. CONCLUSION: Thrombolytic therapy with use of a 2.5-mg rt-PA infusion through each port over a 3-hour period would appear to be a safe method for treating HDC-associated fibrin sheaths. Immediate return of catheter function is achieved in most patients, obviating more invasive techniques. Primary patency rates are relatively short, but catheters that fail can be retreated, resulting in secondary patency rates that are substantial and significantly improved.

The Legs for Life screening for peripheral vascular disease: compliance with physician recommendations in moderate- and high-risk assessed patients.

PURPOSE: To determine compliance within a community with recommendations made by physicians during the 1999 Legs for Life National Screening for Peripheral Vascular Disease (PVD) and Leg Pain. MATERIALS AND METHODS: Patients were initially screened for PVD by completion of the Legs for Life risk factor questionnaire and determination of bilateral ankle/brachial indexes (ABIs). Each patient subsequently met with an interventional radiologist or vascular surgeon. Patients with normal ABIs (>1.0 bilaterally) or mildly abnormal ABI(s) (<1.0 but >0.90) were classified at no and low risk for PVD, respectively; patients with ABI(s) of 0.70-0.89 were classified at moderate risk for PVD; and patients with ABI(s) <0.69 were classified at high risk for PVD. Risk factors for PVD were assessed by the consulting physician and discussed with all patients. Recommendations were made for additional evaluation and/or follow-up care, if necessary. Seven months after screening, patients who were assessed at moderate and high risk for PVD were contacted by telephone to determine if they had pursued additional care or testing. RESULTS: A total of 205 patients were screened for PVD, 48 (23%) of whom were determined to be at moderate to high risk. Forty-four (92%) patients were available for follow-up. At 7 months after screening, 31 (70%) patients had received no further medical advice or treatment. Thirteen (30%) of these patients had completed a follow-up appointment, but only three with a physician specializing in peripheral vascular disease. None of the patients had clinical follow-up with an interventional radiologist. Five (11%) patients had undergone noninvasive Doppler evaluation and one (2%) had undergone diagnostic arteriography. No patient had undergone any form of percutaneous or surgical intervention. CONCLUSION: Patient compliance with physician recommendations after outpatient screening for PVD is low. The Legs for Life screening program could be considered successful in that it provides for patient education and the identification of moderate to high-risk patients. Physicians participating in this program may have to modify their approach to patient screening and follow-up if a concomitant goal is to deliver specialty care.

Hemodialysis catheter-associated fibrin sheaths: treatment with a low-dose rt-PA infusion.

PURPOSE: To prospectively evaluate the efficacy of a low-dose, 3-hour recombinant tissue plasminogen activator (rt-PA) infusion for the treatment of hemodialysis catheter (HDC)-associated fibrin sheaths. MATERIALS AND METHODS: Seventeen patients with end-stage renal disease (female, n = 11; male, n = 6), who were undergoing catheter-directed hemodialysis, were evaluated for 28 episodes of HDC dysfunction. This patient group ranged in age from 25 to 92 years (mean, 57 years). Radiographic contrast and/or clinical evaluation were consistent with the presence of a fibrin sheath on either the arterial and/or venous port in all cases. Patients subsequently underwent a thrombolytic infusion consisting of 2.5 mg rt-PA in 50 mL normal saline at a rate of 17 mL/h (3-hour infusion) per port. All infusions were performed in the interventional radiology recovery room, on an outpatient basis. Patients were followed-up prospectively for technical success, complications, catheter patency, and long-term outcome. RESULTS: The immediate technical success rate, defined as return of manual aspiration and infusion capabilities to both ports, was 100%. No potential patients required exclusion from thrombolytic therapy secondary to contraindications, and no procedure-related complications occurred. The arithmetic mean and median catheter patency at the end of the study was 41 and 25 days, respectively (range, 1-116 days). A Kaplan-Meier survival analysis yielded a 30-, 60-, and 90-day probability of patency of 0.67, 0.61, and 0.51, respectively. At the end of the study period, all 17 patients remained on catheter-directed hemodialysis and 13 (76%) were utilizing the same catheter present at the time of entrance into the study. CONCLUSION: Thrombolytic therapy utilizing a 2.5-mg rt-PA infusion through each port during a 3-hour period would appear to be a safe, efficient method for treating HDC-associated fibrin sheaths. Three-month patency rates are comparable to those reported for other methods of restoring function to HDC catheters, including new catheter placement, catheter exchange over a guide wire, thrombolytic infusions with urokinase, and percutaneous fibrin sheath stripping.

Percutaneous radiologic placement of peritoneal dialysis catheters: long-term results.

PURPOSE: To evaluate the efficacy of percutaneous radiologic placement of peritoneal dialysis (PD) catheters. MATERIAL AND METHODS: Nineteen patients with end-stage renal disease were referred for percutaneous radiologic placement of a commercially available PD catheter. The patient group consisted of 11 men and eight women with a mean age of 58 years (range, 38-85 y). All procedures were performed on an outpatient basis with use of conscious sedation. Patients were followed up to determine technical success, complication rate, catheter survival (continuous patency without infection or mechanical dysfunction), and long-term outcome. RESULTS: The technical success rate for catheter placement was 95% (18 of 19). Complications included one (5%) failed attempt at catheter placement and one (5%) case of postplacement peritonitis. Mean and median catheter survival durations were 320 and 289 days, respectively (range, 33-823 d). A Kaplan-Meier survival analysis yielded 6-, 12-, and 24-month probabilities of catheter survival of 0.89 +/- 0.15, 0.81 +/- 0.20, and 0.81 +/- 0.20, respectively. Long-term outcomes included: continued PD, n = 9 (47%); death, n = 6 (32%; all related to comorbid disease); successful renal transplant, n = 2 (10.5%); and continued or resumed hemodialysis, n = 2 (10.5%). CONCLUSION: Percutaneous radiologic placement of PD catheters has a high technical success rate, low complication rate, and can be performed on an outpatient basis. Catheter survival is comparable to that achieved with surgical and percutaneous endoscopic methods of catheter placement.

Forearm loop, upper arm straight, and brachial-internal jugular vein dialysis grafts: a comparison study of graft survival utilizing a combined percutaneous endovascular and surgical maintenance approach.

PURPOSE: To determine rates and duration of patency achievable in forearm loop, upper arm straight, and brachial-internal jugular (IJ) vein hemodialysis grafts utilizing a combined percutaneous endovascular and surgical maintenance approach. MATERIALS AND METHODS: A retrospective analysis of 74 hemodialysis grafts (forearm loop, n = 22; upper arm straight, n = 34; and brachial-IJ vein, n = 18) in 50 patients with end-stage renal disease was conducted. Operative notes, interventional procedural reports, and hospital records were used to construct a history for each of these grafts from the time of surgical placement until the time the graft was abandoned for an alternative method of dialysis. All procedures performed to maintain and/or restore patency during the usable lifetime of the grafts were documented. RESULTS: Survival analysis using the Kaplan-Meier method demonstrated the following probabilities of primary patency at 6, 12, and 16 months, respectively: forearm loop graft = .46, .26, and .26; upper arm straight graft = .39, .22, and .16; and brachial-IJ vein graft = .19, .06, and .06 (forearm loop vs upper arm straight grafts, P > .05; forearm loop and upper arm straight vs brachial-IJ grafts, P < .001, P < .001, respectively). The probabilities of secondary patency at 12, 24, and 48 months, respectively, were: forearm loop graft = .89, .30, and NA; upper arm straight graft = .52, .35, and .17; and brachial-IJ vein graft = .54, .42, and .21 (P < .05 for all three comparisons: forearm loop > brachial-IJ > upper arm straight). Six percutaneous and two surgical procedures were compared and no significant differences in utilization were determined among the three graft types (ANOVA, P range, .38-.88). CONCLUSION: Kaplan-Meier analysis determined the probability of primary patency for forearm loop grafts to be similar to that for upper arm straight grafts, and both were significantly greater than for brachial-IJ vein grafts. The secondary patency rates for forearm loop grafts are greater than for upper arm and brachial-IJ vein grafts, while that for the brachial-IJ vein graft is greater than the upper arm straight graft. Utilization of interventional and surgical resources required to maintain patency do not significantly differ among the three types of upper extremity hemodialysis grafts.

The biliary manometric perfusion test and clinical trial--long-term predictive value of success after treatment of bile duct strictures: ten-year experience.

PURPOSE: To evaluate the long-term predictive value of the biliary manometric perfusion test and clinical trial for determining patency after treatment of bile duct strictures. MATERIALS AND METHODS: One hundred four patients with benign biliary disease were treated with surgical (n = 59) or percutaneous (n = 45) techniques followed by intubation with large-caliber silicone stents. Prior to removal of the biliary stents, patients underwent a biliary manometric perfusion test (n = 168) and/or a clinical trial (n = 105) to objectively and subjectively evaluate the treated site for potential long-term patency. The patients were followed up for 1-87 months, and clinical outcomes were determined. Kaplan-Meier survival curves were generated for three patient groups, including those who (i) passed either test, (ii) failed either test, and (iii) failed either test, were re-treated, and passed either test. RESULTS: Final successful treatment outcomes were documented in 47 (92%) surgically and 31 (86%) percutaneously treated patients, respectively (P = .001). The Kaplan-Meier survival curves determined the probability of patency at 0, 2, 4, and 6 years after treatment to be 1.0, .96, .78, and .59, respectively, after passing a biliary manometric perfusion test; and 1.0, .91, .78, and .78, respectively, after passing a clinical trial (P > .10). The probability of patency at 4 years after treatment was .45 after failing a biliary manometric perfusion test, and at 6 months was zero after failing a clinical trial (P = .001 and .001, respectively, vs the same test in the passing group). Seventy-nine percent of patients who failed either test required an additional period of repeated stent placement or reoperation. After repeated treatment, the probability of patency at 0-4 years was .80 and .88, respectively, for the biliary manometric perfusion test and clinical trial (P > .05 and P > .10, respectively, vs same test in the group that passed). CONCLUSION: Patients who initially pass either the biliary manometric perfusion test or clinical trial have a significantly increased probability of patency versus those who fail; however, patients who fail either test and who then receive definitive additional treatment have a similar probability of patency as those who initially pass. Although the log rank test demonstrated the Kaplan-Meier survival curves from the biliary manometric perfusion test and clinical trial not to be significantly different in any of the three groups (ie, passing, failing, re-treated), the biliary manometric perfusion test is recommended over the clinical trial because of its simplicity, immediate results, and predicted cost savings.

Entrapment of J-tip guidewires by Venatech and stainless-steel Greenfield vena cava filters during central venous catheter placement: percutaneous management in four patients.

We present four patients in whom bedside placement of a central venous catheter was complicated by entrapment of a J-tip guidewire by a previously placed vena cava (VC) filter. Two Venatech filters were fragmented and displaced into the superior VC or brachiocephalic vein during attempted withdrawal of the entrapped wire. Two stainless-steel Greenfield filters remained in place and intact. Fluoroscopically guided extraction of both wires entrapped by Greenfield filters was successfully performed in the angiography suite.

Single-specimen bile cytology: a prospective study of 80 patients with obstructive jaundice.

PURPOSE: To determine the sensitivity, specificity, and charges associated with single-specimen bile cytologic study in patients with obstructive jaundice. MATERIALS AND METHODS: Eighty consecutive patients with presumed malignant biliary strictures underwent percutaneous biliary drainage (PBD). Cytologic evaluation was performed on a single bile specimen from each patient collected at the time of the PBD. Final diagnoses were obtained from either percutaneous (n = 14) or surgical (n = 66) histologic specimens (gold standard). Both data sets were then compared to determine the sensitivity and specificity of bile cytology. The charges associated with bile cytodiagnosis were compared to those for other biopsy procedures utilized in the same setting. RESULTS: Eighty bile specimens were obtained with a mean of 14 mL (range, 3-65 mL) per patient with 79 (99%) specimens adequate for cytologic processing. Eleven (13%) specimens were acellular. The overall sensitivity was 15% and specificity was 100%; these values were not dependent on the volume of the bile specimen (P > .10) or type of malignancy (P = .10). For bile cytodiagnosis, the mean charge was $160 and the successful biopsy rate (true-positive plus true-negative results/total number procedures) was 27%. CONCLUSION: Single-specimen bile cytology has a low sensitivity; however, because of its convenience, simplicity, atraumatic nature, and low relative charge versus comparable procedures, it may be useful as an adjunct to PBD in patients with suspected malignant biliary disease.

Guide wire directed manipulation of malfunctioning peritoneal dialysis catheters: a critical analysis.

PURPOSE: To evaluate patency rates after guide wire directed manipulation of malfunctioning continuous ambulatory peritoneal dialysis (CAPD) catheters. MATERIALS AND METHODS: During a 58-month period, 23 patients underwent 34 outpatient guide wire directed manipulations of their CAPD catheter to improve function (n = 30) or reduce pain and improve function (n = 4) during dialysis. Catheter patency rates were subsequently determined by review of departmental, hospital, and dialysis center charts; procedural reports; and patient telephone interviews. RESULTS: Among 12 patients who underwent a single guide wire directed manipulation, long-term (> 30 days) catheter patency was achieved in seven (58%). With use of the Kaplan-Meier survival method, the 3-, 6-, and 12-month probability of patency after a single guide wire manipulation was 0.61, 0.54, and 0.11, respectively. The mean duration of patency achieved in this group was 131 days (range, 2-421 days). In those patients (n = 8) who underwent multiple catheter manipulations (n = 19), 11 (58%) procedures resulted in long-term patency, with each patient (100%) achieving at least one such period. The Kaplan-Meier survival method determined the probability of patency in this group at 3, 6, and 12 months to be 0.75, 0.69, and 0.54, respectively. The mean secondary catheter patency was 235 days (range, 2-646 days). Overall, 75% of patients followed up achieved at least one period of long-term catheter patency during the time of this study. One (3%) episode of postprocedure peritonitis occurred. CONCLUSION: Guide wire directed CAPD catheter manipulation is a relatively simple outpatient procedure that restores long-term catheter function for most patients with minimal risk for a major complication. Patients with nonfunctioning CAPD catheters who do not have peritonitis or sepsis will most likely benefit from at least one attempt at radiologic manipulation of their catheter.

Major bile duct injuries during laparoscopic cholecystectomy. Follow-up after combined surgical and radiologic management.

OBJECTIVE: The authors provide the results of follow-up evaluation after combined surgical and radiologic management of 89 patients with major bile duct injuries during laparoscopic cholecystectomy. SUMMARY BACKGROUND DATA: The incidence and mechanism of injury of major bile duct injuries during laparoscopic cholecystectomy has been clearly defined. Furthermore, a number of series have described the management of these injuries by surgical, endoscopic, and radiologic techniques with excellent short-term results. Long-term follow-up data, however, are lacking in the management of these injuries. METHODS: Data were collected prospectively on 89 patients treated at a single institution with major bile duct injuries after laparoscopic cholecystectomy managed between July 1, 1990, and July 1, 1996. Patients referred with injuries underwent early percutaneous transhepatic cholangiography and biliary drainage. Based on the cholangiographic appearance and clinical situation, patients were managed by either percutaneous balloon dilatation or surgical reconstruction with a Roux-en-Y hepaticojejunostomy with transanastomotic stenting. Follow-up was obtained by personal interview during October 1996. RESULTS: Two patients died without an attempt at definitive therapy. Both deaths were caused by sepsis and multisystem organ failure present at the time of transfer to the authors' institution. The remaining 87 patients were managed initially by either balloon dilatation (N = 28) or surgical reconstruction (N = 59). Ten patients have not completed treatment and still have biliary stents in place. Evaluation of 25 patients completing treatment after balloon dilatation (mean follow-up, 27.8 months) showed a success rate of 64%. Evaluation of 52 patients completing treatment after surgical reconstruction (mean follow-up, 33.4 months) showed a success rate of 92%. All failures were managed successfully by either surgical reconstruction or balloon dilatation. CONCLUSIONS: Major bile duct injuries can be managed successfully by combined surgical and radiologic techniques. This series provides, for the first time, significant follow-up on a large number of patients with overall success rates of 64% after balloon dilatation and 92% after surgical reconstruction. The combination of surgery and balloon dilatation resulted in a successful outcome in 100% of patients treated.

Laparoscopic cholecystectomy-related bile duct injuries: a health and financial disaster.

OBJECTIVE: This study was designed to evaluate the total costs associated with repair of laparoscopic cholecystectomy (LC)-related bile duct injuries. SUMMARY BACKGROUND DATA: The popularity of LC with both patients and surgeons is such that this procedure now exceeds open cholecystectomy by a ratio of approximately 4 to 10:1. However, costs associated with LC-related injuries, particularly regarding treatment patterns, have up to now not been explored fully. METHODS: The complete hospital and interventional radiology (IR) billing records for 49 patients who have completed treatment for laparoscopic cholecystectomy-related bile duct injuries were divided into 8 categories. These records were totaled for comparison of costs between patient groups that experienced different injuries and treatment patterns. RESULTS: Patients with LC-related bile duct injuries were billed a mean of $51,411 for all care related to repair of their bile duct injury. Patients incurred an average of 32 days of inpatient hospitalization and 10 outpatient care days. Postoperative treatment included long-term chronic biliary intubation averaging 378 days. Two patients (4%) died as a result of their LC-related complications. Patients with bile duct injuries that were recognized immediately at the time of the initial surgery ultimately experienced a total cost for their repair and hospitalization of 43% to 83% less than for patients in whom recognition of the injury was delayed (p < 0.019 to 0.070). In addition, the total hospitalization and outpatient care days was reduced by as much as 76% with early recognition of an iatrogenic injury. CONCLUSIONS: Repair of cholecystectomy-related bile duct injuries can run 4.5 to 26.0 times the cost of the uncomplicated procedure and carries a significant mortality rate. Intraoperative recognition of such an injury with immediate conversion to an open procedure for definitive repair can result in significant cost savings and relates directly to a decreased morbidity, mortality, length of hospitalization, and number of outpatient care days.

Volumetric evaluation of blood flow in normal renal arteries with a Doppler flow wire: a feasibility study.

PURPOSE: To evaluate the feasibility of direct intravascular determination of renal artery (RA) blood flow with a Doppler probetipped guide wire. MATERIALS AND METHODS: Potential renal donors (n = 10) with normal RAs (n = 23) underwent evaluation of RA blood flow velocity with use of a 0.018-inch, 12-MHz Doppler guide wire. The RA average peak velocity (APV) was obtained with the flow wire. RA diameter was obtained from the filmed images with magnification corrected to a known standard or by a computerized quantification program. These data were used to determine the vessel's cross-sectional area (CSA). RESULTS: The right and left RA APV, CSA, and blood flow differed insignificantly within the group and averaged 9.7 and 9.0 cm/sec (P = .43), 0.417 and 0.357 cm2 (P = .22), and 382 and 370 mL/min (P = .43), respectively. However, in individuals, the RA CSA and total volumetric blood flow varied by a mean of 29% (range, 4%-56%) and 50% (range, 19%-128%), respectively. CONCLUSION: This study demonstrates that direct intravascular determination of RA blood flow with a Doppler-tipped wire is both feasible and relatively uncomplicated. Results indicate that blood flow can vary significantly, both in kidneys within the same individual and from person to person. The Doppler wire may facilitate measurements of RA blood flow during endoluminal interventions and help determine an optimal endpoint for these procedures.

Biliary fistulae and haemorrhage.

Percutaneous biliary drainage is the most common aetiology of haemobilia. Bile duct fistulae can also arise from the hepatic or portal vein, most commonly as a result of trauma. Percutaneous methods for treating haemobilia from all these sources are discussed in detail.