In vitro study and in vivo application of a reusable double-channel sphincterotome.

BACKGROUND AND STUDY AIMS: Therapeutic endoscopic retrograde cholangiopancreatography (ERCP) has been deemed to be a "cost-prohibitive" procedure, based upon the cumulative costs of one-time-use accessories and current reimbursement plans. One-time-use sphincterotomes comprise a significant component of that cost and, accordingly, we evaluated the disability and clinical usefulness of a recently introduced reusable double-channel sphincterotome. MATERIALS AND METHODS: We studied a reusable 6-Fr sphincterotome at baseline and following contamination with 10(6) Bacillus stearothermophilus. Reprocessing included a unique 30-minute ultrasonic cleaning step in lieu of manual cleaning, followed by steam sterilization. Parameters evaluated included sphincterotome function, electrical integrity, and our ability to sterilize the devices for three in vitro trials. In vivo studies included patient demographics and outcomes, procedural findings, and success rates, and the mean number of times the sphincterotome was used, functional grading at time of use, and reasons for sphincterotome malfunction. RESULTS: Ten out of ten sphincterotomes maintained form, function, and electrical integrity in vitro, and all cultures were negative after sterilization. In the initial in vivo study, ten sphincterotomes were used in 50 patients (mean, 5 uses) with a 94% success rate. Reasons for sphincterotome failure included leak or breakage of the accessory port in 70%, wire fracture in 10%, incorrect wire bow in 10%, and clogged injection port in 10%. Following reconfiguration of the insertion-port polymer, an additional ten sphincterotomes were used in 110 patients (mean, 11 uses). Mechanical failure occurred primarily at the wire-insertion port, resulting in progressive friction with reuse. There were neither electrical nor infectious complications associated with reuse. CONCLUSIONS: A reusable double-channel sphincterotome is available which can theoretically be reprocessed and sterilized without the manual cleaning step of the reprocessing process. Contingent upon both provider and patient, multiple reuse can be anticipated, and contingent upon purchase price and reprocessing costs, the potential for procedural cost savings is significant.

Reusable biopsy forceps: a prospective evaluation of cleaning, function, adequacy of tissue specimen, and durability.

BACKGROUND: Recent studies have indicated that reusable biopsy forceps remain contaminated after reprocessing and can only be used a mean of 12 to 25 times without malfunction. Because this contradicts traditional endoscopic practice, our study investigated the ability to sterilize a type of commercially available biopsy forceps and prospectively evaluated their function in vivo until malfunction and/or breakage. METHODS: Thirty reusable biopsy forceps were studied, 15 of which were contaminated for 5 trials each with 10(6) Bacillus stearothermophilus, and 15 of which were prospectively evaluated clinically over an 18-month period (9/98-3/00). Contaminated forceps were reprocessed by using a standard protocol and placed in a sterile bag containing soy broth. The latter was passed through a 0.2 micron filter and was subsequently cultured. In vivo data included biopsy site, size, adequacy, problems obtaining a biopsy specimen, and reasons for ultimate forceps failure. RESULTS: After contamination, all biopsy forceps yielded a heavy growth of B stearothermophilus. No forceps, including 5 that were piecemeal dismantled with a wire cutter, had residual bacteria after reprocessing. In the in vivo study, 1507 biopsy sessions were undertaken in 1339 procedures. Forceps were categorized as new or like-new in 1259 of 1339 (94%) procedures, some loss of function but usable in 72 of 1339 (5.4%), and inadequate function or broken at use in 8 of 1339 (0.6%). Histologically, 1501 specimen sets were adequate (99.6%) and mean specimen size was 2.7 +/- 0.1 mm. Mechanical problems were noted in only 38 of 1507 (3%) sessions to include such things as sticky forceps, and the mean number of uses to malfunction or breakage was 91 +/- 15 (SEM) (range 19-132). CONCLUSIONS: This reusable biopsy forceps can be sterilized and used a mean of 91 times with adequate tissue sampling. Mechanical problems were minor to time of breakage. Contingent on acquisition and reprocessing costs as well as the number of procedures performed, this reusable forceps has the potential for significant cost savings.

Overview of the problem: reprocessing versus disposal of endoscopic accessories.

The decision to reuse or not reuse endoscopic accessories appears simple on the surface. Further reflection, however, reveals that the decision to purchase reusable accessories or reuse disposable accessories is complex and controversial. This article clarifies key issues surrounding the use of disposable and reusable endoscopic accessories. Discussion includes the advantages and disadvantages of disposable and reusable accessories, the issues surrounding reprocessing of reusable equipment, and the issues surrounding the reuse of disposable accessories.

Use of a double-channel duodenoscope for diagnostic and therapeutic ERCP.

BACKGROUND: A double-channel duodenoscope has the potential to shorten ERCP procedure time or improve procedure success rates because tasks can be done in parallel through variably situated accessory channels. METHODS: We prospectively evaluated a prototype double-channel duodenoscope in 102 patients, recording findings, procedural success, and potential advantages or problems associated with the instrument. RESULTS: Both channels were used in 79 cases (77%) and a single-channel in the remainder. Ninety-nine percent (101 of 102) of diagnostic and 95% (87 of 92) of therapeutic ERCPs were successful. Instrument advantages included decreased procedure time (23%), cannulation/procedural ease (16%), and miscellaneous (6%). Disadvantages were noted in 15% of the patients and were related to instrument diameter and diameter of the accessory channels. CONCLUSIONS: Additional study of second generation dual-channel duodenoscopes appears warranted.

In vitro and in vivo evaluation of a reusable double-channel sphincterotome.

BACKGROUND: To date, one reusable, double-channel sphincterotome has been approved by the Food and Drug Administration in the United States. Whether this device can be reprocessed easily and whether it is more durable than currently manufactured disposable sphincterotomes are uncertain. METHODS: Thirty double-channel, 20 mm, braided-wire sphincterotomes approved for multiple uses were studied in vitro/in vivo with regard to durability and sterilization. A cost analysis of reusable, disposable, and reprocessed disposable sphincterotomes was also carried out. RESULTS: Three of 10 sphincterotomes evaluated in vitro broke after 3, 4, and 8 uses. Electrical integrity was preserved after 10 uses in the remaining sphincterotomes. Nine sphincterotomes remained functional for at least 3 uses, five for 6 uses, and one for 10 uses. Culture results after inoculation demonstrated contamination with surviving organisms after manual cleaning and no growth after ethylene oxide sterilization. Sixty-one procedures were performed in vivo using 20 sphincterotomes (mean number of uses 3.1). No evidence of procedurally related infection occurred with reuse. Cost per use of this reusable sphincterotome was calculated to be $62.98; it became cost effective after 2.2 and 7.9 uses when compared with disposable and reprocessed, disposable sphincterotomes, respectively. CONCLUSIONS: This reusable sphincterotome proved to be safe, easily sterilized, and electrically intact after repeated use. In vivo, however, a progressive loss of function limited the mean number of uses to 3.1. In settings that preclude reuse of reprocessed disposable accessories, this reusable sphincterotome may provide a means to decrease costs associated with endoscopic retrograde cholangiopancreatography.

Recurrent pancreatitis in Gardner variant familial polyposis: etiology, diagnostic approach, and interventional results.

HYPOTHESIS: Pancreatitis arising from an obstructing ampullary neoplasm in patients with Gardner variant familial polyposis is an infrequently described clinical entity. We reviewed all patients with Gardner variant polyposis presenting with pancreatitis during a 12-year period in our institution, hoping to better define etiology and the appropriate diagnostic and interventional approach. METHODS: A retrospective record review (1986-1998) defined patient demographics, presenting features, initial and subsequent endoscopic retrograde cholangiopancreatography (ERCP) findings, subsequent treatments, and both immediate and long-term outcomes. Particular consideration was given to initial post-ERCP diagnosis and to endoscopic interventions undertaken at that time. We also looked at those patients who eventually required surgical intervention after long-term failure of medical and endoscopic therapy, the indications for surgery, final pathological characteristics, and follow-up results. RESULTS: Eight patients (6 women and 2 men), with a mean age of 42 years at initial presentation, were found. Each patient was known to have Gardner variant familial polyposis at the time of the initial bout of pancreatitis. All had undergone prior colectomy and 4 of 8 had undergone prior cholecystectomy. None were known to be taking medications or ingesting pancreatoxic substances. Five of 8 patients had obstructing focal or diffuse adenomatous disease involving the ampulla. Two of 8 patients had pancreatitis attributed to other causes (divisum, stones) and a single patient had no clear etiology. Three of 5 patients with ampullary adenomatous disease underwent pancreaticoduodenectomy for recurrent adenomatous encroachment and ampullary stenosis, despite repetitive snare resection and papillotomy. All of these patients had ampullary and other duodenal adenomas, and none had malignant disease. CONCLUSIONS: Patients presenting with pancreatitis in the setting of Gardner variant familial polyposis will frequently have an obstructing ampullary neoplasm, although additional etiologies should be sought. Initial endoscopic therapy affords transient relief but may not be definitive. The abnormal scarring and fibrosis (keloid formation, desmoid reaction) that characterize this disease likely play a large role in endoscopic or subsequent surgical failure. A significant number of these patients will go on to require surgical referral and intervention.

In vitro evaluation of integrity and sterilization of single-use argon beam plasma coagulation probes.

OBJECTIVE: Argon plasma coagulation probes (APC) are currently marketed in the United States as single-use items, and may constitute a significant per-procedure expense. It is unknown whether these probes can be sterilized after endoscopic use and if electrical integrity can be maintained after reprocessing. METHODS: Ten probes (2.3 mm diameter, 220 cm length) manufactured by ERBE Inc., (Marietta, GA) were studied using the ERBE APC 300 at 60 watts. Baseline coagulation depth was measured by coagulating a piece of beefsteak for 60 s. Probes were contaminated with 10(6) Bacillus subtilis spores, cultured, and manually cleaned. Culturing involved introducing 10 cc sterile water through the probes; water was filtered, plated onto blood agar, and incubated for 48 h. After ethylene oxide (ETO), the probes were cultured to determine sterilization. Finally, the per-procedure cost of each probe was assessed. RESULTS: Ten of 10 probes completed 10 testing sessions. One probe split at the proximal end but remained functionally intact. Electrical integrity remained intact for all 10 sessions. All probes grew too numerous to count colonies of B. subtilis after inoculation and no B. subtilis was detected after ETO sterilization. Assuming 10 uses clinically, a total per-procedure equipment cost would approximate $24.00, whereas per-procedure probe cost would equal $42.66 if only five uses were obtained in vivo. CONCLUSIONS: The combination of manual cleaning and ETO sterilization consistently sterilized APC probes. Ninety percent of the probes showed no sign of physical deterioration and 100% maintained their electrical activity after 10 uses. APC probes can potentially be safely and effectively reused up to 10 times, and a significant procedural savings is possible with reuse.

Do age, gender, or tumor location affect outcomes when using metallic stents in the palliative treatment of esophageal carcinoma?

UNLABELLED: Metallic esophageal stents are marketed as easier to insert with fewer placement and subsequent stent-related problems. Accordingly, we retrospectively/prospectively reviewed our experience placing metallic esophageal stents in malignant strictures. METHODS: The research design was a one-group pre- and post-test quasi-experimental design including all patients who underwent metallic stent placement at Virginia Mason Medical Center in Seattle between August 1987 and August 1997. Data included descriptive statistics of patient demographics, pre- and post-dysphagia scores, acute and subacute complications, repeated interventions, and patient survival outcomes. Descriptive statistics, paired t-test assessing pre- and post-dysphagia scores, and 2-way ANOVA were used to analyze the statistical relationship between complications and 1) stent location, 2) patient gender, and 3) patient age. FINDINGS: Seventy-five metallic stents were placed in 63 patients. Improvement in pre- and post-dysphagia scores was statistically significant. Factor analysis was not statistically significant for complications and stent location, complications and gender, and complications and age. CONCLUSIONS: Metallic esophageal stents showed statistically significant improvement for patient dysphagia. Complications of stent placement did not correlate with location of stent, age, or gender.

Reuse of disposable sphincterotomes for diagnostic and therapeutic ERCP: a one-year prospective study.

BACKGROUND: Although previous studies have documented that single-channel sphincterotomes can be sterilized and used up to 10 times, no studies have prospectively evaluated the reuse of disposable sphincterotomes. METHODS: Five- and six-French double-channel sphincterotomes marketed as one-time-use only were manually cleaned, sterilized with ethylene oxide, and used up to 10 times. Over a 1-year period, total number of uses as a cannulation or cutting device, reason for sphincterotome dysfunction, possible infectious complications were evaluated, and cost-savings associated with reuse were evaluated. RESULTS: The mean number of times sphincterotomes could be used in our institution was 3.4; a broken or stiff cutting wire was the main reason for discard. There were two infectious complications during the study period. Both were experienced by patients with unrelieved obstruction treated with new sphincterotomes. Contingent on the method used, total savings related to sphincterotome reuse was approximately $66,000. CONCLUSIONS: Double-channel sphincterotomes marketed as one-time-use items can be reused safely when properly cleaned. Reuse was associated with significant cost savings in our institution.

Risk of contamination of sterile biopsy forceps in disinfected endoscopes.

BACKGROUND: Previous studies have shown that pathogens may persist within bacterial biofilms in endoscope accessory channels despite high-level disinfection. Breaching the gastrointestinal mucosa with biopsy forceps contaminated at time of passage has the potential to cause cross-infection between patients. METHODS: We studied contamination risk of sterilized biopsy forceps passed through endoscopes after high-level disinfection. For each trial, five video colonoscopes, duodenoscopes, and gastroscopes were used. All endoscopes had been previously processed and stored for 10 or more hours. Sterile biopsy forceps were inserted and retrieved followed by vortexing the tips in 15 mL of soy broth. Under a laminar flow hood, the broth was filtered through a 0.2 microm millipore membrane and plated. Because of minimal bacterial growth resulting from the above, soy broth (> 20 mL) was flushed through two video colonoscopes, duodenoscopes, and gastroscopes on two occasions and collected. The effluent was plated using a sample of 0.1 mL dilution. The remaining suspension was passed through a millipore filter, and the filter was cultured. All cultures were incubated more than 48 hours. RESULTS: Biopsy forceps underwent a total of 24 anaerobic and 75 aerobic cultures. Microbacterial growth occurred on 17 plates: 7 from gastroscopes, 5 from colonoscopes, and 5 from duodenoscopes. Fifteen plates grew staphylococcus for a total of 21 colonies, 1 plate grew 1 colony of propionibacter, 2 plates grew diphtheroids for a total of 4 colonies, and 1 plate grew a single colony of lactobacillus. Cultures from soy broth flushed through the various endoscopes grew on 5 plates: 3 from gastroscopes and 2 from duodenoscopes grew a total of 8 colonies of staphylococcus. CONCLUSIONS: With proper cleaning technique, a 20-minute soak in 2% glutaraldehyde is effective in disinfecting endoscopes. Although current procedures for endoscope disinfection remain imperfect, we found that in this clinical setting, infection of pathogenic gastrointestinal flora is unlikely when using sterile biopsy forceps.

Expandable metallic prostheses for malignant obstructions of gastric outlet and proximal small bowel.

BACKGROUND: Data are limited on use of expandable metal stents for treatment of malignant gastric outlet obstruction. Accordingly, we report our experience using these stents to palliate malignant obstructions of the gastric outlet, duodenum, and proximal jejunum. METHODS: Eight patients with malignant strictures causing gastric obstruction underwent endoscopy with fluoroscopic guidance to delineate tumor borders and length followed by expandable metallic prosthesis placement (Wallstent, Z-Stent, Ultraflex, and Endocoil). RESULTS: Symptoms were relieved in seven patients, five of whom had previous surgeries (Whipple, Billroth II, esophagojejunostomy, and gastrojejunostomy) for malignancy. One patient underwent surgical resection of a presumed malignant stricture containing a previously placed Wallstent after a 45-pound weight gain. CONCLUSIONS: Expandable metallic prostheses placed in patients with malignant obstruction of the gastric outlet, duodenum, or proximal jejunum, before or after surgery, effectively palliate obstructive symptoms and may also serve to improve nutrition.

Use of the 25 mm flanged esophageal Z stent for malignant dysphagia: a prospective multicenter trial.

BACKGROUND: An initial multicenter study using a 21 mm flanged esophageal Z stent demonstrated excellent palliation but an 11% immediate complication rate at placement and a 27% migration rate at 1 month. This North American multicenter trial prospectively studied a 25 mm flanged Z stent to define its palliative ability and whether the increased diameter affected placement or migration problems. METHODS: Fifty patients who had esophageal Z stents at seven university or regional referral hospitals were prospectively studied. Indications for prosthesis placement, previous therapy, patient demographics, incidence of concomitant tracheoesophageal fistula, and degree of dysphagia were defined, as were procedural and subsequent stent-related problems, survival times, the ability to occlude a tracheoesophageal fistula, and subsequent degree of dysphagia. RESULTS: Twenty-four patients had infiltrating malignancy (16 exophytic and 10 extrinsic), 9 of whom had concomitant tracheoesophageal fistulas. Ten patients (20%) had misplaced stents requiring retrieval and replacement, 12 patients (24%) had subsequent stent-related problems including exsanguination (2), aspiration (3), tumor overgrowth (3), and postplacement migration (4) (8%). There was statistically significant improvement in prestent versus poststent dysphagia and two thirds of patients had complete occlusion of their tracheoesophageal fistula. CONCLUSIONS: Redesign of the esophageal Z stent has decreased the migration rate without increasing placement or subsequent erosion problems. Its efficacy appears comparable to the currently marketed Z stent for the palliation of malignant dysphagia and occlusion of tracheoesophageal fistula.

Transpapillary stenting for pancreaticocutaneous fistulas.

Because transpapillary stents have been successfully placed to treat the ductal disruptions associated with pseudocysts, pancreatic ascites and pleural effusions, and pancreaticoenteric fistulas, we reviewed our experience with endoscopically placed prostheses in patients who had persistent pancreaticocutaneous fistulas but an otherwise intact duct. Nine patients who underwent endoscopic transpapillary stent placement for ongoing pancreaticocutaneous fistulas at our institution were retrospectively reviewed. Fistulas were present for a mean (+/-SEM) of 35 +/- 11 days and averaged 225 +/- 55 ml of output daily. Etiology of the fistulas included percutaneous pseudocyst drainage in four patients, pancreatic necrosis in two, complications of pancreatic surgery in two, and perforation of the duct of Santorini at the time of minor sphincterotomy in one. All patients had an otherwise intact duct at the time of endoscopic retrograde cholangiopancreatography. Six patients had transpapillary stents placed that did not bridge the area of leakage and three had prostheses placed across the ductal disruption. Eight of nine fistulas were successfully closed by means of this technique including five within 48 hours. There was one instance of stent migration and one patient developed prosthesis occlusion and an infected pseudocyst, which was treated with stent exchange. Stents were retrieved 10 to 14 days after fistula closure and no patient has had a recurrence at a median follow-up of 3 years. Transpapillary stents appear to effect closure of pancreaticocutaneous fistulas that fail to respond to conventional therapy.

Expandable stents: unusual locations.

OBJECTIVES: Expandable metallic prostheses have been used widely for malignant biliary stenoses and are being used increasingly for malignant dysphagia and esophago-airway fistulas. Potentially, such prostheses can be placed for gut neoplasms obstructing beyond the esophagus or gastric cardia. This series reports our experience with expandable metallic stents in the stomach, jejunum, and colon. METHODS: All patients with expandable gastrointestinal stents (other than biliary tree or esophagus) were reviewed. Indications for stent placement, type and location of prosthesis, patient demographics, procedural problems, and data with regard to outcome were defined. RESULTS: Over a 6-yr period, expandable prostheses (Z stent, Esophacoil, and Ultraflex) were placed in nine patients with widespread malignancy (afferent loop obstruction, three; colon obstruction, three; gastric outlet obstruction, two; and esophagojejunal interposition stricture, one). There was one colon perforation as a consequence of dilation prior to placement of the prosthesis. Subsequent problems included migration (one), occlusion by food bolus (one), and tumor overgrowths/ingrowths (three) which were treated with laser. Median survival approximated 9 months. CONCLUSIONS: Expandable metallic prostheses have the potential to palliate malignant gastrointestinal stenoses that have traditionally been treated with surgical bypass or comfort care measures only.

In vitro evaluation of wire integrity and ability to reprocess single-use sphincterotomes.

BACKGROUND: Sphincterotomes are currently marketed as one-time-use items and constitute considerable cumulative expense in a busy endoscopy unit. It is uncertain whether these accessories can be safely reprocessed without loss of form and function. METHODS: We studied disposable sphincterotomes (five 5F, five 6F) in vitro as to their durability, electrical integrity, and ability to be adequately cleaned both manually and with ethylene oxide after contamination with 10(5) to 10(6) Mycobacterium chelonei. RESULTS: Seven of the 10 sphincterotomes withstood the rigors of reuse; three 6F sphincterotomes developed wire fracture between four and eight uses. Electrical integrity, as measured by an electrosurgical analyzer, remained intact up to time of breakage in all sphincterotomes. Manual cleaning followed by glutaraldehyde soak resulted in residual mycobacterial colonies in five 6F sphincterotomes and a single 5F sphincterotome. No instrument had residual organisms cultured following manual cleaning and ethylene oxide sterilization. CONCLUSIONS: The authors conclude that one-time-use sphincterotomes have the potential for safe reuse.

Prospective trial using Virtual Vision as distraction technique in patients undergoing gastric laboratory procedures.

To determine whether distraction techniques using Virtual Vision, an audiovisual system integrated into a modified pair of glasses, improves tolerance to routine gastric laboratory procedures, a prospective study of 50 patients was done. Patients were allowed to view/hear a travelogue tape during testing. Patients and registered nurses, who performed the procedure, were asked their interpretation of Virtual Vision as a device that improved patient tolerance, impaired the patient's ability to follow directions, or impaired the nurse's ability to perform the procedure. Patients graded the procedure from 1 (unbearable) to 5 (reasonably comfortable) and were asked whether they would use the device for a repeat procedure. For those who had undergone previous procedures, patients were queried regarding their preference for routine testing or Virtual Vision. Improved tolerance to procedures was noted in 82% of the patients as perceived by nurses and patients. Eighteen percent of the nurses and 8% of the patients thought that the auditory insert and glasses impaired the patient's ability to cooperate with the test, whereas 88% and 86%, respectively, thought that Virtual Vision was a valuable distraction technique. Using an ordinal scale for testing tolerance, patients graded their test at a mean of 3.8 (SD 1.25) and 82% expressed a desire to use the system in conjunction with future testing. Twenty-six of 33 patients (79%) who had undergone previous gastric laboratory testing preferred distraction techniques using Virtual Vision to conventional testing. Distraction techniques using a system such as Virtual Vision have the potential to improve patient tolerance to routine gastric laboratory studies.

Complications associated with laparoscopic anti-reflux surgery: one multispecialty clinic's experience.

BACKGROUND: The records of all patients with significant complications of laparoscopic anti-reflux surgery (LARS) seen at our institution between June 1993 and September 1996 were reviewed. Specifically excluded were patients who had mild perioperative complications or postoperative dysphagia that either did not require bougienage or responded to one to two dilations. Data collected included patient demographics; type of surgery; complication and its presentation; response to medical, endoscopic, and/or surgical therapy; and outcomes. RESULTS: Nine patients (five men and four women, mean age 59 years) presented at a mean of 4 months post-LARS (seven Nissen fundoplications and two Hill posterior gastropexy repairs). Symptoms included refractory dysphagia (4), intractable gas bloat (5), various degrees of chest or abdominal pain (4), and incapacitating diarrhea (3). Findings included tight repair (4), vagal nerve injury with gastroparesis (4) and/or diarrhea (3), and esophageal or gastric perforation (3). Additional findings included proximal gastric ulcers (3), volvulus (1), and incarcerated intrathoracic hernia (1). Patients with dysphagia had moderate improvement after a mean of 3 +/- 0.6 (standard error of the mean) additional dilations, whereas four patients required five reoperations to date. Seven of the nine patients had clinical improvement, whereas two had refractory symptoms at a mean follow-up of 10 months. CONCLUSIONS: Although LARS has supplanted open surgery in many centers, refractory complications occur in a subset of patients and require reoperation in approximately one half of those referred to a multispecialty clinic.

Prospective multicenter trial of esophageal Z-stent placement for malignant dysphagia and tracheoesophageal fistula.

BACKGROUND: Conventional esophageal prosthesis placement has been associated with a 6% to 8% perforation rate and numerous postplacement complications. Expandable esophageal stents have been developed to preclude the above but there are few studies that have prospectively defined clinical results and subsequent stent-related complications. METHODS: All patients who underwent esophageal Z-stent placement at nine university or referral hospitals were prospectively assessed. Data collected included patient demographics, acute and subacute placement problems, the ability to occlude airway fistulas, prestent and poststent dysphagia scores, and patient survival. RESULTS: Fifty-four of 56 patients (96%) with refractory dysphagia or malignant esophagoairway fistulae had 73 Z-stents successfully inserted. Initial distal deployment occurred in 13% of the patients and an additional 17% required balloon dilation to achieve maximal diameter. Acute placement complications occurred in 11% of patients and included severe pain (3), bleeding from necrotic tumor (2), and hiatal hernia intussusception (1). No perforations occurred. Eight of 11 patients (73%) had complete tracheoesophageal fistula occlusion and mean dysphagia score (+/- SD) improved from 2.6 (0.7) to 1.1 (1.2) (p < 0.01). Fifteen stents (27%) had delayed migration at a mean of 1 month and 3 required surgery for retrieval. Three patients had ultimate stent erosion resulting in bleeding in 2 (exsanguination 1) or fistula (treated with a conventional stent). CONCLUSIONS: The authors conclude that esophageal Z-stents can be placed safely and successfully in the majority of patients. The tendency of distal deployment during placement and subsequent migration problems at a time distant from placement in a patient subset deserve attention and are currently being addressed.