An anastomotic leak is a serious complication of RYGB and can result in life-threatening sequelae (Fig. 2.2). The incidence of leak after laparoscopic RYGB ranges from 0.3 to 4.3 % [7, 8]. A recent multicenter study of 4444 patients who underwent RYGB reported an anastomotic leak rate of 1.0 % [9]. No specific technique for the gastrojejunostomy was associated with an increased rate of leak. However, this study revealed a statistically significant increase in the rate of anastomotic leak among patients who had open surgery, revisional surgery, and placement of an abdominal drain.

A high suspicion for an anastomotic leak must be maintained in a patient with persistent unexplained sustained tachycardia exceeding 120 beats per minute, even in the absence of radiologic findings of a leak [10]. In addition to tachycardia, a sensitive sign of leak is an increase in oxygen requirement of the patient. Other symptoms include abdominal pain, nausea, vomiting, and a feeling of impending doom. It cannot be overemphasized that persistent tachycardia should not be dismissed because a patient is afebrile with a benign abdominal exam, normal white blood cell count, or negative upper gastrointestinal imaging. Negative contrast studies may be falsely negative and should not delay treatment. The sensitivity of swallow studies conducted on postoperative day 1 after RYGB has been reported as 25 %, with a positive predictive value of 31 % [11]. In contrast, sustained tachycardia has been shown to be a reliable indicator of anastomotic leak [10, 12].

Initial management consists of volume resuscitation and coverage with broad-spectrum antibiotics and antifungal therapy. While nonoperative management of anastomotic leak has been described [13, 14], early surgical treatment of leaks is associated with a shorter hospital stay [15]. Standard of care for patients with anastomotic leaks consists of prompt reoperation for abdominal washout, repair of the leak, and drainage. Enteral access should be strongly considered; a gastrostomy tube may be placed in the gastric remnant. Nonoperative management may be appropriate in select cases depending upon the patient’s clinical status and availability of local expertise with the use of nonoperative techniques. If the patient had surgery or is being treated for the complication in a community hospital setting, transfer to a tertiary bariatric center which has the ability to rescue the patient is optimal soon after the patient is initially stabilized.

Early small bowel obstructions after RYGB are most commonly caused by technical problems such as narrowing or angulation at the jejunojejunostomy [16] or bleeding [5, 17]. Other possible etiologies include Roux-en-O configuration [18], a twist of the Roux limb, or obstruction at the transverse mesocolon [16, 19]. Bilious vomiting suggests a Roux-en-O configuration [18, 20], gastrogastric fistula, or obstruction distal to the jejunojejunostomy [20]. Therefore, prompt exploration should be undertaken for early bowel obstructions to rule out technical complications. Nausea, vomiting, and dysphagia may be presenting symptoms of a gastrojejunostomy stricture, which is diagnosed by upper endoscopy. The presentation occurs 1–3 months postoperatively, with 90 % presenting between 30 and 60 days and 10 % between 60 and 90 days [21]. Endoscopic dilation is generally successful after 1–3 dilations [21].

Shimizu et al. reported that 0.5 % of laparoscopic RYGB patients underwent surgical management for small bowel obstruction within 30 days after surgery [17]. In this study, all of the patients had undergone antecolic-antegastric RYGB. The 11 patients with early small bowel obstruction were diagnosed by CT with oral contrast and underwent laparoscopic exploration. Causes included a kink at the jejunojejunostomy, intraluminal blood clot, intra-abdominal hematoma, and pelvic adhesions. Bowel resection was only required in one patient and four patients were converted to laparotomy. Endoscopy was a valuable adjunct for assessing hemostasis, decompressing the Roux limb, and confirming patency at the jejunojejunostomy.

An additional cause of early small bowel obstruction is incarceration of small bowel in an unrepaired ventral hernia. Ventral hernias may be left open to be more optimally repaired after the patient has lost a substantial amount of weight, as the hernia may reoccur if a repair is attempted when the patient still carries a substantial amount of weight. When a ventral hernia is identified prior to surgery, patients should be consented for simultaneous hernia repair with possible mesh placement. The use of synthetic mesh raises a concern for mesh infection because of the GI anastomoses. If a ventral hernia is detected at the time of surgery, a postoperative conversation should ensue with the patient to explain the intraoperative decision making, signs and symptoms of small bowel obstruction, and specific instructions regarding postoperative activity restrictions. In all cases, if a ventral hernia is left unrepaired, a specific note should be made in the operative record, in case the small bowel obstruction presents to another surgeon/team.

The sleeve gastrectomy (SG) was initially described as a component of the duodenal switch [22, 23]. It was subsequently proposed as the initial procedure in a two-stage approach for high-risk patients [24]. Over the past decade, it has gained tremendous popularity as a primary bariatric surgical procedure [25] and has now surpassed the RYGB in volume at academic medical centers [26].

In a recent literature review, there were no significant differences in overall complication rates between RYGB and SG; the reported leak rate for SG was 2.3 % versus 1.9 % in RYGB [27]. The rates of bleeding and stenosis were slightly but significantly higher for SG but there was no significant difference in leak rates between RYGB and SF . In a single-institution study comparing outcomes among SG, RYGB, and DS, the rate of leakage for RYGB and SG was similar but hemorrhage was more frequent after SG [28].

LAGB patients presenting with abdominal pain, nausea, intractable reflux, or intolerance of oral intake should have the fluid from the band withdrawn. This is accomplished by accessing the subcutaneous port with a non-coring needle. A plain abdominal X-ray should also be done to evaluate for band slippage. In addition, one should assess the tubing from band to port on plain films to rule out port disconnection. If emptying the fluid does not resolve symptoms or plain films suggest band slippage, a UGI study may be done for further evaluation (Fig. 2.4).

Complications requiring reoperation from LAGB within the first 30 days are rare. In a study of over 6,000 LAGB patients, 14 (0.2 %) patients required emergency surgery for a complication related to the band [37]. The median time of presentation was 19 months and ranged from 1 to 61 months. The most common complication was band slippage with or without gastric necrosis. Other complications included small bowel obstruction, perforated gastric ulcer, bowel penetration, and port disconnection.

Symptoms of acute band slippage include epigastric pain, dysphagia, gastroesophageal reflux, and vomiting [38]. A contrast study may reveal anterior or posterior prolapse in which the anterior or posterior wall of the stomach herniates above the band. The prolapsed stomach may become ischemic or necrotic [39]. Therefore, emergent surgical intervention is indicated. If gastric ischemia or necrosis is present, urgent explantation would be required. The band may be removed laparoscopically by cutting the tubing first distally which traverses the abdominal wall and then mobilizing the soft tissue where it connects to the band, incising the capsule anterior to the band, and then mobilizing and releasing the band from the stomach. The band is then removed from a 15 mm trocar site and the subcutaneous port is then dissected free and removed. If the stomach is viable, repositioning of the band at the gastroesophageal junction may be feasible [40] only if the patient has had satisfactory weight loss with the band. If the patient is a partial or nonresponder, consideration for removal only should be given, with plans to convert to a metabolic procedure after the acute situation is resolved.

Port-site infections in the early postoperative period may be treated with antibiotics if infection is limited. Band erosion should be suspected if the access port is erythematous or tender [41]; this is best evaluated with upper endoscopy and would require band removal. However, band erosion is rarely seen in the early postoperative period. In the absence of band erosion, if there is an abscess at the port site or if there is no response to antibiotics, the port should be disconnected and removed and the band tubing should be left inside the peritoneal cavity for future reconnection to a new access port [42]. Damage to the port may be diagnosed by loss of volume of the injected fluid and failure of satiety despite band fills. Port leakage or leakage along the band tubing may be diagnosed by injecting contrast into the port under fluoroscopy. Port removal or band replacement may be necessary and this may be done electively.