INTRODUCTION — Bariatric operations are an effective method of reducing obesity and obesity-linked medical illnesses. They can be challenging due to the anatomic and physiologic characteristics and comorbidities of patients with obesity.
This topic will review the major perioperative (intraoperative and early [≤30 day] postoperative) complications and mortality rates of common bariatric surgical procedures, including laparoscopic and open Roux-en-Y gastric bypass, laparoscopic sleeve gastrectomy, biliopancreatic diversion with duodenal switch, and adjustable gastric banding. It will also include available data on the single-anastomosis duodenal ileal bypass with sleeve (SADI-S) and the one-anastomosis gastric bypass (OAGB).
The bariatric procedures, indications and preoperative management, medical outcomes, and long-term complications of laparoscopic operations are reviewed in separate topics.
●(See "Bariatric procedures for the management of severe obesity: Descriptions".)
●(See "Bariatric surgery for management of obesity: Indications and preoperative preparation".)
●(See "Outcomes of bariatric surgery".)
●(See "Late complications of bariatric surgical operations".)
INCIDENCE — Serious intraoperative and early postoperative complications and mortality rates are relatively low when the surgery is performed by experienced surgeons at established bariatric centers.
●Intraoperative morbidity rates – In contemporary series, the overall rate of intraoperative complications from bariatric surgery ranges from 0.69 to 5 percent [1-4]. The rate of intraoperative complications varies with the procedure [1] (see 'Intraoperative complications' below):
•7.3 percent for open Roux-en-Y gastric bypass (RYGB).
•5.5 percent for laparoscopic RYGB.
•3.0 percent for laparoscopic adjustable gastric banding (AGB).
•The rate for laparoscopic sleeve gastrectomy (SG) appears to be somewhere between that of laparoscopic RYGB and laparoscopic AGB [5].
•The data regarding the rate of intraoperative complications for one-anastomosis gastric bypass (OAGB) are less robust, but it is estimated at 0.5 percent in one multicenter study [6] and up to 4.63 percent in a single-center study [7].
•The data for the single-anastomosis duodenal ileal bypass with sleeve (SADI-S) are limited to single-institution cohorts, in which no intraoperative complications were reported [8,9].
●Early postoperative morbidity rates – The overall rate of a major complication occurring within 30 days ranges from approximately 0.2 to 10 percent in contemporary series [4,10-15] and varies with the operative procedure, operative approach (open versus laparoscopic), and other factors [13] (see 'Postoperative surgical complications' below and 'Postoperative medical complications' below):
•The 30 day morbidity rate associated with laparoscopic SG is generally less than is seen with laparoscopic RYGB (0.8 to 5.6 versus 1.4 to 9.4 percent) [5,16,17].
•The 30 day morbidity rate associated with laparoscopic RYGB is lower than that of open RYGB (3.4 versus 7.4 percent) [10].
•Biliopancreatic diversion with duodenal switch (BPD-DS) appears to have the highest morbidity rates [4].
•The 30 day morbidity rate associated with SADI-S is 2.65 [18] to 3.3 [8] percent.
•The 30 day morbidity rate associated with OAGB is 5 percent [19]
●Mortality rates – In a meta-analysis of 58 studies (3.6 million patients), the pooled mortality rate of bariatric surgery was 0.08 percent (95% CI 0.06-0.10) [20]. Thirty-day, 90 day, in-hospital, and overall mortality rates were not statistically different, and neither were the mortality rates reported in trials, large cohort studies, or national registry/database studies. (See 'Mortality' below.)
This has dramatically improved since the early 2000s and now compares favorably with the hospital mortality of other frequently performed major surgical procedures, including hip replacement (0.3 percent), abdominal aneurysm repair (3.9 percent), craniotomy (10.7 percent), esophageal resection (9.1 percent), and pancreatic resection (8.3 percent) [21].
The pooled mortality rates by procedure were [20]:
•0.03 percent for AGB
•0.05 percent for laparoscopic SG
•0.09 percent for OAGB
•0.09 percent for RYGB
•0.41 percent for BPD-DS
•0.6 percent for SADI-S [22]
●Readmission and reoperation rates – Complications from bariatric operations may lead to rehospitalization and reoperation [23-25]. In large database studies, the readmission rate has decreased from 5 percent [26,27] between 2009 and 2015 to 2.75 percent between 2015 and 2018 [28]. Short-term readmission rates are higher for RYGB than for other procedures such as SG [13,28-30]. The OAGB has a similar rate of 5 percent [31]. Hospital readmission rates are higher after open compared with laparoscopic procedures [1,4,10,32,33].
The rate of reoperation within 30 days of the initial procedure ranges from 0.7 to 7.6 percent [4,26,34-37]. The reoperation rates tend to be higher with more complicated procedures, such as BPD-DS and the SADI-S. The rate is more than twice as much for the BPD-DS versus RYGB (3.3 versus 1.5 percent) [4] and for the RYGB (0.7 to 5 percent) versus SG (0.5 to 3 percent) [13,29,30]. The reoperation rate also tends to be higher with the laparoscopic versus the open approach (7.6 versus 6.6 percent) [34]. The SADI-S appears to have a higher reoperation rate than the RYGB (5.0 versus 2.6 percent); however, the SADI-S also appeared to be used in patients with a higher body mass index (BMI), indicating a higher inherent risk [22]. The reoperation rate for the OAGB is estimated at 1.01 percent [38].
●Trend – In a 2022 registry-based analysis of 690,770 patients who underwent RYGB and SG between 2015 and 2019 at centers participating in the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP; 73 percent SG), there was a significant relative reduction in readmissions, end-organ dysfunction, and all-cause mortality in the SG subgroup [39]. There were also noticeable reductions in 30 day readmission, intensive care unit admission, and end-organ dysfunction in the RYGB subgroup. For complications, bleeding, postoperative pneumonia, and stroke were significantly decreased in the RYGB subgroup, but only bleeding was significantly decreased in the SG subgroup during this five-year study period.
RISK FACTORS AND PREVENTION — Higher postoperative complication rates are associated with certain procedures, approaches, and adverse intraoperative events. Certain steps can be taken pre-, intra-, or postoperatively to mitigate the higher risk [1-3,10,40-43].
Risk factors for early postoperative morbidities — General risk factors for early postoperative morbidities include the use of complex procedures, the use of the open approach, and acute intraoperative events (AIEs). A few studies have assessed risk factors for specific complications [35,44-48]. They will be discussed below in the respective sections.
●Complexity of procedures – The morbidity rate is dependent upon the type of procedure performed [4,10,49-60]. In general, biliopancreatic diversion with duodenal switch (BPD-DS) is associated with the highest morbidity rate [49], while laparoscopic adjustable gastric banding (AGB) is associated with the lowest [10]. Laparoscopic Roux-en-Y gastric bypass (RYGB) and laparoscopic sleeve gastrectomy (SG) have fewer complications than BPD-DS but more complications than laparoscopic AGB. Laparoscopic RYGB and laparoscopic SG have similar major complication rates, but laparoscopic RGYB is associated with more minor complications than SG [50]. The single-anastomosis duodenal ileal bypass with sleeve (SADI-S) and one-anastomosis gastric bypass (OAGB) have early complication rates similar to those of the RYGB but also lower than those of the BPD-DS [18,31].
●Open versus laparoscopic approach – The American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP) database showed that patients who had undergone an open approach to an RYGB had significantly higher morbidity than those who underwent a laparoscopic one [10,27]. The patients who had undergone a laparotomy had higher rates of major complications (7.4 versus 3.4 percent), more return visits to the operating room (4.9 versus 3.6 percent), and longer postoperative length of stay (median three versus two days) than patients who had laparoscopic RYGB [10].
●Intraoperative complications – An AIE, along with conversion from a laparoscopic to an open procedure, increases the risk for serious postoperative complications (eg, pulmonary embolus, myocardial infarction) [1,2]. Patients who experience an AIE are more than twice as likely to develop a major postoperative complication compared with patients without an adverse event (8.8 versus 3.9 percent, p<0.001).
Prevention of complications — Certain steps can be taken prior to, during, and after bariatric operations to minimize complications.
Preoperative steps — Preoperative management of bariatric patient is discussed in another topic. (See "Bariatric surgery for management of obesity: Indications and preoperative preparation".)
●Preoperative evaluation – A thorough preoperative assessment of comorbid illnesses (eg, diabetes, hypertension, cirrhosis, obstructive sleep apnea) and effective preoperative management as well as an assessment of anesthetic risks must be performed prior to the operative procedure. A higher risk of morbidity is noted in male bariatric patients, those with a body mass index (BMI) >50 kg/m2, those with dysmetabolic syndrome, and those over age 65 [47]. Immobile or functionally dependent patients, in particular, have a very high risk of complications [61]. (See "Bariatric surgery for management of obesity: Indications and preoperative preparation", section on 'Preoperative assessment'.)
●Preoperative weight loss – There is evidence that a low- (800 to 1200 kcal) or very-low- (400 to 800 kcal) calorie diet for 2 to 12 weeks may induce sufficient weight loss to improve the technical ease of bariatric surgery by reducing the volume of the liver (by 15 to 30 percent) and overall adiposity [62]. However, it is unclear whether preoperative weight loss can reduce the risks of 30 day postoperative complications [63,64]. (See "Bariatric surgery for management of obesity: Indications and preoperative preparation", section on 'Preoperative weight loss'.)
●Venous thromboembolic (VTE) prophylaxis – Prophylaxis against VTE following bariatric operations is discussed separately. (See 'Venous thromboembolism' below.)
Intraoperative steps — Details of common bariatric procedures are discussed in other topics. (See "Laparoscopic Roux-en-Y gastric bypass" and "Laparoscopic sleeve gastrectomy".)
●Instrument and equipment condition – Prior to the patient receiving an anesthetic, the surgeon and appropriate operating room personnel should examine all instruments for availability and function. The usual length of laparoscopic equipment is 38 cm long, but lengths up to 43 cm are available for laparoscopic instruments, surgical energy devices, and staplers. It would be important for the surgeon to know if their hospital has such equipment at its disposal. (See "Instruments and devices used in laparoscopic surgery".)
●Proper placement of trocars – Torque that occurs from improperly placed trocars can be minimized by repositioning or adding additional ports. For example, for a patient with a very large abdomen, it may be difficult to access the gastroesophageal junction and infracolic abdomen to perform an RYGB anastomosis. In this setting, the use of additional ports would likely facilitate the procedure (figure 1). (See "Abdominal access techniques used in laparoscopic surgery", section on 'Foregut surgery'.)
●Tension-free anastomosis – Tension on the gastrojejunal anastomosis when performing an RYGB is a risk factor for anastomotic leak (figure 2). A retrocolic, retrogastric approach creates the least amount of tension on the anastomosis (figure 3). While the antecolic, antegastric route may be technically easier to perform, this route may also create a greater amount of tension on the anastomosis. Bariatric surgeons should be comfortable with performing both Roux limb routes. (See "Bariatric procedures for the management of severe obesity: Descriptions", section on 'Roux-en-Y gastric bypass'.)
●Defining a plane between adhesions and bowel – Regardless of whether adhesions are lysed with an electrosurgical device or scissors, defining and maintaining a clear plane of dissection will minimize risk to the bowel wall. Prompt identification of injury and primary repair of the inadvertent laceration will minimize the risk of a subsequent leak. If the procedure cannot be safely performed laparoscopically because of dense adhesions, conversion to an open approach may enable better visualization and safer lysis of adhesions. (See "Complications of laparoscopic surgery", section on 'Gastrointestinal injury'.)
Postoperative steps — Enhanced recovery after surgery (ERAS) protocols, also called fast-track protocols, are associated with shorter lengths of hospital stay and lower postoperative morbidity [65]. Many ERAS protocols have the added benefit of avoiding the use of narcotics for postoperative pain control [66,67]. ERAS protocols or care pathways have been published for bariatric surgery by major societies [68-72].
INTRAOPERATIVE COMPLICATIONS — The most common intraoperative complications for laparoscopic Roux-en-Y gastric bypass (RYGB) and laparoscopic adjustable gastric banding (AGB) operations, based upon the Longitudinal Assessment of Bariatric Surgery (LABS) database, include [1]:
●Anesthesia events (1.0 percent)
●Instrument/equipment failure (0.8 percent)
●Bowel injury (0.8 percent)
●Hepatic injury (0.4 percent)
●Anastomosis revision (0.3 percent)
●Splenic injury (0.2 percent)
●Major blood vessel injury (0.1 percent)
Based upon the retrospective review of 504 patients in the Bariatric Outcomes Longitudinal Database (BOLD) database undergoing an open RYGB, the overall adverse intraoperative event rate was 7.3 percent [1]. The most common adverse intraoperative complications for open procedures included:
●Revision of anastomosis (2.0 percent)
●Bowel injury (1.0 percent)
●Instrument/equipment failure (0.6 percent)
●Anesthesia events (0.4 percent)
●Hepatic injury (0.2 percent)
●Major blood vessel injury (0.2 percent)
Anesthetic events — Operating on the individual with obesity can be technically challenging even for the most experienced anesthesia and surgical teams. The anatomic and physiologic characteristics of obesity are associated with difficult airway and drug management. The challenges and strategies of anesthetic management during bariatric surgery are discussed separately. (See "Anesthesia for the patient with obesity".)
Laparoscopic access-related injuries — Abdominal access can be challenging in the patient with a thick abdominal wall; however, all types of entry access can be safely performed by experienced surgeons [73-76]. Based upon 17 randomized trials that included 3040 patients, there was no statistical difference in the rates of major complications (eg, vascular or solid organ injury) for open or closed (Veress needle, standard or optical trocar) laparoscopic entry techniques [76]. Access techniques and access-related complications for laparoscopic procedures are reviewed separately. (See "Abdominal access techniques used in laparoscopic surgery", section on 'Peritoneal access'.)
Many bariatric surgeons prefer a combined approach. They first obtain pneumoperitoneum using the Veress needle technique through Palmer's point, then place a port in the left upper quadrant using an optical view method. Placing the patient in the 5 to 10 degree reverse Trendelenburg position may drop some of the intra-abdominal and subcutaneous fat into a more caudad and inferior position, allowing easier access to the upper abdomen.
●When using the Veress needle technique, the left upper quadrant is preferred for initial placement (Palmer's point). The aspiration test, the saline drop test, and an opening pressure of less than 10 mmHg should all be used to confirm proper placement of the needle. The area beneath the Veress needle insertion site inside the abdomen should be inspected for injuries during the initial laparoscopic evaluation of the abdomen [73-76].
●When using the Hassan technique for patients with a large amount of subcutaneous fat, the incision should be made large enough to identify the abdominal wall fascia and peritoneum.
Many structures are at risk of injury, including major vessels and solid and hollow organs, during the introduction of the trocar or Veress needle. Some injuries can be life-threatening. In the supine position, the aorta and inferior vena cava are only 2 to 3 cm below the skin, so blind entry here can result in a major vascular injury. Bariatric patients tend to have a longer distance but may require more force to access the peritoneum because of the amount of subcutaneous tissue. Subsequently, they can still be at risk for vascular injuries.
Bowel injury, such as a puncture or laceration, can occur with insertion of a Veress needle, trocar, lysis of adhesions, or dissection. (See "Complications of laparoscopic surgery", section on 'Gastrointestinal puncture' and "Complications of laparoscopic surgery", section on 'Gastrointestinal injury'.)
Solid viscus organ injury — The liver and spleen can be injured during bariatric surgical procedures because of their close proximity to the stomach (figure 4), especially during sleeve gastrectomy (SG) (picture 1) [77]. Bleeding can usually be controlled with suturing or hemostatic agents, and a partial hepatectomy or splenectomy is rarely required [77].
In addition, patients with cirrhosis, especially if found incidentally, have a higher risk of splenic vein bleeding, and consideration should be given to aborting the operation, especially if previously undiagnosed gastric varices are found. (See "Cirrhosis in adults: Overview of complications, general management, and prognosis" and "Hemostatic abnormalities in patients with liver disease" and "Portal hypertension in adults".)
Portal vein, inferior vena cava injuries — Intraoperative injury to the portal vein and inferior vena cava is rare during bariatric operations but can lead to rapid exsanguination [78]. If thrombosis of the portal vein occurs following an injury, it can lead to mesenteric venous thrombosis, and such patients may require liver transplantation [78]. (See "Complications of laparoscopic surgery", section on 'Major vessels' and "Acute portal vein thrombosis in adults: Clinical manifestations, diagnosis, and management".)
Bowel ischemia — During the performance of RYGB or biliopancreatic diversion with duodenal switch (BPD-DS) procedures in particular, ischemia of the bowel may result if the root of the mesentery containing the major mesenteric vessels is compromised or undermined in one of the following scenarios [79-82]:
●Twisting the mesentery and/or mesenteric root while creating a retrocolic pathway for the Roux limb
●Creating tension on the Roux limb when performing a gastrojejunal anastomosis
●Incising the mesentery of the biliopancreatic limb or Roux limb at an angle such that blood flow to an end of intestinal limb is compromised
●Closing the mesenteric defect with overly generous amounts of tissue between sutures
Misconstruction — Misconstruction of an RYGB or a BPD-DS (figure 5) can occur during a laparoscopic or open approach. Proper bowel orientation must be maintained to avoid misconstruction.
Misconstructions of the RYGB with an anastomosis between the distal aspect of the biliopancreatic limb and the gastric pouch have been reported and can lead to chronic bilious vomiting ("Roux-en-O") (image 1 and image 2 and figure 6 and figure 7) [83]. Other examples of misconstruction include Roux limb inversion (image 3 and figure 8) and short Roux limb (figure 9 and image 4).
POSTOPERATIVE SURGICAL COMPLICATIONS — Early (≤30 day) surgical complications include gastrointestinal leaks, hemorrhage, surgical site infections, marginal ulcers, anastomotic or sleeve stenosis, small bowel obstruction, oral intake intolerance, acute stoma obstruction after adjustable gastric banding (AGB), and portomesenteric vein thrombosis [11,12,34,44,84-92].
Gastrointestinal leak — An anastomotic or staple line leak remains the most dreaded technical complication of bariatric surgery [12,34,85-90,93-98].
●Incidence – The risk of a leak ranges from 0.8 to 6 percent depending on the procedures chosen as well as technical and patient factors involved [4,14,99-102].
•From a Roux-en-Y gastric bypass (RYGB), leaks can occur at the gastrojejunal or jejunojejunal anastomosis at a rate of approximately 1.0 percent [100].
•From a sleeve gastrectomy (SG), leaks can occur at the staple line at a rate of up to 7 percent, but generally the rate has decreased to around 1.5 percent [103].
•From a biliopancreatic diversion with duodenal switch (BPD-DS), leaks can occur from the anastomoses or from the long stomach staple line at a rate of 5 percent [104].
•From a one-anastomosis gastric bypass (OAGB), leaks can occur at the gastrojejunal anastomosis at a rate of 1.8 percent [31].
•From a single-anastomosis duodenal ileal bypass with sleeve (SADI-S), leaks can occur at the duodeno-ileal anastomosis or from the long stomach staple line at a rate of 2.2 percent [22].
•In revisional bariatric surgery, the risk of anastomotic leak approaches 35 percent [99].
●Timing – Most leaks after bariatric surgery occur early, generally within the first week after surgery, but can be either before or after the patient has been discharged [14]. Therefore, vigilant follow-up during the first 30 days is recommended.
●Clinical presentation – Symptoms of a gastrointestinal leak include abdominal pain, oral intolerance, and a "feeling of doom." Sometimes the presentation can be subtle and require vigilance for signs such as low-grade fevers, respiratory compromise, and/or unexplained sustained tachycardia greater than 120 bpm [105]. The same signs may also be present in the setting of pulmonary embolism. (See 'Venous thromboembolism' below.)
●Evaluation – A leak may be radiographically confirmed by barium swallow or contrast computed tomography (CT) (image 5 and image 6) or intraoperatively at the time of surgical exploration [106-110].
●Management – The most important aspect of treatment is early recognition. Any patient with sustained tachycardia above 120 beats per minute should be admitted to the hospital, fluid resuscitated, and closely observed for worsening tachycardia, the development of sepsis, or hemodynamic instability, all conditions that would prompt surgical exploration with possible intraoperative endoscopy to rule out the leak definitively.
If a leak is suspected clinically in patients who are hemodynamically unstable, emergency surgical exploration should be performed, even if the imaging is negative, given that such patients can rapidly progress to sepsis [109,111]. Multiple studies have shown that re-exploration is safe and avoids missing or delaying the treatment of a leak [112-114]. In experienced hands, surgical exploration and management of a leak is often feasible via a laparoscopic approach. A negative exploration that does not reveal a leak should not be considered a complication due to the need for reoperation but rather evidence of good surgical judgment. The surgical principles in treating a leak include (figure 10):
•Broad-spectrum antibiotic coverage
•Identification and repair of the defect
•Irrigation and control of contamination
•Wide external drainage of the contaminated area
•Enteral access for feeding (eg, a gastric feeding tube placed in the gastric remnant for RYGB)
For hemodynamically stable patients who have a small, contained leak, radiology-guided placement of drainage catheters or endoscopic placement of clips, fibrin glue, stents, or negative pressure suction has all been used with success [113,115-118]. (See "Gastrointestinal endoscopy in patients who have undergone bariatric surgery", section on 'Anastomotic leak and fistula'.)
Leaks from an SG generally take a longer time to close than leaks from other bariatric procedures. The sleeved stomach is a high-pressure system due to the presence of sphincters at both ends. The pressure may cause a leak to persist. Additionally, the leak may be caused by a stenosis in the sleeve, a kink in the sleeve's course, or a twist in the sleeve. Consequently, leaks may not close without alleviating the partial obstruction from these entities, which may ultimately require a reoperation [119]. (See 'Stenosis' below.)
●Prevention – Many different intraoperative techniques have been used to decrease the incidence of leaks. Over-sewing staple lines, buttressing materials for the staple line, and fibrin glue have been used, but there is no evidence that these methods reliably decrease the leak rate after RYGB [120-127] or SG [128].
Methylene blue dye or endoscopy can be used intraoperatively to test for a leak during an RYGB or SG. For the dye test, methylene blue-tinged saline is injected via the nasogastric tube while the bowel distal to the gastrojejunostomy is gently occluded. A leak is detected by the presence of methylene blue around the staple line. Alternatively, the pouch can be evaluated for a leak under direct visualization with an endoscope. For endoscopic evaluation, the gastrojejunostomy or sleeve is submerged in saline and the Roux limb or distal stomach is gently occluded, then the endoscope is used to insufflate the pouch. The presence of air bubbles in the saline around the pouch during insufflation indicates a leak may be present. Whether postoperative upper gastrointestinal series should be ordered routinely or selectively remains controversial and is discussed elsewhere [106,109,110]. (See "Imaging studies after bariatric surgery".)
Hemorrhage — Significant hemorrhaging after bariatric surgery has been described in 0.4 to 4.0 percent of patients [10,12,15,34,85-90]. Bleeding can occur from the gastric or short gastric vessels during dissection of the greater curve. Most of the bleeding problems associated with an SG occur from the staple line after transection of the stomach [53,129,130]. The bleeding is most likely a result of the large staples used for the thick tissue in the distal stomach. Large staples are not adequate to seal small vessels [131]. This has led many surgeons to reinforce the staple line by over-sewing, buttressing, or both [52].
Early bleeding can also occur from an anastomosis, which most commonly bleeds intraluminally. Patients frequently present with tachycardia, a decreased hematocrit, hematochezia, or melena [91]. Such bleeding typically resolves without surgical intervention but may require transfusion of blood products and reversal of anticoagulation [12,85,88]. Careful endoscopic examination and therapy is appropriate for ongoing bleeding with high transfusion requirements. (See "Gastrointestinal endoscopy in patients who have undergone bariatric surgery", section on 'Gastrojejunal ulceration and bleeding'.)
Surgery is reserved for hemodynamic instability, intraluminal bleeding not amenable to endoscopic therapy (eg, staple line of the excluded stomach), extraluminal bleeding, or continued bleeding despite restoration of normal coagulation status [91,92]. Patients with early postoperative bleeding had a longer hospital admission (4.8 versus 3.0 days, p<0.001) and a higher mortality rate (7.1 versus 0.9 percent, p<0.01) compared with patients without an early bleed [15].
Surgical site infections — The Bariatric Outcomes Longitudinal Database (BOLD) study reported an overall incidence of surgical site infection (SSI) of 1.1 percent [4]. Rates of wound infection are significantly greater for open gastric bypass procedures (10 to 15 percent) compared with laparoscopic (3 to 4 percent) procedures [34,78,85].
Signs of wound infections include unexplained fever, fluctuance, erythema, or drainage. Treatment consists of opening the affected area for drainage of infected fluid or pus, debridement of any devitalized tissue, and antibiotics if the surrounding skin suggests signs of cellulitis. (See "Complications of abdominal surgical incisions".)
The incidence of wound infections can be decreased by perioperative administration of antibiotics. Data support the use of cefazolin, or in patients allergic to penicillins and cephalosporins, clindamycin PLUS one of the following: ciprofloxacin, levofloxacin, gentamicin, or aztreonam [132]. The choice of antibiotics is reviewed in more detail separately. (See "Antimicrobial prophylaxis for prevention of surgical site infection following gastrointestinal procedures in adults", section on 'Gastroduodenal procedures'.)
Marginal ulcers — Marginal ulcers typically occur within 30 days after the RYGB operations at the gastrojejunal anastomosis (picture 2) [133] and at the gastroileal junction following a BPD-DS [134], despite bypass or removal of most of the acid-secreting cells of the stomach. The risk of a marginal ulcer following bariatric operations based upon the BOLD study was 1.2 percent [4]. The incidence of a marginal ulcer following a BPD-DS was 0.3 percent [134], while the incidence of a marginal ulcer following a BPD only was 3 percent [135].
The clinical presentation of marginal ulcers includes melena, hematochezia, hematemesis, or pain and obstructive symptoms as the swollen mucosa can occlude the anastomosis to the small intestine [136]. Early ulceration most likely is related to ischemia, too large of a pouch, or undiagnosed Helicobacter pylori infection. (See "Bacteriology and epidemiology of Helicobacter pylori infection" and "Indications and diagnostic tests for Helicobacter pylori infection in adults".)
Risk factors for a marginal ulcer include smoking, nonsteroidal anti-inflammatory drugs (NSAIDs), and gastrogastric fistula formation. Most surgeons will place their patients on a proton-pump inhibitor for 3 to 12 months after surgery to help prevent the occurrence of marginal ulcers. Lifetime avoidance of smoking and NSAIDs is advised in RYGB and BPD-DS patients as marginal ulcers can also develop as a late complication. (See "Late complications of bariatric surgical operations", section on 'Marginal ulcers'.)
Stenosis — Stenosis occurs when the gastrojejunal anastomosis of an RYGB or gastroileal anastomosis of a BPD-DS becomes occluded secondary to tissue edema or a narrow anastomosis. Following an SG, stenosis can occur if the sleeve narrows, kinks, or twists. A retrospective review of 230 patients found that the risk of stenosis following an RYGB, a BPD-DS, or an SG is approximately 3.3, 1.2, and 3.5 percent, respectively [137]. Another study of over 180,000 patients from the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP) database reported that the overall incidence of early gastrojejunal stricture after RYGB (within 30 days) was 4.7 per 1000 person-years, of which 85, 40, and 5.6 percent required therapeutic intervention, readmission, and reoperation, respectively [138].
The clinical presentation is consistent with an obstruction and includes nausea, vomiting, abdominal pain, and dysphagia. Endoscopic balloon dilation of the stenotic structures, including those resulting from an SG, is the first management option. [118]. (See "Gastrointestinal endoscopy in patients who have undergone bariatric surgery", section on 'Stomal (anastomotic) stenosis'.)
Small bowel obstruction — Small bowel obstruction (SBO) can occur from misconstruction during RYGB or BPD-DS operations, internal herniation, adhesion formation (in both open and laparoscopic procedures), or port site hernia. Early SBO is usually caused by a technical error such as the failure to reapproximate the fascia from a port site or potential internal hernia sites. After an RYGB, an internal hernia can occur through the mesentery of the jejunojejunal anastomosis and posterior to the gastrojejunal anastomosis. If a retrocolic approach is used, an internal hernia can occur through the transverse mesocolon defect and through Petersen's defect, which is the space between the transverse mesocolon and the Roux limb as it passes through the mesocolon [139]. (See "Late complications of bariatric surgical operations", section on 'Internal hernias'.)
In a prospective study that included 2126 patients undergoing RYGB, early SBO (within 30 days) following bariatric procedures occurred in approximately 0.5 percent of patients [140]. Among 184,660 patients undergoing RYGB in the MBSAQIP database, 1189 (0.64 percent) required readmission due to SBO within 30 days, of which 69 percent required a reoperation and 1.3 percent expired [141]. The reasons for readmission were identified as unspecified intestinal obstruction (82.5 percent), incisional hernia (9.1 percent), and internal hernia (8.4 percent). The SADI-S appears to have fewer SBOs than the BPD-DS [18]. For the OAGB, the rate of bowel obstruction appears to be about 2.8 percent [142].
The clinical presentation and management of an SBO is reviewed elsewhere. (See "Etiologies, clinical manifestations, and diagnosis of mechanical small bowel obstruction in adults" and "Management of small bowel obstruction in adults".)
Nausea, vomiting, and poor oral intake — After bariatric surgery, it is not unusual for patients to experience some extent of nausea, vomiting, and mild food intolerance [29]. Generally, though, if patients can stay hydrated and tolerate a thin liquid diet, they will not need readmission or workup for the other causes of early morbidities. Patients with mild gastrointestinal symptoms are generally kept on a liquid diet for one to two weeks after the operation and slowly transition to more solid forms of food over the next one to two months. Surgical edema is typically the cause of these symptoms, which should be self-limited. Patients who continue to present with recurring or worsening symptoms should undergo further workup for more serious complications as discussed above.
Stomal obstruction — Acute stomal obstruction is an early complication that can occur in up to 14 percent of AGB patients [143-145]. Obstruction is usually caused by inclusion of excess perigastric fat, use of a band of insufficient diameter for the thickness of the tissue, or significant tissue edema. Patients usually present with persistent nausea, vomiting, and inability to tolerate secretions or oral intake. The diagnosis is confirmed with an upper gastrointestinal series demonstrating no passage of contrast beyond the band.
Acute stomal obstruction due to edema can initially be treated conservatively with nasogastric tube decompression until the edema subsides, although the potential for aspiration pneumonia and gastric ischemia exists [146]. Persistent obstruction requires surgical revision or removal of the band. Meticulous removal of excess perigastric fat at the time of initial band placement may help prevent this complication [147]. The use of larger-diameter bands may also help to reduce the incidence of acute postoperative obstruction.
Portal, splenic, and mesenteric vein thrombosis — Portomesenteric vein thrombosis is a rare complication of laparoscopic bariatric surgery, most commonly after SG. The risk is about 0.4 percent, and it may lead to increased mortality. Patients typically present with abdominal pain, nausea, vomiting, fevers, tachycardia, and leukocytosis. Diagnosis can be made by CT scan of the abdomen [148-150]. Therapy typically consists of anticoagulation therapy. (See "Laparoscopic sleeve gastrectomy", section on 'Portal vein thrombosis'.)
POSTOPERATIVE MEDICAL COMPLICATIONS — Early (≤30 day) medical complications include deep vein thrombosis, pulmonary embolism, myocardial infarction, and pulmonary complications.
Venous thromboembolism — Based upon information from multiple databases, the overall incidence of venous thromboembolism (VTE), including deep vein thrombosis (DVT) and pulmonary embolism (PE), ranges from 0.17 to 0.4 percent [3,151]. Furthermore, most VTEs occurred following discharge; 2 percent occurred intraoperatively, and 25 percent occurred prior to hospital discharge.
●Incidence – The rate of VTE identified in patients undergoing a Roux-en-Y gastric bypass (RYGB) and adjustable gastric banding (AGB) was 0.46 and 0.14 percent, respectively. The highest rates of VTE were identified in 4 percent of patients undergoing a biliopancreatic diversion (BPD) and 2.2 percent of patients undergoing a BPD with duodenal switch (BPD-DS). The incidence of VTE was 0.29 percent for laparoscopic procedures and 1.2 percent for open procedures [152].
●Risk factors – Risk factors for VTEs include body mass index (BMI) >50 kg/m2, a history of VTE, a history of a hypercoagulable disorder, pulmonary hypertension, venous stasis disease, poor functional status, open or revision surgery, and operative time >3 hours [151,153,154]. The most common risk factors associated with fatal PE include severe venous stasis disease, BMI >60 kg/m2, truncal obesity, and obesity-hypoventilation syndrome [154,155]. Patients undergoing revision bariatric surgery are at a higher risk of VTE [152].
●Evaluation – Diagnosis of PE in patients with obesity can be problematic because the use of standard diagnostic modalities (such as a nuclear lung scan, computed tomographic [CT] angiography, pulmonary angiography, and/or lower extremity duplex scan) may not be physically feasible.
●Management – Immediate anticoagulation is prescribed for patients for whom there is a high level of clinical suspicion. In patients in whom anticoagulation is contraindicated, a mechanical filter can be placed in the inferior vena cava to lower the risk of continued clot embolization (see "Placement of vena cava filters and their complications"). However, per the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP) data, the use of an inferior vena cava (IVC) filter even in high-risk patients did not seem to have a protective benefit [156].
●Prevention – A single optimal strategy for preventing VTE in the bariatric surgery setting has not been established [157]. Most bariatric surgeons use both pneumatic compression devices in conjunction with subcutaneous unfractionated or low-molecular-weight heparin and ambulation on the day of surgery [158]. The current American Society of Bariatric and Metabolic Surgeons (ASMBS) guidelines regarding thromboprophylaxis state that all bariatric patients receive mechanical prophylaxis and are ordered for early ambulation. Additionally, the surgeon may routinely use pharmacologic prophylaxis of either low-molecular-weight heparin or unfractionated heparin [159]. In one retrospective study of over 24,000 patients undergoing bariatric procedures, low-molecular-weight heparin was more effective than unfractionated heparin in preventing postoperative VTE while not increasing the rate of bleeding [160].
Preoperative risk stratification may be a useful tool for identifying high-risk patients for more aggressive prophylaxis [161]. In patients identified preoperatively as high risk for PE, due to previous medical history of VTE, PE, or venous stasis disease or inability to ambulate, perioperative chemoprophylaxis of low-molecular-weight heparin or unfractionated heparin may be prescribed for an extended period even after discharge from the hospital. As the average time to developing VTE is 21 to 28 days, many bariatric surgeons continue chemoprophylaxis for six weeks postoperatively. The use of preoperatively placed IVC filters is associated with higher rates of VTE, PE, other complications, and death. Their routine use is discouraged [162].
Prevention of postoperative VTE is reviewed in detail separately. (See "Prevention of venous thromboembolic disease in adult nonorthopedic surgical patients" and "Prevention of venous thromboembolism in adults undergoing hip fracture repair or hip or knee replacement".)
Myocardial infarction — The incidence of a postoperative myocardial infarction after a bariatric operation is approximately 0.2 percent [10]. (See "Perioperative myocardial infarction or injury after noncardiac surgery".)
Pulmonary complications — Major postoperative pulmonary complications include pneumonia and acute respiratory failure. Atelectasis is common after all types of surgery that require general anesthesia and is more prevalent in patients with obesity.
The following studies illustrate the incidence rates and risks for pulmonary complications and the sequelae:
●Based upon a retrospective review of the Bariatric Outcomes Longitudinal Database (BOLD) database, the overall incidence of postoperative pneumonia was 0.4 percent [4]. The incidence of pneumonia was significantly higher for patients treated with a BPD-DS (n = 1545) compared with patients treated with RYGB (0.9 versus 0.4 percent, p = 0.001).
●In a retrospective review of the National Inpatient Sample database that included 304,515 patients undergoing a bariatric operation, the overall rate of acute respiratory failure (ARF) was 1.35 percent [163]. The incidence was higher with an open compared with a laparoscopic procedure (3.87 versus 0.94 percent); RYGB compared with non-RYGB procedures (1.54 versus 0.82 percent); and for patients with preoperative heart failure, chronic renal failure, peripheral vascular disease, chronic obstructive pulmonary disease (COPD), tobacco smoking, and diabetes. ARF is associated with a significantly higher in-hospital mortality rate compared with those who do not develop the complication (5.69 versus 0.04 percent). (See 'Mortality' below.)
●In a retrospective review of the American College of Surgeons' National Surgical Quality Improvement Program that included 32,889 bariatric surgical patients, the incidence of pneumonia and ARF was 0.6 percent each [164]. While the incidence is relatively low, patients with pulmonary complications were more likely to have additional in-hospital complications (eg, surgical site infections, PE, myocardial infarction), and length of stay was significantly longer (six versus three days, median length of stay). The 30 day mortality rate for patients with pneumonia and ARF was significantly higher compared with patients without the complication (4.3 versus 0.16 percent and 13.7 versus 0.10 percent, respectively). (See 'Mortality' below.)
Specific risk factors for pneumonia included new or preoperative medical history of heart failure, COPD, bleeding disorder, tobacco smoking history, increasing age, open versus laparoscopic operation, and RYGB and BPD-DS when compared with AGB.
Specific risk factors for ARF included COPD, percutaneous coronary intervention, diabetes, increasing age, prolonged anesthetic time, open versus laparoscopic operation, and RYGB and BPD-DS when compared with AGB.
●Similarly, risk factors for pulmonary disease based upon the findings from the National Inpatient Sample database that included 304,515 patients included open procedures and comorbid factors that included heart failure, chronic renal failure, gastric bypass, peripheral vascular disease, male sex, age older than 50 years, alcohol drinking, smoking, chronic lung disease, and diabetes [163]. Patients undergoing a revision bariatric procedure also have a high risk of morbidity.
●Early ambulation and incentive spirometry after surgery are important for decreasing the incidence of pulmonary complications. Preoperative identification of the presence of significant obstructive sleep apnea and initiation of continuous positive airway pressure (CPAP) therapy also reduce the risk of pulmonary complications in the postoperative period [165]. CPAP should be used in the early postoperative period if it is clinically necessary without concern for causing a leak. (See "Overview of the management of postoperative pulmonary complications" and "Clinical presentation and diagnosis of obstructive sleep apnea in adults".)
MORTALITY — Mortality rates are higher for patients undergoing more complex operative procedures or those who sustain adverse intraoperative events or serious in-hospital complications (such as pulmonary embolism [PE] or acute respiratory failure [ARF]) as well as those with a body mass index (BMI) ≥50 kg/m2, male sex, age ≥65, medical comorbidities, or poor functional status [10,34,44,84,164,166-179]. In addition, the mortality rate was higher for patients undergoing an open (versus laparoscopic) procedure (0.79 versus 0.17 percent) [10] or undergoing bariatric surgery by a low-volume surgeon or at a low-volume hospital.
●Complexity of procedure – Higher mortality rates were identified for patients undergoing the more complex biliopancreatic diversion with duodenal switch (BPD-DS) compared with adjustable gastric banding (AGB; 1.2 versus 0.3 percent) [34]. (See 'Incidence' above.)
●Serious intraoperative or in-hospital complications – Higher mortality rates are also identified for patients with postoperative PE (7.1 percent) [151], pneumonia (4.3 percent) or respiratory failure (13.7 percent) [164], or sepsis (n = 6; 33 percent of all deaths) [168] with respect to those patients without such complications. Additionally, complications of an anastomotic leak also increase the risk of postoperative mortality [164]. If not diagnosed in a timely fashion, the mortality rate from an anastomotic leak can be as high as 15 percent [99].
•PE remains the most common cause of mortality in the perioperative period after bariatric surgery and accounts for approximately 30 to 50 percent of deaths [84,180]. The mortality rate of a PE is dependent upon the severity of presentation and time to diagnosis/treatment and varies widely from 1 to 95 percent [181]. The majority of deaths occur within the first one to two hours after the embolism. (See "Overview of acute pulmonary embolism in adults" and "Treatment, prognosis, and follow-up of acute pulmonary embolism in adults".)
•Cardiac complications, including myocardial infarction and cardiac failure, are a common cause of mortality in the perioperative period [171,182]. An analysis of 13,871 patients from a national registry of bariatric surgery reported that the mortality from cardiovascular events ranged from 12.5 to 17.6 percent [182]. (See "Management of cardiac risk for noncardiac surgery".)
•ARF accounts for approximately 11 percent of perioperative mortality after bariatric operations [96,182]. (See "Overview of the management of postoperative pulmonary complications".)
•Sepsis related to an anastomotic leak, if not diagnosed in a timely fashion, is associated with a mortality rate as high as 15 percent [99]. (See "Sepsis syndromes in adults: Epidemiology, definitions, clinical presentation, diagnosis, and prognosis".)
●Patient-specific factors – Based upon retrospective reviews of United States (US) national databases and other series, the following examples illustrate patient-specific risk factors for mortality:
•Male sex – Data from the National Inpatient Sample database on 304,515 patients showed that male sex (odds ratio 1.7) was associated with greater mortality [174]. In a Medicare study, males had higher rates of early death than females at 30 days (3.7 versus 1.5 percent), 90 days (4.8 versus 2.1 percent), and one year (7.5 versus 3.7 percent) [167]. Similar results have been reported in a large population-based study [176].
•Older age – Higher mortality rates have been reported in patients over the age of 65 years. A study of 16,155 Medicare beneficiaries who underwent bariatric procedures reported 30 day, 90 day, and one-year mortality rates of 2.0, 2.8, and 4.6 percent, respectively [167]. Mortality rates were significantly higher for those aged 65 and older compared with younger patients (4.8 versus 1.7 percent at 30 days, 6.9 versus 2.3 percent at 90 days, and 11.1 versus 3.9 percent at one year). Similar results were reported in a second study of a similar population [175]. This is also confirmed by the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP) database analysis of over 266,000 patients [183].
•Excessive obesity – There are conflicting reports relating excessive (or super) obesity to postoperative mortality. In a retrospective review of 856 patients undergoing bariatric surgery, patients who died (n = 54) were more likely to have a BMI ≥50 kg/m2 compared with patients who survived (30/54 patients [55.6 percent] versus 278/802 patients [34.7 percent], hazard ratio [HR] 1.8, 95% CI 1.01-3.09) [44].
However, in another retrospective review, BMI ≥60 kg/m2 was not a risk factor for postoperative mortality. A review of 3692 patients undergoing a bariatric operation found that 291 patients with a BMI ≥60 kg/m2 had a similar in-hospital mortality rate compared with patients with a BMI <60 kg/m2 (0.34 versus 0.12 percent) [45].
•Comorbid illnesses – A diagnostic cost group (DCG) score is a risk adjustment measure that reflects the level of medical comorbidities and one-year mortality rates and predicts health expenditures in US veterans [177-179]. In the retrospective study previously described [44], DCG score ≥2 (higher-than-average expected expenditures) was significantly associated with an increased risk of death after bariatric surgery compared with a DCG score <2 (HR 3.4, 95% CI 1.8-6.5). Metabolic syndrome is associated with three times the mortality compared with patients who do not have metabolic syndrome [184,185]. Cirrhosis is also associated with an increased risk of mortality. Based upon the US Nationwide Inpatient Sample, patients without cirrhosis had significantly lower mortality rates compared with patients with compensated and decompensated cirrhosis (0.3 versus 0.9 versus 16.3 percent) [186].
•Poor functional status – Functional status is an underappreciated comorbidity of obesity as well as a predictor of postoperative complications. Functional status can be described as independent, partially dependent, and completely dependent. A totally dependent person needs assistance for all activities of daily living. A retrospective review of 44,408 bariatric surgery patients from the American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP) database from 2007 to 2009 found the overall mortality rate was 0.14 percent [187]. For patients with partially dependent preoperative function, the adjusted odds ratio (AOR) for 30 day mortality was increased 2.6-fold, while totally dependent patients had a very high risk with an increased AOR of 27.6.
●Surgeon-/hospital-specific factors – The following examples illustrate surgeon- or hospital-specific factors that contribute to a higher mortality:
•Low-volume surgeons and hospitals – Increased mortality is associated with low surgeon and hospital volume of bariatric procedures. Both in-hospital and 30 day mortality are decreased when bariatric surgery is performed by surgeons and hospitals that perform more than 100 procedures a year [166,176]. For these reasons, volume is one of the criteria utilized by the American Society of Metabolic and Bariatric Surgery and the ACS for the accreditation of bariatric center of excellence (COE) programs. Currently, a minimum of 50 stapling cases per year is required to be certified as a comprehensive bariatric COE by the governing bodies. (See "Hospital accreditation, accommodations, and staffing for care of the bariatric surgical patient", section on 'Surgeons and case volume'.)
•Surgical approach – Improvements in perioperative mortality have been documented with the introduction of laparoscopic and robot-assisted techniques [10,188-190]. Most series report a lower mortality rate with a laparoscopic versus an open approach, although mortality rates for both are less than 1 percent [10]. The increased mortality with the open approach may be related to technical challenges that made the patient ineligible for the laparoscopic approach, adverse intraoperative events, and/or laparoscopic expertise of the surgeon [37]. There is no difference between a laparoscopic and a robot-assisted laparoscopic approach in terms of morbidity and mortality.
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Bariatric surgery".)
SUMMARY AND RECOMMENDATIONS
●Incidence – When performed by experienced surgeons at established bariatric centers, the risk of serious intraoperative and early (≤30 days) postoperative complications and mortality rates are relatively low (see 'Incidence' above):
•Intraoperative complication rate – 0.69 to 5 percent
•Early postoperative complication rate – 0.2 to 10 percent
•Perioperative mortality rate – 0.08 percent (95% CI 0.06-0.10)
•Early readmission rate – 2.75 percent
•Early reoperation rate – 0.7 to 7.6 percent
●Risk factors and prevention – Higher postoperative complication rates are expected with more complex procedures, an open approach, and the occurrence of adverse intraoperative events. Certain steps can be taken pre-, intra-, or postoperatively to mitigate the higher risks. (See 'Risk factors and prevention' above.)
●Intraoperative complications – The most common intraoperative complications include anesthetic events, access-related injuries, liver or splenic injuries, vascular injuries, bowel ischemia, and anastomotic misconstruction. (See 'Intraoperative complications' above.)
●Early postoperative surgical complications – Early (≤30 day) surgical complications include gastrointestinal leaks, hemorrhage, surgical site infections, marginal ulcers, anastomotic or sleeve stenosis, small bowel obstruction, oral intake intolerance, acute stoma obstruction after adjustable gastric banding (AGB), and portomesenteric vein thrombosis. (See 'Postoperative surgical complications' above.)
●Early postoperative medical complications – Early (≤30 day) medical complications include deep vein thrombosis, pulmonary embolism, myocardial infarction, and pulmonary complications. (See 'Postoperative medical complications' above.)
●Mortality – Higher mortality rates are expected with (see 'Mortality' above):
•More complex operative procedures (eg, biliopancreatic diversion with duodenal switch)
•Adverse intraoperative events or serious in-hospital complications (eg, pulmonary embolism, myocardial infarction, acute respiratory failure, or sepsis from a leak)
•Patient-specific risk factors such as body mass index ≥50 kg/m2, male sex, age ≥65, medical comorbidities, or poor functional status
•Use of the open (as opposed to minimally invasive) approach
•Low-volume surgeon or low-volume hospital