A method for radiologic-assisted gastrostomy when percutaneous endoscopic gastrostomy is contraindicated.

BACKGROUND: Percutaneous endoscopic gastrostomy (PEG) has been established as a faster and safer procedure than open surgical gastrostomy. It cannot be done, however, for many patients with partially obstructing pharyngeal or esophageal carcinoma, previous gastrectomy, upper abdominal surgery, or bowel distension from distal obstruction. PATIENTS AND METHODS: We attempted percutaneous radiologic-assisted gastrostomy (RAG) in 231 patients referred for gastrostomy, 38 of whom had a relative contraindication for PEG. The procedure involves passing, under radiologic guidance, an orogastric inflation tube that contains a snare. We used a 5-inch long, 18-gauge needle to transabdominally insert a wire into the stomach, avoiding loops of bowel visualized by air contrast. Retrieving the transabdominal wire by snare allowed retrograde passage of the gastrostomy tube as done in standard PEG. RESULTS: The procedure was successful in 230 of 231 cases, including 37 of the 38 patients with contraindications. We could not gain gastric access in 1 patient with a 75% gastrectomy. Overall, 6 patients developed complications and 1 died. There was no procedure-related morbidity or mortality in the patients with contraindications to PEG who underwent successful RAG. Subsequent laparotomy indicated tube passage through the liver in 2 of these cases and small bowel mesentery in 1 case without clinical problems. We performed a percutaneous jejunostomy in the efferent limb of the gastrojejunostomy in 1 patient with a previous gastrectomy. CONCLUSION: The snare technique is simpler and faster than the usual radiologic gastropexy technique, and safer than an endoscopic procedure. It has become our procedure of choice for gaining gastric access.

Effect of the mode of lipid administration on parenteral nutrition-related infections.

OBJECTIVE: To determine if total nutrient admixtures (TNAs) influence the rate of infection in clinical practice. DESIGN: Prospective, randomized trial. SETTING: Department of Veterans Affairs Medical Center. PATIENTS: All patients were administered parenteral nutrition (PN) via a central venous catheter and received daily lipids. INTERVENTION: Patients were randomized as to the mode of administration of lipids. Lipids were either administered with other PN components in a TNA or were piggybacked (PB) into the PN solution. MAIN OUTCOME MEASURES: Treatment groups were compared for the rate of occurrence of PN-related infections. Infections were classified as catheter infections or catheter sepsis. RESULTS: Ninety-eight patients were entered into the trial. Data from 96 patients (44 TNA, 52 PB) were available for analysis. Treatment groups were well matched for age, baseline albumin, days of PN, predicted basal metabolic rate, and calorie and protein requirements. TNA patients received a significantly greater percentage of nonprotein calories as lipid. The incidence of infection was 12.6 and 10.3 per 1000 days of PN in the TNA and PB groups, respectively (p = 0.89). The microorganisms responsible for infection and the type of infections that developed were similar in both groups. CONCLUSIONS: Use of TNAs does not influence the rate of infection in patients receiving PN.

Systemic appearance of nutrients placed into the peritoneal cavity.

The use of the peritoneal cavity as a route for delivery of nutrients in patients on continuous ambulatory peritoneal dialysis as well as patients not in renal failure has been proposed. In earlier studies we found dogs could be maintained for 30 days through continuous peritoneal infusion of a solution of amino acids, glucose, and lipids. We then conducted experiments to observe the rate of absorption of nutrients from the gut compared to the peritoneal cavity, the relative rate absorption of amino acids, sugars, and fats, and the absorption of of glucose polymers from the peritoneal cavity. These latter experiments are the subject of this report. In a dog model we observed D-xylose was more rapidly absorbed through the peritoneal cavity than through the gut, but amino acids were absorbed at the same rate. Glucose and amino acid are absorbed at the same rate through the peritoneal cavity, but lipids do not appear in the system circulation for 2 1/2 hr. Intraperitoneal glucose polymer produces an increase in serum glucose concentration that is lower and more sustained than what occurs after giving an equal amount of plain glucose. It is concluded that the peritoneal cavity rapidly transports glucose, amino acids, and glucose polymers and slowly transports lipids into the systemic circulation. Consideration of this route for delivery of nutritional support seems reasonable, but further studies of appropriate concentrations and rates of delivery will be required.

Percutaneous needle pharyngostomy.

Pharyngostomy is a technique of proven usefulness in patients requiring extra oral enteric access. Techniques to perform this have required an operating room, general anesthesia, and the associated risks. We have developed a technique for needle pharyngostomy that can be performed at the bedside of an awake patient. We utilize a central venous catheter tray and a small bore feeding tube to do this procedure. We have performed this on 17 patients without a procedural complication. This form of enteric access has allowed total nutritional support for up to 2 yr. Two late infections at the pharyngostomy site have occurred; both were easily controlled with oral antibiotics. This technique is recommended for patients who repeatedly pull out feeding tubes or are planning home enteral nutrition.

Total nutritional support through the peritoneal cavity.

We prepared ten dogs with intraperitoneal catheters to allow total nutritional support for 30 days by constant infusion of a solution containing 1.5% amino acid, 10% glucose, and 1% lipids at a rate of 4 ml/kg/hr. Seven dogs survived in apparent good health but with a 13.9 +/- 1.3% weight loss. There was no correlation between actual caloric input and percentage of weight loss. Serum albumin concentration fell to 2.6 +/- 0.3 gm/dl but other chemistries remained near normal. The peritoneum showed significant inflammatory reaction but this resolved by 30 days. Three dogs died in the first week from what appeared to be circulatory collapse from rapid fluid shifts. We conclude that a significant amount of nutrients can be delivered through the peritoneal cavity but whether total nutritional support is feasible remains to be proven.

Spontaneous resolution of intestinal obstruction while receiving home parenteral nutrition.

Six patients with intestinal obstruction secondary to benign adhesions so dense and vascular that the operating surgeon could not free them, were placed on a home total parenteral nutrition program. Four patients had enterocutaneous fistulas, and two had their bowels divided and stomas created to divert intestinal contents from distal enterotomies made during the attempt to free the intestine. The enterocutaneous fistulas closed in four patients within 2 weeks to 5 months and the obstructions spontaneously resolved in 2 to 3.5 months. At reoperation on the two patients whose intestinal tracts were purposely divided, the previously matted bowels with dense adhesions were easily freed. We believe complete gastrointestinal rest allows adhesions to mature into long avascular collagen fibers in the absence of a persistent inflammatory reaction that accompanies partial or total small bowel obstruction. We recommend that patients receive 3 months of home total parenteral nutrition before a second operation for persistent obstruction after recent enterolysis should be considered. Spontaneous resolution should occur, but if not, reoperation can be performed safely after the 3 month interval.

Creating a continent tubeless gastrostomy.

Standard tube feeding gastrostomy is associated with a myriad of problems among which leakage of gastric contents is the most serious. Over the years many methods of creating tubeless gastrostomy have been unsuccessful because of the extent of surgery required and the persistent leakage of gastric fluids. By using gastrointestinal staplers to create a gastric tube and by incorporating a reverse intussusception valve at its base, we created a continent tubeless feeding gastrostomy in 23 patients aged 10 to 81 years. Eight had head injuries, 12 had cerebrovascular strokes, and 3 had head and neck disorders leading to aphagopraxia. Twelve of the procedures were performed with local anesthesia and the remainder with general anesthesia. The GIA stapler was used to create a 5 cm gastric tube after which circumferential stitches were placed to intussuscept 2 cm of the tube into the stomach to create a valve. A 1 cm stoma was created at skin level. Operative time was 47 to 90 minutes and there were no operative deaths or complications. Stomal necrosis occurred in two patients so they were converted to a standard tube gastrostomy by leaving a Foley catheter in place. After 16 months of follow-up, one stricture developed at 4 months and two were electively closed under local anesthesia. None is incontinent. The tubeless continent feeding gastrostomy is convenient for both patients and nursing staff and should replace the standard feeding gastrostomy when long-term nutritional support is needed.

A tube for enteral nutrition of patients with aphagopraxia and patients with ventilator assistance.

A new feeding tube was designed for use in patients who cannot swallow. A comparison of our ability to pass a commercially available, mercury weighted, small feeding tube or the new, nonweighted feeding tube was made. Forty-one consecutive patients who had endotracheal intubation and who had mechanical ventilation assistance or who had suffered injuries to the central nervous system, producing aphagopraxia were compared. In the 22 patients in whom the guided tube system was first tried, enteric support was possible in 20. Gastric placement was possible in only 12 of 19 patients in whom the mercury weighted tubes were first tried and in only one of these patients did the tube pass into the small intestine beyond the ligament of Treitz. Seventeen of 20 nonweighted tubes passed into the small intestine. The newly designed small feeding tube system should be used as the initial means of gaining access to the intestine for enteric nutritional support of patients in intensive care units and after strokes or neurologic injuries when the patient cannot swallow.

Early postoperative nutritional support using the serosal tunnel jejunostomy.

A silicone rubber serosal tunnel jejunostomy for postoperative nutritional therapy was placed in 38 patients who had major operations. A dilute, chemically defined diet was begun within 48 hours of surgery and an attempt made to advance it slowly to full strength by the fifth day. No catheter-related complications occurred. Intravenous fluid therapy was shortened to less than three days in 11 (29%) patients. Only 17 (45%) patients tolerated full-strength feedings within the protocol time. Three (8%) patients depended on their tube feedings for over 30 days and the need for parenteral nutrition was avoided. Of six septic patients, four had complete intolerance of the diet; two of these developed massive gastric hypersecretion. Serum albumin was a statistically significant indicator of whether a patient could tolerate tube feedings. No patient with an albumin less than 3 g/dl tolerated full-strength feedings and the pded. Of six septic patients, four had complete intolerance of the diet; two of these developed massive gastric hypersecretion. Serum albumin was a statistically significant indicator of whether a patient could tolerate tube feedings. No patient with an albumin less than 3 g/dl tolerated full-strength feedings and the patients with albumin greater than 4 g/dl had no problems. Postoperative feeding utilizing the serosal tunnel jejunostomy is technically safe, but feeding difficulties can be anticipated in those patients who are septic or severely malnourished; these are the patients whose nutritional needs are the greatest. The greatest benefits accrue to those patients who are in need of long-term nutritional support. We recommend routine placement of these catheters in major operations.

Protein-calorie malnutrition in a community hospital.

An assessment of protein-calorie status was performed on 200 consecutive adult nonobstetric admissions to a private hospital from two group family practices. Anthropometric measurements, serum albumin level, and total lymphocyte count were determined at admission and weekly if the patient remained in the hospital. Nonnutritional factors affecting muscle protein stores and serum albumin level were taken into account. A total of 63 patients (31.5%) were found to be malnourished. The most common conditions associated with decreased protein stores were hypoxic cardiopulmonary disease, gastrointestinal disease, neuromuscular or arthritic impairment, organic brain syndrome, and febrile illness. Although nonnutritional factors accounted for many instances of protein depletion, expecially milder degrees of depletion, nevertheless protein calorie malnutrition was highly prevalent in this population.

Cholecystosonography for the decision to operate when acute cholecystitis is suspected.

Forty-four patients admitted with a diagnosis of acute cholecystitis underwent immediate cholecystosonography, and if the findings were interpreted as showing gallstones, surgery was performed. The accuracy was 91 per cent. The routine use of cholecystosonography in acutely ill patients is recommended.

Management of enterocutaneous fistulas and problem stomas with silicone casting of the abdominal wall defect.

Silicone casting of abdominal wall defects around enteric fistulas in six patients and problem stomas in three patients proved to be an effective means of controlling the output of the fistulas, reducing wound care time, and reducing or eliminating parenteral nutrition needs. Outpatient management was possible in seven of the nine patients. It is observed that the wounds healed rapidly with this method of fistula control. Epithelialization occurred more rapidly than expected. This method of management may tend to make the fistulas remain open longer than by other means of care, but the significant increase in patient comfort, the financial savings, and the relative safety warrant continued utilization and observation of this method of management.

The management of infected abdominal aortic grafts and graft-enteric fistulas.

Seven cases of infected aortic grafts or aorto-enteric fistulas following resection of an abdominal aortic aneurysm are reviewed. All cases were treated with axillo-femoral bypass and graft removal. Patients had recurrent fever, chills, and abdominal pain (5 patients), or massive gastrointestinal hemorrhage (2 patients). The two patients with massive gastrointestinal hemorrhage died. Three of the 5 long-term survivors had a recurrence of the retroperitoneal abscess after graft removal; one of these died. One axillo-femoral bypass graft required early thrombectomy. There have been no subsequent problems with any of the grafts in the 2.5 to 4 year followup period. Axillo-femoral bypass immediately before graft removal is the treatment of choice for infected or fistulous aortic grafts. Any delay in graft removal after the onset of symptoms should be avoided.

Repair of massive septic abdominal wall defects with Marlex mesh.

Marlex mesh was used to close the abdominal wall defect in six patients with septic wound dehiscence and intra-abdominal infection. The mesh was implanted under local anesthesia and served as a protective covering for the bowel and allowed early ambulation, including prone positioning of the patient for easier wound care. In four surviving patients, the Marlex mesh was covered by full thickness skin flaps after granulation tissue had covered the material. No patients had infected sinus tract formation or extrusion. Two patients had incisional hernias develop when the Marlex mesh was not sutured to the abdominal wall permanently. The use of Marlex mesh to cover infected defects in the abdominal wall when primary closure cannot be accomplished is suggested by our experience.

Diagnostic peritoneal lavage in evaluating acute abdominal pain.

A study was performed to determine the value of peritoneal lavage in the acute abdomen not related to trauma. Lavage was performed in 33 patients in the evaluation of abdominal pain of sufficient degree to warrant consideration for surgical intervention. Peritoneal lavage was truly positive or truly negative in 64% of the cases. It showed false negative results in 28% and false positive results in 8%. The lavage was most accurate in the evaluation of appendicitis, colonic disease, and intra abdominal bleeding. It was highly inaccurate in the evaluation of cholecystitis and peptic ulcer disease. It was concluded that the peritoneal lavage can be a useful adjunct in the evaluation of patients with abdominal pain and should be considered in difficult diagnostic problems but not routinely employed.

Results of routine preoperative sigmoidoscopy and barium enema on patients with inguinal hernia.

The diagnostic findings of history, physical examination, rectal examination, barium enema, and sigmoidoscopy in 684 patients with the initial complaint of inguinal hernia were examined. History and physical examination led to the diagnosis of four cancers of the colon and one villous adenoma. No significant new findings were made by sigmoidoscopy when symptoms elicited by history did not indicate some abnormality. A cost analysis of finding significant disease in these patients with inguinal hernias indicates that in-hospital evaluation by barium enema and sigmoidoscopy is probably not warranted.