Placement of percutaneous endoscopic gastrostomy or jejunostomy is a safe procedure with low periprocedural mortality, but overall mortality rates are high because of underlying disease conditions. These procedures are also associated with postprocedure complications. The clinically significant adverse events related to the procedures include infection (at tube site and peritonitis), bleeding, and aspiration. More rare associated events include buried bumpers, injury to adjacent viscera with subsequent fistula formation, and tumor seeding. There is a lack of guidelines about these procedures other than those concerning the use of antibiotics and the management of antithrombotics and anticoagulation before the procedure.
Key points
- •
There is a high incidence of complications after percutaneous enteral access but most are minor.
- •
There is a high mortality in patients undergoing percutaneous enteral access but this is usually caused by underlying illness.
- •
American Society for Gastrointestinal Endoscopy guidelines exist regarding the use of antibiotics and management of antithrombotics and anticoagulation before percutaneous endoscopic gastrostomies but data about other complications are limited.
- •
Careful patient selection is imperative and should involve detailed evaluation of indications, contraindications, risks and benefits, possible technical difficulties, and ethical issues involved with long-term enteral access.
Minimizing the adverse events associated with percutaneous enteral access
Percutaneous endoscopic enteral access is required in patients who need long-term enteral nutrition or decompression of the gastrointestinal tract for distal obstruction. It can be achieved by various modalities and includes the most commonly used percutaneous endoscopic gastrostomy (PEG) along with placement of direct jejunostomy (DPEJ) or placement of a jejunal extension through the PEG (PEG-J). The technique of PEG placement without the need for a laparotomy was first described in 1980 by Gauderer and Ponsky. Since then there have been numerous modifications and although the procedure-related mortality is low (less than 1%), several studies have reported short-term mortality in these patients ranging from 10% to 25%. A recent study showed decreased 30-day and 1-year mortality rates in patients who received PEG compared with those who were offered but deferred PEG placement. In a recent multicenter study from South Korea involving 1625 patients undergoing PEG placement with pull technique, the overall complication rate was reported to be 13% with most of these being minor. Earlier studies done in the United States have shown complication rates ranging from 13% to 32%. Even though the procedure-related mortality and morbidity are low, overall mortality is high in these patients (approaching 24% at 1-year in one study) and is usually related to underlying medical conditions. These findings were confirmed by a recent study from Europe in which it was shown that 24-week mortality in patients undergoing PEG placement depended on the indication and was significantly higher in patients who underwent PEG placement for neurologic reasons compared with those with tumor-related indications (60% vs 27.7%, respectively).
Minimizing the adverse events associated with percutaneous enteral access
Percutaneous endoscopic enteral access is required in patients who need long-term enteral nutrition or decompression of the gastrointestinal tract for distal obstruction. It can be achieved by various modalities and includes the most commonly used percutaneous endoscopic gastrostomy (PEG) along with placement of direct jejunostomy (DPEJ) or placement of a jejunal extension through the PEG (PEG-J). The technique of PEG placement without the need for a laparotomy was first described in 1980 by Gauderer and Ponsky. Since then there have been numerous modifications and although the procedure-related mortality is low (less than 1%), several studies have reported short-term mortality in these patients ranging from 10% to 25%. A recent study showed decreased 30-day and 1-year mortality rates in patients who received PEG compared with those who were offered but deferred PEG placement. In a recent multicenter study from South Korea involving 1625 patients undergoing PEG placement with pull technique, the overall complication rate was reported to be 13% with most of these being minor. Earlier studies done in the United States have shown complication rates ranging from 13% to 32%. Even though the procedure-related mortality and morbidity are low, overall mortality is high in these patients (approaching 24% at 1-year in one study) and is usually related to underlying medical conditions. These findings were confirmed by a recent study from Europe in which it was shown that 24-week mortality in patients undergoing PEG placement depended on the indication and was significantly higher in patients who underwent PEG placement for neurologic reasons compared with those with tumor-related indications (60% vs 27.7%, respectively).
Adverse events associated with upper endoscopy
Most of the data about complications related to upper endoscopy are obtained from relatively healthy patients who undergo the procedure in an outpatient setting, whereas PEG placement is commonly performed in relatively sicker patients with much higher comorbidities and lower life expectancy. The adverse events associated with upper endoscopy and their management strategies are discussed elsewhere in this issue and are not discussed here.
Pneumoperitoneum and peritonitis
Pneumoperitoneum is reportedly common after PEG placement and is usually a temporary and benign condition. Previous studies have shown an incidence of around 40% for endoscopic and around 55% for radiologically placed gastrostomies. In a prospective study involving 65 patients undergoing PEG placement, pneumoperitoneum was noted on radiographs in 20% of patients at 3 hours. Most of these resolved but pneumoperitoneum persisted in three patients beyond 72 hours without any clinical significance in any of these patients. In a retrospective study, Blum and colleagues showed that free intra-abdominal air was noted in 12% of patients on radiographic imaging within 5 days of PEG or PEG-J placement. A total of 85% of these patients did not have any complications, whereas 15% developed signs and symptoms concerning for peritonitis and required laparotomy. Hence clinical condition postprocedure should dictate management rather than the finding of free air that is of no clinical significance in most cases.
In a retrospective review of 322 surgical intensive care unit patients undergoing PEG placement, 5% developed peritonitis requiring laparotomy. Body mass index higher than 30 kg/m 2 and serum albumin lower than 2.5 gm/dL were associated with increased risk of peritonitis. Use of carbon dioxide for insufflation during PEG placement has been shown to be associated with lower incidence of pneumoperitoneum and lesser distention of small bowel.
Aspiration
Aspiration is one of the most common major complications related to the placement of PEG tubes and can occur during or after the procedure. The rate of aspiration events occurring periprocedurally has been reported to be around 1%, whereas postprocedure aspiration has been reported in up to 20% to 30% of patients and is associated with high mortality. Most of the periprocedural aspiration events are related to aspiration of oropharyngeal contents, whereas delayed aspiration events also involve gastric contents including tube feeds. Periprocedural aspiration events can be affected by the depth of sedation, frequency of suctioning during procedure, and level of head elevation. All of these should be closely monitored and optimized during all endoscopic procedures. Postprocedural aspiration events have been shown to be related to the presence of reflux esophagitis, hiatal hernia, and impaired swallowing, the last one being an indication itself for PEG placement in many of these patients. Placement of PEG tubes for prevention of aspiration pneumonia in patients with cognitive impairment is not supported by present literature because there are no conclusive data that PEG feedings decrease aspiration rates.
Injury to adjacent viscera
Because placement of a PEG tube relies on transillumination and finding the site of insertion by finger indentation on the abdominal wall, there is a very low risk of injury to the organs adjacent to the anterior abdominal wall including colon, small bowel, liver, and spleen. These adverse events are uncommon and hence exact incidence is not known but there are multiple case reports. Bowel injury in patients undergoing PEG placement is usually seen in very young or elderly patients because of laxity of the mesentery that can lead to bowel transposition between the abdominal wall and anterior wall of the stomach. Excessive insufflation of the stomach and small bowel can also lead to gastric rotation and bowel transposition. Making sure that there is 1:1 indentation before needle insertion can prevent this event. The needle should always be advanced into the gastric lumen under direct endoscopic visualization while pulling back on the syringe as the needle is advanced. Free suction of air is noted at the same time as the needle is visualized in the stomach; this ensures the air being suctioned is not coming from the bowel lumen.
Although some published reports indicate that transhepatic insertion can be managed conservatively, there is a case report describing massive intraperitoneal bleeding caused by liver injury. Colonic injury during PEG insertion can be initially asymptomatic and difficult to diagnose clinically unless there is a high clinical suspicion but can lead to formation of cologastric or colocutaneous fistulas over time (described later).
Bleeding
Placement of endoscopic gastrostomy or jejunostomy tubes can lead to early bleeding from the abdominal wall or delayed bleeding from ulcer formation. In a retrospective review of 1625 patients with PEG placement in South Korea, 1.2% of patients developed bleeding. Use of anticoagulant agents and the presence of diabetes mellitus were associated with an increased risk of bleeding. Two retrospective reviews involving more than 1000 patients each from the United States showed post–PEG placement bleeding rates ranging from 2.8% to 3.3%. Immediate gastric bleeding after PEG placement is usually caused by an injury to the gastroepiploic artery itself or one of its branches. It can be controlled by external pressure at the PEG site. Sometimes tightening of the PEG bumper against the abdominal wall is needed to tamponade the bleeding. However, it is important to loosen the bumper after 48 hours to prevent ischemic injury and ulceration of the gastric mucosa at the PEG site. Severe intraperitoneal hemorrhage can occur because of liver laceration and presents as severe postprocedure hypotension with or without peritonitis. The management is primarily surgical and it has high mortality based on the scant data from case reports.
American Society for Gastrointestinal Endoscopy Standards of Practice Committee 2009 guidelines consider PEG placement a high-risk procedure for gastrointestinal bleeding and hence recommend holding newer antiplatelet medications (theinopyridines) for 7 to 10 days before the procedure in patients who are low risk for cardiovascular complications. In patients who are high risk for cardiovascular complications, PEG placement should be deferred until it is safe to hold anticoagulation and antiplatelet therapy and alternate methods for feeding should be considered. Those patients who are on anticoagulation should have their anticoagulation held and bridging with short-acting anticoagulants should be considered. Nonsteroidal anti-inflammatory drugs and aspirin can be continued before PEG placement.
Common endoscopic findings in patients undergoing endoscopy for upper gastrointestinal bleeding after PEG placement include reflux esophagitis (most common finding in two studies), mucosal tears, and gastric erosions or ulcers. Gastric ulceration from PEG placement can occur on anterior and posterior walls of the stomach. Anterior ulceration is usually caused by pressure necrosis from excessive tension between the bumper and face plate. It is of great importance to evaluate the mucosa below the gastric bumper or balloon after external manipulation during the endoscopy ( Fig. 1 ). Excessive traction should be avoided to prevent these ulcers after PEG placement and it is our practice to loosen the bumper by 1 cm approximately 24 hours after PEG placement to ensure it is freely rotating at the skin surface.
Infections
Infectious complications are common without the routine use of prophylactic antibiotics, with one study reporting an incidence of around 38%. Most of the infections are minor and include local wound infections but serious infections including peritonitis and necrotizing fasciitis can also occur after PEG placement. In a randomized clinical trial, Ahmad and colleagues showed that use of periprocedural antibiotics decreased the risk of peristomal infections in the first week after PEG placement from 18% to 3%. A meta-analysis of 10 randomized controlled trials involving more than 1000 patients showed that penicillin- and cephalosporin-based prophylaxis are effective in decreasing the risk of wound infections after PEG placement. Overall relative risk reduction was 64% with absolute risk reduction of 15% and number needed to treat was 8 to prevent one wound infection. Penicillin-based prophylaxis was slightly superior to cephalosporin-based prophylaxis (absolute risk reduction of 13% vs 10% and number needed to treat 8 vs 10). Current American Society for Gastrointestinal Endoscopy guidelines recommend parental cefazolin (or an antibiotic with equivalent coverage) to be administered in all patients 30 minutes before PEG placement. With the increasing prevalence of methicillin-resistant Staphylococcus aureus (MRSA) colonization in hospitalized patients, MRSA wound infections have been recognized as an important cause of gastrostomy wound infections. Thomas and colleagues have shown that MRSA screening before PEG placement and subsequent decontamination can decrease the incidence of MRSA wound infections but more conclusive data are needed before this can be recommended in all patients undergoing PEG placement. An important factor to prevent the local infectious complications is adherence to sterile technique.
Most of the gastrostomy wound infections can be managed by local wound care but close monitoring to make sure that it does not progress to deep tissue involvement is important. Necrotizing fasciitis is a rare but potentially fatal complication that can occur from infection at the PEG site. Most of the cases of necrotizing fasciitis have been documented as case reports and hence actual incidence is not known. Management of necrotizing fasciitis involves surgical debridement along with broad antibiotic coverage and is associated with high morbidity and mortality.
Buried bumper syndrome
Buried bumper refers to displacement of the gastrostomy tube internal bumper into the PEG tract anywhere between the gastric and anterior abdominal walls. It is a rare but serious complication with reported incidence of 0.3% to 2.4% in most studies, although one study reported an incidence of 9%. Buried bumpers usually present with leakage of tube feeds around the PEG site, inability to push feeds through the PEG, pain and swelling at the site, and in some cases with frank peritonitis ( Fig. 2 ). The diagnosis should be highly suspected in patients who have pain out of proportion to the abdominal examination findings and in those with signs of peritonitis and can be confirmed by either endoscopy or by computed tomography scan that can show the exact location of the bumper. All buried bumpers should be removed, even in asymptomatic patients. Sometimes the PEG can be removed by using the pull technique only or it can be pushed back in to the stomach (if the location is intramural) but in cases of subcutaneous infections or extramural location of the bumper, surgical intervention is needed. There are some case reports that have shown removal of buried bumper by inserting a through-the-scope balloon dilator and inflating it over a wire before pulling the tube out of the patient’s mouth or by making a small gastric incision using a needle knife. The main reason for the buried bumper is believed to be excessive traction between the bumper and the abdominal wall face plate leading to pressure necrosis and subsequent migration. The factors that can contribute to this process include tightening of the bumper after placement, weight gain after PEG placement, obesity, chronic cough (leading to increased intra-abdominal pressure), and placement of gauze or bandage below the external skin bumper to keep the site clean and dry adding tension to the bumper. We recommend that the external skin bumper should always be kept loose and easily rotatable at the skin site to prevent buried bumper and ischemic ulceration.
Related posts:
Comprehensive Review of Adverse Events Associated with GI Endoscopy
Minimizing, Recognizing, and Managing Endoscopic Adverse Events
Complications of Enteroscopy: How to Avoid Them and Manage Them When They Arise
Endoscopic Retrograde Cholangiopancreatography–Related Adverse Events
Endoscopic Retrograde Cholangiopancreatography (ERCP)-Related Adverse Events
Closing Perforations and Postperforation Management in Endoscopy
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
