Systematic review of transgastric ERCP in Roux-en-Y gastric bypass patients.

Balloon-assisted endoscopic retrograde cholangiopancreatoscopy (ERCP) in Roux-en-Y gastric bypass (RYGB) patients is technically challenging due to anatomic and accessory constraints, thus success rates are modest. Transgastric ERCP (TG-ERCP) offers a viable alternative. We aimed to systematically review the literature on TG-ERCP in RYGB patients to better define the technical approaches, success rates, and adverse events of this procedure. A computer-assisted search of the Embase and PubMed databases was performed to identify studies that focused on the techniques and clinical outcomes of TG-ERCP. Two investigators independently identified studies and abstracted relevant data. The literature search yielded 26 eligible studies comprising 509 TG-ERCP cases. Access to the excluded stomach to facilitate ERCP was achieved laparoscopically in 58% of reported cases, via open surgery (6% of reported cases), by antecedent placement of a percutaneous gastrostomy tube (33%), or with endoscopic ultrasound assistance (3%). Successful gastric access was reported in 100% of cases and successful ductal cannulation in 98.5%. Adverse events were reported in 14% of cases; 80% of these were related to gastrostomy creation and the rest were attributable to ERCP. Wound infections (n = 19, 3.7%) were the most common gastrostomy-related adverse event, and post-ERCP pancreatitis (n = 7, 1.4%) was the most common ERCP-related adverse event. No deaths were reported. Based on existing observational studies, TG-ERCP appears to be a safe and highly effective approach in patients with RYGB anatomy. Additional research and clinical experience are needed to more precisely define the risk-benefit ratio and optimal technique of TG-ERCP.

Laparoscopic repair of gastro-duodenal fistula secondary to band erosion.

Laparoscopic gastric banding is one of the most common surgical treatments for morbid obesity performed worldwide. The procedure, however, has many well-documented risks and complications, including band erosion. We present here a gastric banding patient who was referred to our tertiary care centre after secondarily forming an entero-enteric fistula with complaints of pain, nausea, vomiting and severe reflux. She was successfully treated with laparoscopic dissection and due to her existing anatomy, and the patient's desire for continued weight loss, she was converted to Roux-en-Y gastric bypass.

Food cravings among bariatric surgery candidates.

PURPOSE: Food cravings are common, more prevalent in the obese, and may differ in those who pursue surgical treatment for obesity. Food craving tools are most often validated in non-clinical, non-obese samples. METHODS: In this retrospective study, 227 bariatric surgery candidates at a large medical center completed the Food Cravings Questionnaire-Trait (FCQ-T). The aim was to explore the factor structure of the FCQ-T. RESULTS: Principal components analysis with varimax rotation revealed a seven-factor structure that explained 70.89 % of the variance. The seven factors were: (1) preoccupation with food, (2) emotional triggers, (3) environmental cues, (4) loss of control, (5) relief from negative emotions, (6) guilt, and (7) physiological response. The preoccupation with food factor accounted for 49.46 % of the variance in responses. CONCLUSIONS: Unlike other populations, food cravings in bariatric surgery candidates appear to be related most to preoccupations with food.

Executive control circuitry differentiates degree of success in weight loss following gastric-bypass surgery.

OBJECTIVE: While overall success rates of bariatric surgery are high, approximately 20% of patients either regain or never lose the expected amount of weight. The purpose of this study was to determine whether, after gastric-bypass surgery, the degree of weight loss can be differentiated based on the neural response to food cues. DESIGN AND METHODS: In this functional MRI study, 31 post-surgical patients viewed food and neutral images in two counterbalanced runs during which they were either instructed to "crave" or to "resist" craving. The neural response to food cues was assessed within and between runs for all participants, and further analyzed between more successful (n = 24) and less successful (n = 7) groups. More successful was defined by meeting 50% excess weight loss. RESULTS: Overall, instructions to "crave" elicited significant activity in the dorsomedial prefrontal cortex (PFC) whereas "resist" elicited significant activity in the dorsolateral PFC (DLPFC). Between groups there was no brain difference when instructed to "crave." The more successful participants however had significantly more activity in the DLPFC when instructed to "resist." CONCLUSIONS: These findings suggest that the ability to mobilize neural circuits involved in executive control post-gastric-bypass surgery may be a unique component of successful outcome post-surgery.

Socially desirable responding by bariatric surgery candidates during psychological assessment.

BACKGROUND: Most bariatric surgery programs in the United States require preoperative psychological evaluations for candidates for surgery. Among those who perform these evaluations is concern that many patients engage in "impression management" or minimizing the symptoms of distress to receive a recommendation to proceed with surgery from the mental health professional. We sought to assess the prevalence of socially desirable responding and its associations with measures of psychological functioning among bariatric surgery candidates at 2 academic medical centers in the United States. METHODS: The participants were male (n = 66) and female (n = 293) bariatric surgery candidates who presented for psychological evaluation. The participants completed 2 measures of socially desirable response styles (Marlowe-Crowne Social Desirability Scale and Personality Assessment Inventory Positive Impression Management scale) and standardized measures of anxiety, depression, and alcohol-related problems. RESULTS: The participants exhibited elevated scores on the social desirability indicators, with 33.3-39.8% scoring above the recommended cut-score on the Personality Assessment Inventory Positive Impression Management scale and 62.3-67% scoring 1 standard deviation above the standardization mean on the Marlowe-Crowne Social Desirability Scale. Scores on the Marlowe-Crowne Social Desirability Scale and Personality Assessment Inventory Positive Impression Management scale correlated inversely with the clinical measures of anxiety and depression, and the high/low scorers on the social desirability indices exhibited significant differences in anxiety and depression. Thus, elevated scores on the social desirability indices were associated with underreporting of certain clinical symptoms. CONCLUSION: A substantial proportion of bariatric surgery candidates appear to present themselves in an overly favorable light during the psychological evaluation. This response style is associated with less reporting of psychological problems and might interfere with the accurate assessment of patient functioning.

The Food Craving Questionnaire-Trait in a bariatric surgery seeking population and ability to predict post-surgery weight loss at six months.

Food cravings have been understudied in bariatric surgery patients and the Food Craving Questionnaire-Trait has not been validated in this population. Reliability and validity of the FCQ-T were examined and a regression analysis was run to determine whether or not preoperative scores on individual subscales of the instrument could predict weight loss at 6 months. The FCQ-T demonstrated excellent internal consistency in bariatric surgery-seeking patients, and individual subscales measuring emotion and mood were correlated with other measures of depression and anxiety. Endorsement of binge eating or emotional eating behaviors during a clinical interview was correlated with similar subscales on the FCQ-T. Higher scores on the subscale 'cues that may trigger food cravings' were associated with greater weight loss at 6 months post-surgery and higher scores on the subscale 'guilt from cravings and/or giving into them' was associated with less weight loss. Management of external cues may predict successful outcomes while emotional impact of cravings may indicate the need for further intervention to help manage specific food craving traits.

Preoperative hemoglobin A1c and postoperative glucose control in outcomes after gastric bypass for obesity.

BACKGROUND: Hemoglobin A1c (HbA1c) is a reliable marker for long-term glycemic control in obese diabetic patients. Roux-en-Y gastric bypass improves HbA1c levels over time. However, it is not clear whether the preoperative HbA1c level is a predictor of the outcome in these patients. Our objectives were to understand the predictive capacity of the preoperative HbA1c level in gastric bypass patients at a single university-based Bariatric Center of Excellence. METHODS: We performed a retrospective review of 468 charts from 2006 to 2009 of patients who had undergone Roux-en-Y gastric bypass. Using their preoperative HbA1c status, the patients were categorized and the postoperative outcomes compared. RESULTS: Of the 468 patients reviewed, 310 (66.2%) had a HbA1c of <6.5% (group 1), 92 (19.4%) had a HbA1c of 6.5-7.9% (group 2), and 66 (14.1%) had a HbA1c level of >8.0% (group 3). No difference was found among the 3 groups in baseline body mass index, race, procedure type, length of stay, hospital cost, and smoking status. Groups 2 and 3 were associated with older age, male gender, and higher baseline creatinine. Groups 2 and 3 also had a proportionally greater inpatient postoperative blood glucose level. An elevated postoperative glucose level was independently associated with wound infection (P = .008) and acute renal failure (P = .04). Also, group 3 experienced worse outcomes, including less weight loss at 18 months and fewer diabetic remissions. Over time, however, the vast majority in all groups achieved excellent chronic glycemic control, with HbA1c <6.5% after Roux-en-Y gastric bypass. CONCLUSION: Poor preoperative glycemic control is associated with worse glucose level control postoperatively, fewer diabetic remissions, and less weight loss. An elevated mean postoperative glucose level is independently associated with increased morbidity.

Patients who endorse more binge eating triggers before gastric bypass lose less weight at 6 months.

BACKGROUND: Patients presenting for gastric bypass surgery often demonstrate binge eating behaviors. The present study sought to determine whether binge eating triggers are predictive of weight loss outcomes in bariatric surgery patients at 6 months postoperatively in the setting of a university hospital. METHODS: A total of 48 patients presenting for gastric bypass surgery at an academic medical center in the Southeastern United States and who had returned for the 6-month follow-up visit were included in the present study. The patients were mostly women (85%), white (71%), and middle-aged (mean age 47 years), with an average weight of 100.9 kg. The patients completed the Inventory of Binge Eating Situations at baseline, and weight loss outcomes were assessed at 6 months. Weight loss success was indexed using 2 methods: the percentage of excess weight lost (continuous variable) and whether the patient was on track with their weight loss as defined by a ≥ 50% excess weight loss (dichotomous variable). RESULTS: A significant negative correlation (r = -.31, P = .03) was found between the preoperative Inventory of Binge Eating Situations scores and the percentage of excess weight loss at 6 months after gastric bypass surgery. Logistic regression analysis showed that "on track" status at 6 months was predicted by the Inventory of Binge Eating Situations score at baseline (Wald chi-square = 3.97, df = 1, P = .046). CONCLUSION: Careful assessment of binge eating situations could serve as a potential predictor of poor weight loss outcomes in patients seeking gastric bypass surgery. These findings support the baseline assessment of binge eating triggers and future research to examine the effectiveness of interventions for coping with binge eating triggers for gastric bypass surgery patients.

Gastrointestinal bleeding after gastric bypass surgery: nuisance or catastrophe?

BACKGROUND: Bleeding in the early postoperative period after gastric bypass surgery is a rare complication reported in 1-4% of cases. In most cases, the bleeding will be self-limited, resolving after discontinuation of anticoagulation for deep venous thrombosis prophylaxis. Occasionally, however, bleeding can be problematic, resulting in significant postoperative morbidity or even mortality. The purpose of the present study was to review cases of early postoperative gastrointestinal bleeding after gastric bypass to identify possible risk factors and examine outcomes. METHODS: A prospective weight loss surgery patient database was reviewed to examine the cases occurring from January 2005 to July 2008. Patients with early postoperative bleeding were identified. The demographics and co-morbidities were compared between those with and without postoperative bleeding. The outcomes of interest included the transfusion requirement, postoperative morbidity, need for reoperation, length of hospital stay, intensive care unit admission, and 30-day mortality. RESULTS: A total of 776 patients had undergone gastric bypass during the study period. Of these, 26 (3.3%) were identified with early postoperative bleeding. No significant differences were found in the demographics between the patients with early postoperative bleed and those without. Of the patients with early postoperative bleeding, the most common presenting sign or symptom was tachycardia (46%) followed by melena (32%) and hematemesis (18%). The transfusion requirement range was 0-11 U (average 3.2 U/patient). Of the 26 patients, 4 (15%) required intensive care unit admission and 8 (31%) required reoperation. Four gastrointestinal leaks, one stomal ulceration, and one gastrogastric fistula were identified in the early postoperative bleeding group. One patient (3.8%) in the early postoperative bleeding group died. Compared with nonbleeding patients, those with early postoperative bleeding had a significantly longer hospital stay (4.8 versus 3, P <.0001) and greater mortality rate (7.1% versus .9%, P <.01). CONCLUSION: Postoperative bleeding is a significant complication after gastric bypass surgery. Although in many cases, the bleeding will be self-limited, it can result in additional notable complications. In the present study, postoperative bleeding contributed to significantly longer hospital stays and resulted in 1 death. A patient with bleeding after gastric bypass should raise due concern, with attention to the potential attendant complications.

Surgical treatment of obesity.

Once an obese patient has failed attempts at diet modification, physical activity, pharmacologic treatment, and possibly even complementary and alternative therapies, the next step is to consider surgical management. Treatment plans must be customized for individual patients and should involve evaluation by the primary care provider, a dietician, psychologist, and surgeon. Then depending on the individual's needs, comorbidities, and candidacy, a specific surgical intervention may be necessary. These procedures are restrictive, malabsorptive, and a combination of both. Each procedure has its own short-term and long-term complications and must be monitored for the rest of the individual's life.

Sphincter of Oddi dysfunction after Roux-en-Y gastric bypass.

BACKGROUND: Patients who have undergone Roux-en-Y gastric bypass for morbid obesity may develop postoperative abdominal pain disorders that require surgical evaluation. Chronic pancreatitis and pain associated with sphincter of Oddi dysfunction (SOD) is an uncommon disorder whose clinical diagnosis is problematic without sphincter of Oddi manometry. To evaluate the diagnosis and treatment of SOD in the gastric bypass population, a retrospective review and analysis of gastric bypass patients who had undergone transduodenal sphincteroplasty (TS) for SOD was undertaken. METHODS: The medical records of patients who had undergone TS after gastric bypass at the Medical University of South Carolina Digestive Disease Center from January 2002 to December 2006 were evaluated for outcomes-based data with the approval of the institutional review board for the evaluation of human subjects. Long-term patient outcomes were assessed using the Medical Outcomes Study Short Form 36-item, version 2, quality-of-life survey. RESULTS: A total of 16 women (median age 49 years) were identified who had undergone TS with biliary sphincteroplasty and pancreatic ductal septoplasty for SOD. The indications for surgery included pain (100%), nausea (31%), weight loss (13%), and recurrent pancreatitis (31%). The diagnosis of SOD was supported by magnetic resonance cholangiopancreatography with secretin stimulation. Three postoperative complications (18.8%) developed, but no mortality. The average length of hospital stay was 5 days (range 2-9). Of the 16 patients, 13 (81%) responded to the survey follow-up. The mean length of follow-up was 28 months (range 16-57). Of the 13 patients, 11 (85%) reported pain improvement after surgery. The survey's norm-based scores were similar to those of a representative population. CONCLUSION: SOD should be considered in the differential diagnosis of gastric bypass patients with pancreatobiliary pain after cholecystectomy. When the clinical history is supported by laboratory and magnetic resonance cholangiopancreatography data, TS can be undertaken with low morbidity and good patient outcomes. SOD is a notable disorder in the gastric bypass population. With appropriate patient selection, TS can be beneficial.

The practice of pancreatic resection after Roux-en-Y gastric bypass.

The morbid obesity epidemic in the United States has resulted in increasing numbers of patients who have undergone Roux-en-Y gastric bypass who require surgical management of nonbariatric disorders. When pancreatic resection is indicated in bariatric patients, consideration of the altered foregut anatomy can be applied to the principles of pancreatic resection to foster effective techniques that minimize operative complications. A retrospective review and analysis of bariatric patients who underwent pancreatic resection at the Medical University of South Carolina Digestive Center over a 2-year period (2006 to 2007) was conducted to assess indications for operation, operative techniques, and postoperative outcome in patients with previous Roux-en-Y gastric bypass. There were five patients (four female, one male) identified with a mean age of 35 years (range, 32-50 years). The mean time interval from gastric bypass to pancreatic resection was 42.6 months (range, 10-72 months). Indications for pancreatic operations were islet hyperplasia in two patients, chronic pancreatitis in two, and serous cystadenoma in one. Two patients underwent duodenal-preserving pancreatic head resection (Beger procedure) and three underwent distal pancreatectomy and splenectomy. Mean length of hospital stay was 11.4 days (range, 5-22 days). Two patients had extended hospital stay as a result of gastrointestinal ileus. There was no other operative morbidity or mortality. Mean length of patient follow up was 9.8 months (range, 1-17 months). Specific operative techniques used in pancreatic head resection were duodenal preservation, pancreatic drainage with an omega loop constructed from a mid-Roux limb, and excluded stomach gastrostomy. Techniques used in pancreatic tail and body resection were splenectomy discontinuous from pancreatectomy, division of the splenic vein and artery at the pancreatic neck early in surgery, retrograde dissection of the pancreas body and tail, and dissection of the body and tail posterior to the Roux limb leaving the Roux limb intact. Pancreatic resection after Roux-en-Y gastric bypass is safe and effective when using prescribed operative principles that minimize disruption of the foregut reconstruction and adds protection to the gastric remnant with a gastrostomy for decompression and access for enteral alimentation when necessary.

Validation of the obesity surgery mortality risk score in a multicenter study proves it stratifies mortality risk in patients undergoing gastric bypass for morbid obesity.

BACKGROUND: A scoring system for clinical assessment of mortality risk has been previously proposed for bariatric surgery (Demaria EJ, Portenier D, Wolfe L, Surg Obes Relat Dis. 2007;3:34-40.). The Obesity Surgery Mortality Risk Score (OS-MRS) was developed from a single institution experience of 2075 patients. The current study provides multicenter validation of the value of the OS-MRS. The OS-MRS assigns 1 point to each of 5 preoperative variables, including body mass index>or=50 kg/m2, male gender, hypertension, known risk factors for pulmonary embolism (previous thromboembolism, preoperative vena cava filter, hypoventilation, pulmonary hypertension), and age>or=45 years. Patients with total score of 0 to 1 are classified as 'A' (lowest) risk group, score 2 to 3 as 'B' (intermediate) risk group, and score 4 to 5 as 'C' (high) risk group. METHODS: Prospectively-collected data from 4431 consecutive patients undergoing a primary gastric bypass at 4 bariatric programs recruited to validate the proposed system were analyzed to assess OS-MRS as a means of stratifying surgical mortality risk. RESULTS: There were 33 total deaths for an overall mortality for the validation cohort of 0.7% consistent with published standards. Mortality for 2164 class A patients was 0.2%, for 2142 class B patients was 1.1%, and for 125 class C patients was 2.4%. Mortality was significantly different between each of the class A, B, and C groupings (P<0.05, chi2). Mortality was 5-fold greater in the class B group than in class A. Only 6 patients with all 5 risk factors were identified. Class C patients (n=125, 3% of total cohort) were characterized by a 12-fold greater mortality than the lowest risk group (A) and a disproportionate 9% of all mortalities. CONCLUSION: The OS-MRS was found to stratify mortality risk in 4431 patients from 4 validation centers that were nonparticipants in the original defining cohort study. The score represents the first validated scoring system for risk stratification in bariatric surgery and is anticipated to aid informed consent discussions, guide surgical decision-making, and allow standardization of outcome comparisons between treatment centers.

Predictors of endoscopic findings after Roux-en-Y gastric bypass.

OBJECTIVES: To evaluate predictors of endoscopic findings in symptomatic patients after Roux-en-Y gastric bypass (RYGBP) for obesity. METHODS: A retrospective chart review of 1,001 RYGBP procedures was performed. Two hundred twenty-six (23%) patients were identified as having endoscopy to evaluate upper gastrointestinal symptoms following surgery. Polychotomous logistic regression analysis was used to assess predictors of normal endoscopy, marginal ulcers, stomal stenosis, and staple-line dehiscence. RESULTS: The most common endoscopic findings were 99 (44%) normal postsurgical anatomy, 81 (36%) marginal ulcer, 29 (13%) stomal stenosis, and 8 (4%) staple-line dehiscence. Factors that significantly increase the risk of marginal ulcers following surgery include smoking (AOR = 30.6, 95% CI 6.4-146) and NSAID use (AOR = 11.5, 95% CI 4.8-28). PPI therapy following surgery was protective against marginal ulcers (AOR = 0.33, 95% CI 0.11-0.97). Median time for diagnosis of marginal ulcers following surgery was 2 months, and 77 of 81 (95%) presented within 12 months. CONCLUSIONS: Following RYGBP surgery for obesity, smoking and NSAID use significantly increase the risk of marginal ulceration, and PPI therapy is protective. Because a significant majority of marginal ulcers present within 12 months of surgery, it may be reasonable to consider prophylactic PPI therapy during this time period, especially for high risk patients.

Postoperative left prefrontal repetitive transcranial magnetic stimulation reduces patient-controlled analgesia use.

BACKGROUND: Several recent studies suggest that repetitive transcranial magnetic stimulation can temporarily reduce pain perception in neuropathic pain patients and in healthy adults using laboratory pain models. No studies have investigated the effects of prefrontal cortex stimulation using transcranial magnetic stimulation on postoperative pain. METHODS: Twenty gastric bypass surgery patients were randomly assigned to receive 20 min of either active or sham left prefrontal repetitive transcranial magnetic stimulation immediately after surgery. Patient-controlled analgesia pump use was tracked, and patients also rated pain and mood twice per day using visual analog scales. RESULTS: Groups were similar at baseline in terms of body mass index, age, mood ratings, pain ratings, surgery duration, time under anesthesia, and surgical anesthesia methods. Significant effects were observed for surgery type (open vs. laparoscopic) and condition (active vs. sham transcranial magnetic stimulation) on the cumulative amount of patient-delivered morphine during the 44 h after surgery. Active prefrontal repetitive transcranial magnetic stimulation was associated with a 40% reduction in total morphine use compared with sham during the 44 h after surgery. The effect seemed to be most prominent during the first 24 h after cortical stimulation delivery. No effects were observed for repetitive transcranial magnetic stimulation on mood ratings. CONCLUSIONS: A single session of postoperative prefrontal repetitive transcranial magnetic stimulation was associated with a reduction in patient-controlled analgesia pump use in gastric bypass surgery patients. This is important because the risks associated with postoperative morphine use are high, especially among obese patients who frequently have obstructive sleep apnea, right ventricular dysfunction, and pulmonary hypertension. These preliminary findings suggest a potential new noninvasive method for managing postoperative morphine use.

Placement of inferior vena cava filters in bariatric surgical patients--possible indications and technical considerations.

BACKGROUND: Surgical treatment of the morbidly obese has assumed an increasingly important role in both the academic and community setting, while postoperative pulmonary embolism remains a devastating complication. Since the overall incidence remains low, the role for vena cava filter placement in this group is not yet well defined. In addition, the technical challenges and techniques for insertion have not been well-described. We present our experience with filter placement among patients with gastric bypass and the evolution of technique to facilitate safe placement in this group. METHODS: From 1995 to August 2003, 586 patients underwent gastric bypass for morbid obesity. Review of registries and records from this period was accomplished to identify patients at MUSC who underwent both the gastric bypass and placement of an inferior vena cava filter. 12 patients were identified by this method. RESULTS: Technical challenges with venous access and imaging are described. 6 patients were identified as potential high risk for thromboembolic complications and had a filter placed preoperatively with a mean postoperative stay of 5.3 days. The 6 patients who required filter placement in the postoperative period as part of the management of postoperative complications had a mean hospital stay of 24.5 days. There were no long-term complications associated with filter placement at a mean follow-up interval of 19 months. CONCLUSION: Inferior vena cava filter placement is not only feasible and safe for the morbidly obese individual undergoing gastric bypass, but should be strongly considered for patients with risk factors for thromboembolic complications or who experience postoperative complications requiring ICU stay or prolonged immobility.

Gender, race, and obesity-related quality of life at extreme levels of obesity.

OBJECTIVE: Research investigating obesity-related quality of life (QOL) has shown that at increasing levels of overweight, individuals report more impaired QOL. Further, some research has indicated that white women suffer more impairment than men and African Americans. The current study sought to expand the existing literature by investigating an extreme subsample of the obese population. It was expected that participants in the current study would report more impaired obesity-related QOL than in previous research conducted with less obese individuals. It was also hypothesized that race and gender groups would differ in obesity-related QOL and that the relationship between degree of overweight and QOL would not be consistent across race and gender groups. RESEARCH METHODS AND PROCEDURES: Impact of Weight on Quality of Life Questionnaire-Lite Version data were collected from 512 individuals seeking gastric bypass surgery (mean BMI = 53.3) RESULTS: Results confirmed the study hypotheses. In general, white women reported the most QOL impairment, despite having significantly lower BMI than other race/gender groups. Compared with previous studies, the observed relationships between BMI and QOL were somewhat attenuated. DISCUSSION: Various domains of QOL may be differentially affected by degree of obesity; these relationships are not homogeneous throughout the obese population.

Percutaneous endoscopic gastrostomy complications in a tertiary-care center.

Since its introduction in 1980 the percutaneous endoscopic gastrostomy (PEG) has become the procedure of choice for establishing enteral access. However, there is still a relatively high complication rate associated with PEG placement. We reviewed the complications associated with PEG placement at our tertiary-care referral center. A retrospective chart review was conducted on patients over 17 years of age undergoing PEG placement between January 1, 1994 and March 1, 1996. Indications for surgery, antibiotic use, and postoperative complications were determined. There were 166 PEGs placed during this time and 27 (16.3%) complications. There was one death (0.6%) directly related to PEG placement. Thirteen patients (7.8%) died within 30 days of PEG placement and an additional 12 patients (7.2%) died before leaving the hospital. Wound infections occurred in nine (5.4%) patients including one case of necrotizing fasciitis. Only four of 153 (2.6%) patients who received preoperative antibiotics developed wound infections, whereas five of 13 (38.5%) patients without antibiotic prophylaxis developed infections. We conclude that percutaneous endoscopic gastrostomy is a safe and effective way of establishing enteral access in most patients. A relatively high mortality rate can be expected as a result of underlying medical problems. Antibiotics should be given to help prevent local wound infections.