TE: This policy is used to determine coverage of bariatric surgery for members whose plan documents indclude this benefit. Please review the individual plan document to determine if coverage is available.

Description:
Bariatric surgery is a treatment for morbid obesity in patients who fail to lose weight with conservative measures. There are numerous gastric and intestinal surgical techniques available. While these techniques have heterogeneous mechanisms of action, the result is a smaller gastric pouch that leads to restricted eating. However, these surgeries may lead to malabsorption of nutrients or eventually to metabolic changes.

Adults With Morbid Obesity
For individuals who are adults with morbid obesity who receive gastric bypass, the evidence includes randomized controlled trials (RCTs), observational studies, and systematic reviews. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. TEC Assessments and other systematic reviews of RCTs and observational studies found that gastric bypass improves health outcomes, including weight loss and remission of Type 2 diabetes. A TEC Assessment found similar weight loss with open and laparoscopic gastric bypass. The evidence is sufficient to determine that the technology results in a meaningful improvement in the net health outcome.

For individuals who are adults with morbid obesity who receive laparoscopic adjustable gastric banding (LAGB), the evidence includes RCTs, observational studies, and systematic reviews. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. Systematic reviews of RCTs and observational studies have found that LAGB is a reasonable alternative to gastric bypass. There is less weight loss with LAGB than with gastric bypass, but LAGB is less invasive and is associated with fewer serious adverse events. 

For individuals who are adults with morbid obesity who receive sleeve gastrectomy (SG), the evidence includes RCTs, observational studies (evaluating SG alone and comparing SG with gastric bypass), as well as systematic reviews. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. Systematic reviews of RCTs and observational studies have found that SG results in substantial weight loss and that this weight loss is durable for at least 5 years. A meta-analysis found that short-term weight loss was similar after SG compared with gastric bypass. Long-term weight loss was greater after gastric bypass, but SG is associated with fewer adverse events. The evidence is sufficient to determine that the technology results in a meaningful improvement in the net health outcome.

For individuals who are adults with morbid obesity who receive biliopancreatic diversion (BPD) with duodenal switch, the evidence includes nonrandomized comparative studies, observational studies, and a systematic review. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. Nonrandomized comparative studies have found significantly higher weight loss after BPD with duodenal switch compared with gastric bypass at 1 year. A large case series found sustained weight loss after 7 years. The evidence is sufficient to determine that the technology results in a meaningful improvement in the net health outcome.

For individuals who are adults with morbid obesity who receive BPD without duodenal switch, the evidence includes observational studies and systematic reviews. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. A TEC Assessment reviewed the available observational studies and concluded that weight loss was similar after BPD without a duodenal switch or gastric bypass. However, concerns have been raised about complications associated with BPD without duodenal switch, especially long-term nutritional and vitamin deficiencies. The evidence is insufficient to determine the effects of the technology on health outcomes.

For individuals who are adults with morbid obesity who receive vertical-banded gastroplasty, the evidence includes observational studies and systematic reviews. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. A TEC Assessment identified 8 nonrandomized comparative studies evaluating vertical-banded gastroplasty, and these studies found that weight loss was significantly greater with open gastric bypass. Moreover, vertical-banded gastroplasty has relatively high rates of complications, revisions, and reoperations. The evidence is insufficient to determine the effects of the technology on health outcomes.

For individuals who are adults with morbid obesity who receive 2-stage bariatric surgery procedures, the evidence includes a small RCT and observational studies. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. There is a lack of evidence that 2-stage bariatric procedures improve outcomes compared with 1-stage procedures. The small RCT compared intragastric balloon (IGB) plus gastric bypass with the standard of care plus gastric bypass and did not detect a difference in weight loss at 6 months postsurgery. Case series have shown relatively high complication rates in 2-stage procedures, and patients are at risk of complications in both stages. The evidence is insufficient to determine the effects of the technology on health outcomes.

For individuals who are adults with morbid obesity who receive laparoscopic gastric plication, the evidence includes 2 RCTs, observational studies, and systematic reviews. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. A 2014 systematic review only identified a small nonrandomized comparative study comparing laparoscopic gastric plication with other bariatric surgery procedures. Since the systematic review, 2 RCTs have been published, one comparing laparoscopic gastric plication with a sham procedure and another comparing laparoscopic gastric plication with SG. Laparoscopic gastric plication was more effective than sham at 1-year follow-up and equally effective as SG at 2-year follow-up. Additional comparative studies and RCTs with longer follow-up are needed to permit conclusions about the safety and efficacy of laparoscopic gastric plication. The evidence is insufficient to determine the effects of the technology on health outcomes.

For individuals who are adults with morbid obesity who receive single anastomosis duodenoileal bypass with SG (SADI-S), the evidence includes observational studies. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. No controlled trials have evaluated SADI-S. There are a few case series, the largest of which had fewer than 100 patients. A retrospective chart review of patients receiving gastric bypass, BPD, and SADI-S, reported that among patients without diabetes, SADI-S was more effective in weight loss and cholesterol outcomes than gastric bypass. Among patients with diabetes, SADI-S and BDP had higher remission rates than gastric bypass. Comparative studies and especially RCTs are needed to permit conclusions about the safety and efficacy of SADI-S. The evidence is insufficient to determine the effects of the technology on health outcomes.

For individuals who are adults with morbid obesity who receive duodenojejunal sleeve, the evidence includes RCTs and systematic reviews. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. A systematic review of duodenojejunal sleeves included 5 RCTs and found significantly greater short-term weight loss (12 – 24 weeks) with the sleeves compared with medical therapy. There was no significant difference in symptoms associated with diabetes. All RCTs were small and judged by systematic reviewers to be at high-risk of bias. High-quality comparative studies are needed to permit conclusions on the safety and efficacy of the procedure. The evidence is insufficient to determine the effects of the technology on health outcomes.

For individuals who are adults with morbid obesity who receive IGB devices, the evidence includes RCTs, systematic reviews, and case series. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. RCTs assessing the 2 IGB devices approved by the Food and Drug Administration have found significantly greater weight loss with IGB than with sham treatment or lifestyle therapy alone after 6 months (maximum length of device use). Some adverse events were reported, mainly related to accommodation of the balloon in the stomach; in a minority of cases, these adverse events were severe. One RCT followed patients for an additional 6 months after IGB removal and found sustained weight loss. There are limited data on the durability of weight loss in the long-term. Comparative data are lacking. A large case series found that patients gradually regained weight over time. Moreover, it is unclear how 6 months of IGB use would fit into a long-term weight loss and maintenance intervention. The evidence is insufficient to determine the effects of the technology on health outcomes.

For individuals who are adults with morbid obesity who receive an aspiration therapy device, the evidence includes an RCT and case series. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. The RCT found significantly greater weight loss with aspiration therapy than lifestyle therapy at 1 year. One small case series reported on 15 patients at 2 years. The total amount of data on aspiration therapy remains limited and additional studies are needed before conclusions can be drawn about the effects of treatment on weight loss, metabolism and nutrition and long-term durability of treatment. The evidence is insufficient to determine the effects of the technology on health outcomes.

Revision Bariatric Surgery
For individuals who are adults with morbid obesity and failed bariatric surgery who receive revision bariatric surgery, the evidence includes case series and registry data. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. Case series have shown that patients receiving revision bariatric surgery experienced satisfactory weight loss. Data from a multinational bariatric surgery database has found that corrective procedures following primary bariatric surgery are relatively uncommon but generally safe and efficacious. The evidence is sufficient to determine that the technology results in a meaningful improvement in the net health outcome.

Adults With Type 2 Diabetes
For individuals who are diabetic and not morbidly obese who receive gastric bypass, SG, BPD or LAGB, the evidence includes RCTs, nonrandomized comparative studies, and case series. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. Systematic reviews of RCTs and observational studies have found that certain types of bariatric surgery are more efficacious than medical therapy as a treatment for Type 2 diabetes in obese patients, including those with a body mass index (BMI) between 30 and 34.9 kg/m². The greatest amount of evidence is on gastric bypass. Systematic reviews have found significantly greater remission rates of diabetes, decrease in hemoglobin A1c levels, and decrease in BMI with bariatric surgery than with nonsurgical treatment. The efficacy of surgery is balanced against the short-term risks of the surgical procedure. Most RCTs in this population have 1 to 3 years of follow-up; 1 RCT that included patients with BMI between 30 and 34.9 kg/m² had 5-year follow-up data. The evidence is sufficient to determine that the technology results in a meaningful improvement in the net health outcome.

There are clinical concerns about durability and long-term outcomes at 5 to 10 years as well as potential variation in observed outcomes in community practice vs. clinical trials. As a result, bariatric surgery for individuals who are diabetic and not morbidly obese is considered not medically necessary.

Nondiabetic and Nonobese Adults
For individuals who are not diabetic and not morbidly obese who receive any bariatric surgery procedure, the evidence includes RCTs, nonrandomized comparative studies, and case series. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. There is limited evidence for bariatric surgery in patients who are not diabetic or morbidly obese. A few small RCTs and case series have reported a loss of weight and improvements in comorbidities for this population. However, the evidence does not permit conclusions on the long-term risk-benefit ratio of bariatric surgery in this population. The evidence is insufficient to determine the effects of the technology on health outcomes.

Adolescent Children With Morbid Obesity
Gastric Bypass, LAGB or SG
For individuals who are adolescent children with morbid obesity who receive gastric bypass or LAGB or SG, the evidence includes RCTs, observational studies, and systematic reviews. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. Systematic reviews of studies on bariatric surgery in adolescents, who mainly received gastric bypass or LAGB or SG, found significant weight loss and reductions in comorbidity outcomes with bariatric surgery. For bariatric surgery in the adolescent population, although data are limited on some procedures, studies have generally reported that weight loss and reduction in risk factors for adolescents are similar to that for adults. Most experts and clinical practice guidelines have recommended that bariatric surgery in adolescents be reserved for individuals with severe comorbidities, or for individuals with a BMI greater than 50 kg/m². Also, greater consideration should be placed on the patient developmental stage, on the psychosocial aspects of obesity and surgery, and on ensuring that the patient can provide fully informed consent. The evidence is sufficient to determine that the technology results in a meaningful improvement in the net health outcome.

Bariatric Surgery Other Than Gastric Bypass, LAGB or SG
For individuals who are adolescent children with morbid obesity who receive bariatric surgery other than gastric bypass or LAGB or SG, the evidence includes systematic reviews and a cohort study. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. Studies using bariatric surgery other than gastric bypass, LAGB or SG, have small sample sizes. Results from a meta-analysis including patients using other procedures have shown significant improvements in BMI reduction, fasting blood insulin, and total cholesterol, although the estimates have wide confidence intervals, limiting interpretation. The evidence is insufficient to determine the effects of the technology on health outcomes.

Preadolescent Children With Morbid Obesity
For individuals who are preadolescent children with morbid obesity who receive bariatric surgery, the evidence includes no studies focused on this population. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. Several studies of bariatric surgery in adolescents have also included children younger than 12 years old, but findings were not reported separately for preadolescent children. Moreover, clinical practice guidelines have recommended against bariatric surgery for preadolescent children. The evidence is insufficient to determine the effects of the technology on health outcomes.

Hiatal Hernia Repair With Bariatric Surgery
For individuals with morbid obesity and a preoperative diagnosis of a hiatal hernia who receive hiatal hernia repair with bariatric surgery, the evidence includes cohort studies and case series. Relevant outcomes are overall survival, change in disease status, functional outcomes, health status measures, quality of life, and treatment-related mortality and morbidity. Results from the cohort studies and case series have shown that, when a preoperative diagnosis of a hiatal hernia has been present, repairing the hiatal hernia during bariatric surgery resulted in fewer complications. However, the results are limited to individuals with a preoperative diagnosis. There was no evidence on the use of hiatal hernia repair when the hiatal hernia diagnosis is incidental. The evidence is sufficient to determine that the technology results in an improvement in the net health outcome.

BACKGROUND 
BARIATRIC SURGERY
Bariatric surgery is performed to treat morbid (clinically severe) obesity. Morbid obesity is defined as a body mass index (BMI) greater than 40 kg/m² or a BMI greater than 35 kg/m² with associated complications including, but not limited to, diabetes, hypertension, or obstructive sleep apnea. Morbid obesity results in a very high risk for weight-related complications, such as diabetes, hypertension, obstructive sleep apnea, and various types of cancers (for men: colon, rectal, prostate; for women: breast, uterine, ovarian), and a shortened life span. A morbidly obese man at age 20 can expect to live 13 fewer years than his counterpart with a normal BMI, which equates to a 22% reduction in life expectancy. 

The first treatment of morbid obesity is dietary and lifestyle changes. Although this strategy may be effective in some patients, only a few morbidly obese individuals can reduce and control weight through diet and exercise. Most patients find it difficult to comply with these lifestyle modifications on a long-term basis.

When conservative measures fail, some patients may consider surgical approaches. A 1991 National Institutes of Health Consensus Conference defined surgical candidates as “those patients with a BMI of greater than 40 kg/m², or greater than 35 kg/m² in conjunction with severe comorbidities such as cardiopulmonary complications or severe diabetes.”¹ 

Resolution (cure) or improvement of Type 2 diabetes (T2D) after bariatric surgery and observations that glycemic control may improve immediately after surgery, before a significant amount of weight is lost, have promoted interest in a surgical approach to treatment of T2D. The various surgical procedures have different effects, and gastrointestinal rearrangement seems to confer additional antidiabetic benefits independent of weight loss and caloric restriction. The precise mechanisms are not clear, and multiple mechanisms may be involved. Gastrointestinal peptides, e.g., glucagon-like peptide-1 (1GLP-1), glucose-dependent insulinotropic peptide (GIP), and peptide YY (PYY), are secreted in response to contact with unabsorbed nutrients and by vagally mediated parasympathetic neural mechanisms. GLP-1 is secreted by the L cells of the distal ileum in response to ingested nutrients and acts on pancreatic islets to augment glucose-dependent insulin secretion. It also slows gastric emptying, which delays digestion, blunts postprandial glycemia, and acts on the central nervous system to induce satiety and decrease food intake. Other effects may improve insulin sensitivity. GIP acts on pancreatic beta cells to increase insulin secretion through the same mechanisms as GLP-1, although it is less potent. PYY is also secreted by the L cells of the distal intestine and increases satiety and delays gastric emptying.

Types of Bariatric Surgery Procedures
The following summarizes the most common types of bariatric surgery procedures.

Open Gastric Bypass
The original gastric bypass surgeries were based on the observation that post-gastrectomy patients tended to lose weight. The current procedure (CPT code 43846) involves both a restrictive and a malabsorptive component, with horizontal or vertical partition of the stomach performed in association with a Roux-en-Y procedure (i.e., a gastrojejunal anastomosis). Thus, the flow of food bypasses the duodenum and proximal small bowel. The procedure may also be associated with an unpleasant “dumping syndrome,” in which a large osmotic load delivered directly to the jejunum from the stomach produces abdominal pain and/or vomiting. The dumping syndrome may further reduce intake, particularly in “sweets eaters.” Surgical complications include leakage and operative margin ulceration at the anastomotic site. Because the normal flow of food is disrupted, there are more metabolic complications than with other gastric restrictive procedures, including iron deficiency anemia, vitamin B12 deficiency, and hypocalcemia, all of which can be corrected by oral supplementation. Another concern is the ability to evaluate the “blind” bypassed portion of the stomach. Gastric bypass may be performed with either an open or laparoscopic technique. 

Note: In 2005, the CPT code 43846 was revised to indicate that the short limb must be 150 cm or less, compared with the previous 100 cm. This change reflects the common practice in which the alimentary (i.e., jejunal limb) of a gastric bypass has been lengthened to 150 cm. This length also serves to distinguish a standard gastric bypass with a very long, or very, very long gastric bypass, as discussed further here.

Laparoscopic Gastric Bypass
CPT code 43644 was introduced in 2005 and described the same procedure as open gastric bypass (CPT code 43846), but performed laparoscopically.

Adjustable Gastric Banding
Adjustable gastric banding (CPT code 43770) involves placing a gastric band around the exterior of the stomach. The band is attached to a reservoir implanted subcutaneously in the rectus sheath. Injecting the reservoir with saline will alter the diameter of the gastric band; therefore, the rate-limiting stoma in the stomach can be progressively narrowed to induce greater weight loss, or expanded if complications develop. Because the stomach is not entered, the surgery and any revisions, if necessary, are relatively simple.

Complications include slippage of the external band or band erosion through the gastric wall. Adjustable gastric banding has been widely used in Europe. Two banding devices are approved by the Food and Drug Administration (FDA) for marketing in the United States. The first to receive FDA approval was the LAP-BAND (original applicant, Allergan, BioEnterics, Carpinteria, CA; now Apollo Endosurgery, Austin, TX). The labeled indications for this device are as follows:

"The LAP-BAND^® system is indicated for use in weight reduction for severely obese patients with a body mass index (BMI) of at least 40 or a BMI of at least 35 with one or more severe comorbid conditions, or those who are 100 lb or more over their estimated ideal weight according to the 1983 Metropolitan Life Insurance Tables (use the midpoint for medium frame). It is indicated for use only in severely obese adult patients who have failed more conservative weight-reduction alternatives, such as supervised diet, exercise and behavior modification programs. Patients who elect to have this surgery must make the commitment to accept significant changes in their eating habits for the rest of their lives."

In 2011, FDA-labelled indications for the LAP-BAND were expanded to include patients with a BMI from 30 to 34 kg/m² with at least 1 obesity-related comorbid condition.

The second adjustable gastric banding device approved by FDA through the premarket approval process is the REALIZE^® model (Ethicon Endo-Surgery, Cincinnati, Ohio). Labeled indications for this device are:

“The [REALIZE] device is indicated for weight reduction for morbidly obese patients and is indicated for individuals with a Body Mass Index of at least 40 kg/m², or a BMI of at least 35 kg/m² with one or more comorbid conditions. The Band is indicated for use only in morbidly obese adult patients who have failed more conservative weight-reduction alternatives, such as supervised diet, exercise, and behavior modification programs.”

Sleeve Gastrectomy
A sleeve gastrectomy (CPT code 43775) is an alternative approach to gastrectomy that can be performed on its own or in combination with malabsorptive procedures (most commonly biliopancreatic diversion [BPD] with duodenal switch). In this procedure, the greater curvature of the stomach is resected from the angle of His to the distal antrum, resulting in a stomach remnant shaped like a tube or sleeve. The pyloric sphincter is preserved, resulting in a more physiologic transit of food from the stomach to the duodenum and avoiding the dumping syndrome (overly rapid transport of food through stomach into intestines) seen with distal gastrectomy. This procedure is relatively simple to perform and can be done as an open or laparoscopic procedure. Some surgeons have proposed the sleeve gastrectomy as the first in a 2-stage procedure for very high-risk patients. Weight loss following sleeve gastrectomy may improve a patient’s overall medical status and, thus, reduce the risk of a subsequent more extensive malabsorptive procedure (e.g., BPD).

Biliopancreatic Bypass Diversion
The BPD procedure (also known as the Scopinaro procedure; CPT code 43847) developed and used extensively in Italy, was designed to address drawbacks of the original intestinal bypass procedures that have been abandoned due to unacceptable metabolic complications. Many complications were thought to be related to bacterial overgrowth and toxin production in the blind, bypassed segment. In contrast, BPD consists of a subtotal gastrectomy and diversion of the biliopancreatic juices into the distal ileum by a long Roux-en-Y procedure. The procedure consists of the following components:

1. A distal gastrectomy induces a temporary early satiety and/or the dumping syndrome in the early postoperative period, both of which limit food intake.
2. A 200-cm long “alimentary tract” consists of 200 cm of ileum connecting the stomach to a common distal segment. 
3. A 300- to 400-cm “biliary tract” connects the duodenum, jejunum, and remaining ileum to the common distal segment.
4. A 50- to 100-cm “common tract” is where food from the alimentary tract mixes with biliopancreatic juices from the biliary tract. Food digestion and absorption, particularly of fats and starches, are therefore limited to this small segment of bowel, i.e., creating a selective malabsorption. The length of the common segment will influence the degree of malabsorption. 
5. Because of the high incidence of cholelithiasis associated with the procedure, patients typically undergo an associated cholecystectomy. 

Many potential metabolic complications are related to BPD, including, most prominently, iron deficiency anemia, protein malnutrition, hypocalcemia, and bone demineralization. Protein malnutrition may require treatment with total parenteral nutrition. In addition, several case reports have noted liver failure resulting in death or liver transplant. 

BPD With Duodenal Switch
CPT code 43845, which specifically identifies the duodenal switch procedure, was introduced in 2005. The duodenal switch procedure is a variant of the BPD previously described. In this procedure, instead of performing a distal gastrectomy, a sleeve gastrectomy is performed along the vertical axis of the stomach. This approach preserves the pylorus and initial segment of the duodenum, which is then anastomosed to a segment of the ileum, similar to the BPD, to create the alimentary limb. Preservation of the pyloric sphincter is intended to ameliorate the dumping syndrome and decrease the incidence of ulcers at the duodenoileal anastomosis by providing a more physiologic transfer of stomach contents to the duodenum. The sleeve gastrectomy also decreases the volume of the stomach and decreases the parietal cell mass. However, the basic principle of the procedure is similar to that of the BPD, i.e., producing selective malabsorption by limiting the food digestion and absorption to a short common ileal segment.

Vertical-Banded Gastroplasty
Vertical-banded gastroplasty (VBG; CPT code 43842) was formerly one of the most common gastric restrictive procedures performed in the United States, but has now been replaced by other restrictive procedures due to high rates of revisions and reoperations. In this procedure, the stomach is segmented along its vertical axis. To create a durable reinforced and rate-limiting stoma at the distal end of the pouch, a plug of stomach is removed, and a propylene collar is placed through this hole and then stapled to itself. Because the normal flow of food is preserved, metabolic complications are uncommon. Complications include esophageal reflux, dilation, or obstruction of the stoma, with the latter 2 requiring reoperation. Dilation of the stoma is a common reason for weight regain. VBG may be performed using an open or laparoscopic approach.

Long-Limb Gastric Bypass (i.e., > 150 cm)
Variations of gastric bypass procedures have been described, consisting primarily of long-limb Roux-en-Y procedures (CPT code 43847), which vary in the length of the alimentary and common limbs. For example, the stomach may be divided with a long segment of the jejunum (instead of ileum) anastomosed to the proximal gastric stump, creating the alimentary limb. The remaining pancreaticobiliary limb, consisting of stomach remnant, duodenum, and length of proximal jejunum, is then anastomosed to the ileum, creating a common limb of variable length in which the ingested food mixes with the pancreaticobiliary juices. While the long alimentary limb permits absorption of most nutrients, the short common limb primarily limits absorption of fats. The stomach may be bypassed in a variety of ways (e.g., resection or stapling along the horizontal or vertical axis). Unlike the traditional gastric bypass, which is a gastric restrictive procedure, these very long-limb Roux-en-Y gastric bypasses combine gastric restriction with some element of malabsorptive procedure, depending on the location of the anastomoses. Note that CPT code for gastric bypass (43846) explicitly describes a short limb ( < 150 cm) Roux-en-Y gastroenterostomy, and thus would not apply to long-limb gastric bypass.

Laparoscopic Malabsorptive Procedure
CPT code 43645 was introduced in 2005 to specifically describe a laparoscopic malabsorptive procedure. However, the code does not specifically describe any specific malabsorptive procedure.

Weight Loss Outcomes
There is no uniform standard for reporting results of weight loss or for describing a successful procedure. Common methods of reporting the amount of body weight loss are percent of ideal body weight achieved or percent of excess body weight (EBW) loss, with the latter most commonly reported. These 2 methods are generally preferred over the absolute amount of weight loss, because they reflect the ultimate goal of surgery: to reduce weight into a range that minimizes obesity-related morbidity. Obviously, an increasing degree of obesity will require a greater amount of weight loss to achieve these target goals. There are different definitions of successful outcomes, but a successful procedure is often considered one in which at least 50% of EBW is lost, or when the patient returns to within 30% of ideal body weight. The results may also be expressed as the percentage of patients losing at least 50% of EBW. Table 1 summarizes the variations in reporting weight loss outcomes. 

Table 1. Weight Loss Outcomes

Outcome Measure	Definition	Clinical Significance
Decrease in weight	Absolute difference in weight pre- and post-treatment	Unclear relation to outcomes, especially in morbidly obese
Decrease in BMI	Absolute difference in BMI pre- and post-treatment	May be clinically significant if change in BMI clearly leads to change in risk category
Percent EBW loss	Amount of weight loss divided by EBW	Has anchor to help frame clinical significance; unclear threshold for clinical significance
Percent patients losing > 50% of EBW	No. of patients losing > 50% EBW divided by total patients	Additional advantage of framing on per patient basis. Threshold for significance (> 50%) arbitrary.
Percent ideal body weight	Final weight divided by ideal body weight	Has anchor to help frame clinical significance; unclear threshold for clinical significance

BMI: body mass index; EBW: excess body weight.

Durability of Weight Loss
Weight change (i.e., gain or loss) at yearly intervals is often reported. Weight loss at 1 year is considered the minimum length of time for evaluating these procedures; weight loss at 3 to 5 years is considered an intermediate time period for evaluating weight loss; and weight loss at 5 to 10 years or more is considered to represent long-term weight loss following bariatric surgery.

Short-Term Complications (Operative and Perioperative Complications < 30 Days)
In general, the incidence of operative and perioperative complications is increased in obese patients, particularly in thromboembolism and wound healing. Other perioperative complications include anastomotic leaks, bleeding, bowel obstruction, and cardiopulmonary complications (e.g., pneumonia, myocardial infarction).

Reoperation Rate
Reoperation may be required to either “take down” or revise the original procedure. Reoperation may be particularly common in VBG due to pouch dilation. 

Long-Term Complications (Metabolic Adverse Events, Nutritional Deficiencies)
Metabolic adverse events are of particular concern in malabsorptive procedures. Other long-term complications include anastomotic ulcers, esophagitis, and procedure-specific complications such as band erosion or migration for gastric-banding surgeries.

Improved Health Outcomes in Terms of Weight-Related Comorbidities
Aside from psychosocial concerns, which may be considerable, 1 motivation for bariatric surgery is to decrease the incidence of complications of obesity, such as diabetes, cardiovascular risk factors (i.e., increased cholesterol, hypertension), obstructive sleep apnea, or arthritis. Unfortunately, these final health outcomes are not consistently reported. 

REGULATORY STATUS
Forms of bariatric surgery performed without specific implantable devices are surgical procedures and, as such, are not subject to regulation by the FDA.

Table 2 shows forms of bariatric surgery with implantable devices approved by FDA through the premarket approval process.

Table 2: FDA-Approved Bariatric Surgery Devices

Device	Manufacturer	PMA Date	Labeled Indications
AspireAssist System	Aspire Bariatrics	Jun 2016	For long-term use in conjunction with lifestyle therapy and continuous medical monitoring in obese adults > 22 y, with a BMI of 35.0 to 55.0 kg/m2 and no contraindications to the procedure who have failed to achieve and maintain weight loss with nonsurgical weight loss therapy
ORBERA® intragastric balloon system	Apollo Endosurgery	Aug 2015	For use in obese adults (BMI, 30 – 40 kg/m2 ) who have failed weight reduction with diet and exercise, and have no contraindications. Maximum placement time is 6 mo. Balloon placed endoscopically and inflated with saline.
ReShape® Integrated Dual Balloon System	ReShape Medical	Jul 2015	For use in obese adults (BMI, 30 – 40 kg/m2) and ≥ 1 comorbid conditions who have failed weight reduction with diet and exercise, and have no contraindications. Maximum placement time is 6 mo. Balloon delivered transorally and inflated with saline.
LAP-BAND® Adjustable Gastric Banding System	Apollo Endosurgery (original applicant: Allergan)	Apr 2010	For use in weight reduction for severely obese adults with BMI of at least 40 kg/m2 or a BMI of at least 30 kg/m2 with ≥ 1 severe comorbid conditions who have failed more conservative weight-reduction alternatives (e.g., supervised diet, exercise, behavior modification programs).
REALIZE® Adjustable Gastric Band	Ethicon Endosurgery	Nov 2007	For use in weight reduction for morbidly obese patients and for individuals with BMI of at least 40 kg/m2, or a BMI of at least 35 kg/m2 with ≥ 1 comorbid conditions, or those who are ≥ 45.4 kg over their estimated ideal weight. Indicated for use only in morbidly obese adults who have failed more conservative weight-reduction alternatives (e.g., supervised diet, exercise, behavior modification programs).

BMI: body mass index: FDA: Food and Drug Administration; PMA: premarket approval. 

In February 2017, the FDA issued a letter to health care providers discussing the potential risks with liquid-filled intragastric balloons in response to reports of 2 types of adverse events related to the balloons. Several dozen reports concerned spontaneous overinflation of the balloons, which caused pain, swelling, and vomiting. The second set of adverse event reports indicated that acute pancreatitis developed in several patients due to compression of gastrointestinal structures. These reports involved both ReShape and ORBERA brands. The adverse events may require premature removal of the balloons.

In August 2017, the FDA issued a second letter to health care providers informing them of 5 unanticipated deaths occurring from 2016 through the time of the letter, due to intragastric balloons. The FDA recommended close monitoring of patients receiving these devices.

Related Policies
70173 Gastric Electrical Stimulation

Policy:
BARIATRIC SURGERY IN ADULTS WITH MORBID OBESITY
The following bariatric surgery procedures may be considered MEDICALLY NECESSARY for the treatment of morbid obesity  in adults who have successfully completed a comprehensive nonsurgical weight management program that includes but is not limited to nutrition counseling, counseling for increase in activity, cognitive behavior therapy and possible medication support (see Policy Guidelines section for patient selection criteria). Bariatric surgery should be performed in appropriately selected patients, by surgeons who are adequately trained and experienced in the specific techniques used, and in institutions that support a comprehensive bariatric surgery program, including long-term monitoring and follow-up post-surgery and that are accredited by Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (https://www.facs.org/quality programs/mbsaqip). 

• Gastric bypass using a Roux-en-Y anastomosis, either open or laparoscopic
• Sleeve gastrectomy
• Open or laparoscopic biliopancreatic bypass (i.e., Scopinaro procedure) with duodenal switch based on specific prior authorization review 

NOTE:  BARIATRIC SURGERY PROCEDURES ARE LIMITED TO ONE PER LIFETIME.

The following bariatric surgery procedures are considered investigational/unproven therefore is considered NOT MEDICALLY NECESSARY for the treatment of morbid obesity in adults who have failed weight loss by non-surgical measures: 

• Vertical-banded gastroplasty
• Laparoscopic adjustable gastric banding   
• Gastric bypass using a Billroth II type of anastomosis (mini-gastric bypass) 
• Biliopancreatic bypass without duodenal switch 
• Long-limb gastric bypass procedure (i.e., > 150 cm) 
• Two-stage bariatric surgery procedures (e.g., sleeve gastrectomy as initial procedure followed by biliopancreatic diversion at a later time) 
• Laparoscopic gastric plication 
• Single anastomosis duodeno-ileal bypass with sleeve gastrectomy.

The following endoscopic procedures are investigational/unproven therefore is considered NOT MEDICALLY NECESSARY as a primary bariatric procedure or as a revision procedure (i.e., to treat weight gain after bariatric surgery to remedy large gastric stoma or large gastric pouches):

• Insertion of the StomaphyX™ device
• Endoscopic gastroplasty
• Use of an endoscopically placed duodenojejunal sleeve 
• Intragastric balloons  
• Aspiration therapy device. 

BARIATRIC SURGERY IN PREADOLESCENT CHILDREN
Bariatric surgery is considered investigational/unproven therefore is considered NOT MEDICALLY NECESSARY for the treatment of morbid obesity in preadolescent children. 

BARIATRIC SURGERY IN PATIENTS WITH A BMI LESS THAN 35 KG/M²
Bariatric surgery is considered NOT MEDICALY NECESARY for patients with a BMI less than 35 kg/m². 

REVISION BARIATRIC SURGERY
Revision surgery to address perioperative or late complications of the original bariatric procedure may be considered MEDICALLY NECESSARY. They include, but are not limited to, staple-line failure, obstruction, stricture, non-absorption resulting in hypoglycemia or malnutrition, weight loss of 20% or more below ideal body weight, and band slippage that cannot be corrected with manipulation or adjustment (see Policy Guidelines section). 

Revision of a primary bariatric procedure that has failed due to dilation of the gastric pouch or dilation proximal to an adjustable gastric band (documented by upper gastrointestinal examination or endoscopy) is considered MEDICALLY NECESSARY if the initial procedure was successful in inducing weight loss prior to pouch dilation, and the patient has been compliant with a prescribed nutrition and exercise program.  Converting from one surgical intervention to a second, particularly because the desired weight loss has not been realized, is considered a second procedure and not a revision and is subject to the limitation of one procedure per lifetime.

BARIATRIC SURGERY IN ADOLESCENTS
Bariatric surgery in adolescents may be considered MEDICALLY NECESSARY according to the same weight based criteria used for adults, but greater consideration should be given to psychosocial and informed consent issues (see Policy Guidelines section). In addition, any devices used for bariatric surgery must be used in accordance with the U.S. Food and Drug Administration-approved indications.

CONCOMITANT HIATAL HERNIA REPAIR WITH BARIATRIC SURGERY
Repair of a hiatal hernia at the time of bariatric surgery may be considered MEDICALLY NECESSARY for patients who have a preoperatively diagnosed hiatal hernia with indications for surgical repair (see Policy Guidelines section). 

Repair of a hiatal hernia that is diagnosed at the time of bariatric surgery, or repair of a preoperatively diagnosed hiatal hernia in patients who do not have indications for surgical repair, is considered investigational/unproven therefore is considered NOT MEDICALLY NECESSARY.

Policy Guidelines:
Patient Selection Criteria 
Morbid obesity is defined as a body mass index (BMI) 40 kg/m² or more or a BMI 35 kg/m² or more with at least 1 clinically significant obesity-related disease such as diabetes, obstructive sleep apnea, coronary artery disease, or hypertension for which these complications or diseases are not controlled by best practice medical management.

While there is limited evidence on which to assess the long-term impacts of bariatric surgery for patients younger than age 18 years, severely obese (BMI ≥ 40 kg/m² or 140% of the 95th percentile for age and sex, whichever is lower) adolescents with commonly present though not required comorbidities, or who have a BMI of 35 kg/m² or greater (or 120% of the 95th percentile for age and sex, whichever is lower) with clinically significant disease may be considered for bariatric surgery according to the American Academy of Pediatrics.¹ U.S. Food and Drug Administration (FDA) premarket approval for the LAP-BAND^® System indicates it is intended for severely obese adults. (The clinical study submitted to FDA for approval of the LAP-BAND was restricted to adults ages 18 – 55 years.)

Patients should have documented success base on active participation in comprehensive non-surgical measures for weight reduction prior to consideration of bariatric surgery.  A physician supervised 4 – 6 month comprehensive non-surgical weight management program will include but not be limited to nutritional counseling, counseling for a healthy increase in activity, cognitive behavior therapy and possibly medication support. Special consideration may be given to individuals with Type 2 diabetes regarding program duration and documented weight loss.

Coding
Please see the Codes table for details.

Benefit Application
BlueCard/National Account Issues
State mandates and contractual exclusions may apply to coverage eligibility of bariatric surgery in general.

State or federal mandates (e.g., FEP) may dictate that all devices approved by the U.S. Food and Drug Administration (FDA) (i.e., the Lap-Band device) may not be considered investigational, and thus coverage eligibility of these devices may be assessed only on the basis of their medical necessity.

Rationale
This evidence review was created in July 1996 and has been updated regularly with searches of the PubMed database. The most recent literature update was performed through Dec. 16, 2021.

Evidence reviews assess the clinical evidence to determine whether the use of technology improves the net health outcome. Broadly defined, health outcomes are the length of life, quality of life, and ability to function including benefits and harms. Every clinical condition has specific outcomes that are important to patients and managing the course of that condition. Validated outcome measures are necessary to ascertain whether a condition improves or worsens; and whether the magnitude of that change is clinically significant. The net health outcome is a balance of benefits and harms.

To assess whether the evidence is sufficient to draw conclusions about the net health outcome of technology, 2 domains are examined: the relevance, and quality and credibility. To be relevant, studies must represent 1 or more intended clinical use of the technology in the intended population and compare an effective and appropriate alternative at a comparable intensity. For some conditions, the alternative will be supportive care or surveillance. The quality and credibility of the evidence depend on study design and conduct, minimizing bias and confounding that can generate incorrect findings. The randomized controlled trial (RCT) is preferred to assess efficacy; however, in some circumstances, nonrandomized studies may be adequate. RCTs are rarely large enough or long enough to capture less common adverse events and long-term effects. Other types of studies can be used for these purposes and to assess generalizability to broader clinical populations and settings of clinical practice. The following is a summary of the key literature to date.

Overview: Bariatric Surgery in Adults With Morbid Obesity
There is a vast literature on bariatric surgery for adults with morbid obesity. This literature is characterized by a preponderance of single-arm clinical series from individual institutions. These types of studies can be used to determine the amount of weight loss expected from surgery, the durability of the weight loss, and the rate of adverse events. However, these studies are not adequate for determining the comparative efficacy of bariatric surgery versus conservative treatment, or the comparative efficacy of different bariatric surgery techniques. Some comparative trials, including randomized and nonrandomized designs, compare bariatric surgery with conservative therapy and/or compare outcomes of different bariatric surgery procedures. RCTs of bariatric surgery have been performed but are limited and insufficient to draw conclusions about comparisons of bariatric surgery and conservative treatments for weight loss.¹ RCTs are difficult in bariatric surgery because many experts consider it inappropriate or unethical to randomize patients to bariatric surgery. Also, most patients and clinicians have strong preferences for treatment, which result in a select population that might agree to randomization and, therefore, limited generalizability. As a result, the emphasis for this evidence review is on comparative nonrandomized trials of bariatric surgery and nonsurgical therapy or of different types of bariatric surgery procedures.

Swedish Obese Subjects Trial
The Swedish Obese Subjects (SOS) trial is the most influential study of bariatric surgery versus conservative treatment. The SOS trial started in 1987 with a registry containing a detailed questionnaire and clinical data on obese patients with a body mass index (BMI) greater than 34 kg/m² at 480 primary health care centers in Sweden. From this registry, patients who met eligibility criteria were recruited and offered bariatric surgery. Thus, SOS patients self-selected into treatment, and there were baseline differences between groups, primarily reflecting more excess weight and a higher incidence of comorbidities in the surgery group. A total of 2,010 people chose surgery, and 2,037 people chose conservative care. Each surgical patient was matched on 18 clinical variables with a patient from the registry who received nonsurgical treatment (usual care). Each surgeon chose the surgical procedure offered. Most procedures were vertical-banded gastroplasty (VBG; > 70%), with gastric bypass (6%) and gastric banding (23%) procedures performed as well. Usual care in the SOS trial was the local practice of the primary care center and usually did not include pharmacologic treatment. Patients were followed at regular intervals with repeat questionnaires and physical examinations for at least 10 years.

Many publications from this trial have reported on methods, weight loss, and clinical outcomes.^2,3,4,5 The following general conclusions can be drawn from the SOS study:

• Weight loss was greater with bariatric surgery than with conservative treatment. At 10 years of follow-up, weight loss in the surgery group was 16% of total body weight compared with a weight gain of 1.6% in the conservative treatment group.
• There was significant improvement in glucose control for diabetics and reduced incidence of new cases of diabetes.
• The effect on other cardiovascular risk factors (e.g., hypertension, lipidemia) was also positive, but less marked than that seen for diabetes.
• Mortality was reduced by 29% after a mean follow-up of 10.9 years.
• Quality of life improved in the 2- to 10-year follow-up period, with the degree of improvement in quality of life correlating with the amount of weight loss.

Longitudinal Assessment of Bariatric Surgery Consortium
The Longitudinal Assessment of Bariatric Surgery Consortium study is a large prospective, longitudinal, noncomparative study of patients who underwent Roux-en-Y gastric bypass (RYGB) or laparoscopic adjustable gastric banding (LAGB) with follow-up through 3 years post-procedure.⁶ The study enrolled 2458 subjects, with a median BMI of 45.9 kg/m2 (interquartile range [IQR], 41.7 to 51.5 kg/m²). For their first bariatric surgical procedure, 1,738 participants underwent RYGB, 610 LAGB, and 110 other procedures. At 3-year follow-up, for 1,533 RYGB patients with available data, the percentage of baseline weight lost was 31.5% (IQR, 24.6% to 38.4%). For the 439 LAGB patients with available data at 3 years, the percentage of baseline weight loss was 15.9% (IQR, 7.9% to 23.0%). At 3 years post-surgery, 67.5% and 28.5% of RYGB and LAGB patients, respectively, had at least partial diabetes remission. Dyslipidemia was in remission in 61.9% and 27.1% of RYGB and LAGB patients, respectively. Subsequent bariatric procedures (revision or reversal) were required in 0.3% (95% confidence interval [CI], 0.1% to 0.9%) of the RYGB patients and in 17.5% (95% CI, 13.8% to 21.9%) of LAGB patients

National Patient-Centered Clinical Research Network — Bariatric Study
The National Patient-Centered Clinical Research Network (PCORnet) Bariatric Study is a large retrospective, comparative study of 65,093 patients aged 20 to 79 years who underwent RYGB (n = 32,208), LAGB (n = 29,693), or sleeve gastrectomy (SG)(n = 3,192) with follow-up through 5 years post-procedure.⁷ Mean estimated percent total weight loss (% TWL) was calculated at 1, 3 and 5 years in addition to 30-day rates of major adverse events. Study results are summarized in Table 3. This study demonstrates that RYGB is associated with a greater weight loss than SG (p < .001) and that LAGB is associated with the lowest amount of weight loss as observed in a large and diverse patient cohort.

Table 3. National Patient-Centered Clinical Research Network — Bariatric Study Results

	Mean TWL, % (95% CI)			MAE Rate,% (95% CI)
Group (na)	1 Year	3 Years	5 Years	30 Days
RYGB (19,029; 9225; 3676)	-31.2 (-31.3 to -31.1)	-29.0 (-29.2 to -28.8)	-25.5 (-25.9 to -25.1)	5.0 (NR)
LAGB (1681; 943; 337)	-13.7 (-14.0 to -13.3)	-12.7 (-13.5 to -12.0)	-11.7 (-13.1 to -10.2)	2.9 (NR)
SG (14,929; 5304; 1088)	-25.2 (-25.4 to -25.1)	-21.0 (-21.3 to -20.7)	-18.8 (-19.6 to -18.0)	2.6 (NR)

CI: confidence interval; LAGB: laparoscopic adjustable gastric banding; MAE: major adverse event; NR: not reported; RYGB: Roux-en-Y gastric bypass; SG: sleeve gastrectomy; TWL: total weight loss.
^a Number of patients evaluated at 1, 3 and 5 years, respectively. 

Systematic Reviews
Numerous systematic reviews have compared the efficacy of bariatric surgery with conservative therapy or compared different types of bariatric surgery techniques, some of which are older and/or do not extend across the full range of available studies.^8,9,10,11

Cosentino et al. (2021) performed a network meta-analysis of 43 RCTs comparing the efficacy of bariatric surgery versus medical therapy, as well as comparing different types of bariatric surgery techniques.¹² Most included trials were 1 year in duration but a few extended to 5 years. Results demonstrated that surgery reduced BMI more effectively than medical therapy (mean difference [MD], -6.632 kg/m2; 95% CI, -8.29 to -4.97), but increased risk for severe adverse events (odds ratio [OR], 3.06; 95% CI, 1.09 to 8.57). When comparing different procedures to medical therapy, duodenal switch (DS) and bilio-pancreatic diversion (BPD) appeared to be more effective than other procedures, whereas greater curvature plication, LAGB, and laparoscopic vertical banded gastroplasty produced a smaller weight loss than other interventions. When comparing different types of bariatric surgery techniques on BMI change, RYGB was superior to LAGB (MD, -4.26; 95% CI, -6.02 to -2.50; n = 2 studies) and LVGB (MD, -3.05; 95% CI, -5.88 to -0.21; n = 2 studies); the difference between RYGB and SG (n = 12 studies), BPD (n = 2 studies), gastric plication (n = 3 studies), and one anastomosis/gastric bypass (OAGB; n = 2 studies) did not reach statistical significance. Roux-en-Y gastric bypass was inferior to DS for BMI change (MD, 7.55; 95% CI, 6.35 to 8.75; 2 studies).

Park et al. (2019) conducted a systematic review with a network meta-analysis evaluating the comparative efficacy of various bariatric surgery techniques against standard-of-care in the treatment of morbid obesity and diabetes.¹³ The literature search was conducted through February 2018, identifying 45 RCTs for inclusion on RYGB (2 studies versus control), SG (3 studies versus control), LAGB (5 studies versus control), and biliopancreatic diversion (BPD) with DS (BPD-DS; 3 studies versus RYGB). Based on 33 trials, superior efficacy for percent excess weight loss (EWL) compared to standard-of-care was seen for BPD-DS (MD, 38.2%; 95% CI, 7.3 to 69.1), RYGB (MD 32.1%; 95% CI, 3.1 to 61.1), and SG (MD 32.5%; 95% CI, 5.5 to 59.5) at 6 months post-procedure. LAGB was not superior to standard-of-care (MD -0.2%; -19.6 to 19.2). At 3 years post-procedure, superior efficacy for percent EWL compared to standard-of-care was seen for RYGB (MD 45%; 95% CI, 21.8 to 68.2) and SG (MD 39.2%; 95% CI, 15.2 to 63.3). BPD-DS (relative risk [RR] 7.51; 95% CI, 1.91 to 29.54), RYGB (RR 7.51; 95% CI, 1.98 to 28.46), and SG (RR 6.69; 95% CI, 1.75 to 25.57) were all superior to standard-of-care with respect to remission rates at 3 to 5 years post-procedure and remission rates were not significantly different among procedures. SG was found to have a relatively lower risk of adverse events compared to RYGB.

Kang et al. (2017) conducted a systematic review with a network meta-analysis that compared the 3 most common types of bariatric surgery techniques: RYGB, SG and LAGB.¹⁴ The literature search, conducted through July 2016, identified 11 RCTs for inclusion (8 RYGB versus SG; 2 RYGB versus LAGB; 1 SG versus LAGB). Quality of the trials was assessed using the Jadad score, based on allocation concealment, blinding, intention-to-treat analysis, power calculation, and funding. Most trials had a Jadad score of 3 (scale range, 1 to 5). A meta-analysis for the outcome of BMI reduction (6 trials) showed that there was no difference between SG and RYGB (0.7; 95% CI, -1.6 to 3.1). A meta-analysis of RYGB and LAGB (2 trials) and a single trial of SG and LAGB showed that LAGB was not as effective as RYGB or SG (5.8; 95% CI, 2.3 to 9.1; and 5.1; 95% CI, 0.9 to 8.9; respectively). Meta-analyses for the outcome of percent EWL showed the same pattern, no difference comparing SG and RYGB (5 trials; -4.0; 95% CI, -14.0 to 8.2), and both SG and RYGB more effective than LAGB (2 trials; 22.0; 95% CI, 6.5 to 34.0; 1 trial; 26.0; 95% CI, 6.4 to 41.0; respectively).

Colquitt et al. (2014) updated 2003 and 2009 Cochrane reviews of bariatric surgery for obesity.¹⁵ The authors identified 22 randomized trials that compared bariatric surgery with nonsurgical obesity management or that compared different bariatric surgery procedures (N = 1,798 participants; sample size range, 15 to 250 participants). All 7 RCTs comparing surgery with nonsurgical interventions found benefits of surgery on measures of weight change at 1- to 2-year follow-ups. However, reviewers noted that adverse event rates and reoperation rates were poorly reported across trials, and long-term follow-up (beyond 1 to 2 years) was limited. Gloy et al. (2013) conducted a systematic review and meta-analysis of RCTs comparing current bariatric surgery techniques with nonsurgical treatment for patients with a BMI of 30 kg/m2 or more.¹⁶ Eleven studies (N = 796 patients) were included. Overall, patients after bariatric surgery lost more body weight than patients after nonsurgical treatment (MD, -26 kg; 95% CI, -31 to -21; p < .001). Remission of Type 2 diabetes (T2D) was more likely for bariatric surgery patients than for nonsurgical patients (RR of T2D remission, 22.1; 95% CI, 3.2 to 154.3; p < .000); similarly, remission of metabolic syndrome was more likely for bariatric surgery patients (RR, 2.4; 95% CI, 1.6 to 3.6; p < .001). After bariatric surgery, 21 (8%) of 261 patients required reoperations (5/124 after LAGB, 4/69 after RYGB, 1/49 after SG, 1/19 after BPD). Similar to the Colquitt et al. (2014) meta-analysis, no studies reported longer-term follow-up (> 2 years) and heterogeneity between studies were high. Chang et al. (2014) published a systematic review and meta-analysis of RCTs and observational studies to evaluate the effectiveness and risks of bariatric surgery.¹⁷ Reviewers included 164 studies (37 RCTs, 127 observational studies), with a total of 161,756 patients. Mean pre-surgery BMI was 45.62 kg/m² and, among the studies that provided information about obesity-related comorbidities, 26% of patients had T2D, 47% had hypertension, 28% had dyslipidemia, 7% had cardiovascular disease, and 25% had obstructive sleep apnea (OSA). Perioperative complications were relatively low, with a perioperative mortality rate in RCTs of 0.08% (95% CI, 0.01 to 0.24) and in observational studies of 0.22% (95% CI, 0.14 to 0.31). Complication rates were 17% (95% CI, 11 to 23) for RCTs and 10% for observational studies (95% CI, 7 to 13). At 1-year follow-up, mean change in BMI was -13.53 kg/m2 (95% CI, -15.51 to -11.55) in RCTs and -11.79 kg/m2 (95% CI, -13.89 to -9.69) in observational studies. Decreases in BMI were generally sustained over 2 to 4 years of follow-up among studies reporting this outcome.

Many systematic reviews have reported improvements in specific obesity-related comorbidities following bariatric surgery. These reviews have relied primarily on the results of observational studies and included the outcomes of hypertension, T2D, hyperlipidemia, cardiovascular events, quality of life, cancer, knee pain, and liver disease.^{18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37}

Liu et al. (2021) performed a network meta-analysis of 35 RCTs (N = 2198) to compare the effects of bariatric surgery versus lifestyle/medical interventions on dyslipidemia and insulin resistance in patients who are overweight with or without T2D.³⁸ Compared with lifestyle/medical interventions, the Homeostasis Model Assessment for Insulin Resistance (HOMA-IR; a product of fasting circulating insulin and glucose concentrations divided by 22.5) was significantly lower with RYGB (MD, -3.93; 95% credible interval [CrI], -6.20 to -2.17), single anastomosis (mini-) gastric bypass (SAGB) (MD, -4.45; 95% CrI, -9.04 to -0.34), and SG (MD, -4.32; 95% CrI, -6.74 to -2.22). Compared with lifestyle/medical interventions, a statistically significant difference in the reduction of LDL-C was only reached with RYGB (MD, -0.51; 95% CrI, -0.85 to -0.16) and DS (MD, -0.90; 95% CrI, -1.66 to -0.16).

Wiggins et al. (2020) analyzed large-scale population studies to evaluate the association between bariatric surgery and long-term mortality and the incidence of new-onset obesity-related disease at a national level.³⁹ The analysis included 18 national or regional administrative database cohort studies involving patients who had undergone any bariatric procedure compared to an appropriate control group with a minimum follow-up of 18 months. Overall, 1,539,904 patients were included: 269,818 receiving a bariatric procedure and 1,270,086 controls. Results revealed that bariatric surgery was associated with a significant improvement in all-cause mortality (pooled odds ratio [POR], 0.62; 95% CI, 0.55 to 0.69; p < .001), cardiovascular mortality (POR, 0.5; 95% CI, 0.35 to 0.71; p < .001), T2D incidence (POR, 0.39; 95% CI, 0.18 to 0.83; p = .01), hypertension (POR, 0.36; 95% CI, 0.32 to 0.4; p < .001), dyslipidemia (POR, 0.33; 95% CI, 0.14 to 0.8; p = .01), and ischemic heart disease (POR, 0.46; 95% CI, 0.29 to 0.73; p = .001). Limitations of this analysis included inability to account for unmeasured variables, which may have not been equally distributed between patient groups due to the nonrandomized design of included studies, heterogeneity between studies regarding the nature of the control group utilized, and unexamined potential adverse effects related to bariatric surgery due to a lack of data.

Puzziferri et al. (2014) conducted a systematic review of studies of bariatric surgery reporting follow-up beyond 2 years, which included 29 studies (N = 7,971 patients).⁴⁰ At follow-up, which ranged from 2 to 5 years post-procedure, the mean sample size‒weighted percentage of EWL was higher for gastric bypass (65.7%) than for gastric banding (45.0%). Reviewers noted that few studies reported sufficient long-term results to minimize bias.

Section Summary: Bariatric Surgery in Adults With Morbid Obesity
There is a lack of large-scale RCTs with long-term follow-up comparing bariatric surgery with nonsurgical treatment for the general population of patients with morbid obesity. Evidence from nonrandomized comparative studies and case series and meta-analyses of existing RCTs has consistently reported that bariatric surgery results in substantially greater weight loss than nonsurgical therapy. Data from the largest comparative study (the SOS study) has reported that bariatric surgery is associated with improvements in mortality, diabetes, cardiovascular risk factors, and quality of life.

Evidence for Specific Types of Bariatric Surgery Procedures
Gastric Bypass for Adults With Morbid Obesity
Clinical Context and Therapy Purpose
The purpose of gastric bypass is to provide a treatment option that is an alternative to or an improvement on existing therapies, in patients who are adults with morbid obesity.

The question addressed in this evidence review is: Does gastric bypass improve the net health outcome in adults who are obese?

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest are individuals who are adults with morbid obesity. Morbid obesity is defined as a BMI 40 kg/m² or more or a BMI 35 kg/m² or more with at least 1 clinically significant obesity-related disease such as diabetes, OSA, coronary artery disease, or hypertension for which these complications or diseases are not controlled by best practice medical management.

Interventions
The therapy being considered is gastric bypass. The procedure involves both a restrictive and a malabsorptive component, with the horizontal or vertical partition of the stomach performed in association with a Roux-en-Y procedure (i.e., a gastrojejunal); thus, food bypasses the duodenum and proximal small bowel.

Comparators
Comparators of interest include standard medical care. Treatment for adults with morbid obesity includes physical exercise, low carbohydrate dieting, and low-fat dieting.

Outcomes
The general outcomes of interest are overall survival (OS), change in disease status, functional outcomes, health status measures, quality of life, treatment-related mortality, and treatment-related morbidity.

Negative outcomes can include surgical complications, including leakage and operative margin ulceration at the, and metabolic complications, including iron deficiency anemia, vitamin B12 deficiency, and hypocalcemia.

The existing literature evaluating gastric bypass as a treatment for morbid obesity has varying lengths of follow-up, ranging from 1 to 10 years. While studies described below all reported at least 1 outcome of interest, longer follow-up was necessary to fully observe outcomes. Therefore, 1-year follow-up is necessary to demonstrate weight loss efficacy. Longer follow-up of 5 to 10 years is desirable to assess maintenance of weight loss, impact on co-occurring conditions, and appearance of long-term complications.

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

• To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs.
• In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
• To assess longer-term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
• Consistent with a "best available evidence approach," within each category of study design, studies with larger sample sizes and longer durations were sought.
• Studies with duplicative or overlapping populations were excluded.

Review of Evidence
Systematic Reviews
Cui et al. (2021) published a systematic review of 7 RCTs comparing long-term outcomes of RYGB (n = 239) versus medical therapy (n = 238) in obese patients with T2D.⁴¹ Results demonstrated a higher likelihood of T2D remission with RYGB versus medical therapy at 1 year (RR, 18.01; 95% CI, 4.53 to 71.70), 3 years (RR, 29.58; 95% CI, 5.92 to 147.82), and 5 years (RR, 16.92; 95% CI, 4.15 to 69.00). The probability of achieving American Diabetes Association (ADA) treatment goals was also more likely with RYGB versus medical therapy at 1 year (RR, 3.99; 95% CI, 1.01 to 15.82), 3 years (RR, 3.16; 95% CI, 1.33 to 7.49), and 5 years (RR, 6.18; 95% CI, 1.69 to 22.68).

Yan et al. (2016) published a systematic review of RCTs comparing gastric bypass with medical treatment in obese patients (i.e., BMI ≥ 30 kg/m²) who had T2D.⁴² The primary study outcome was remission of T2D, which was reported in 5 of the 6 studies. A pooled analysis found a significantly higher remission rate after gastric bypass than after medical treatment (OR, 76.37; 95% CI, 20.70 to 271.73; p < .001). Also, a pooled analysis found a significantly lower final BMI in the gastric bypass group than in the medical treatment group (MD, -6.54 kg/m2; 95% CI, -9.28 to -3.80 kg/m2; p < .001).

A 2005 TEC Assessment focused on laparoscopic gastric bypass, which intends to reproduce the open procedure via minimally invasive techniques.⁴³ This technically complex surgery requires a dedicated team and a relatively high degree of skill and experience in laparoscopic technique. This Assessment reviewed 7 comparative trials of the open gastric bypass and laparoscopic gastric bypass, including 3 RCTs. Also, 18 large clinical series of laparoscopic gastric bypass were included. The Assessment concluded that weight loss at 1 year was similar for laparoscopic and open gastric bypass approaches.⁴³ Longer follow-up periods were less well-reported but appeared to be similar for both approaches. While comparisons of complication rates were less certain, some patterns were evident and consistent across the data examined. The profile of adverse events differed between the 2 approaches, with each having advantages and disadvantages. Laparoscopic gastric bypass offered a less invasive procedure associated with decreased hospital stay and earlier return to usual activities. Mortality might be lower with the laparoscopic approach, although both procedures had mortality rates less than 1%. Postoperative wound infections and incisional hernias were also less frequent with laparoscopic gastric bypass. However, anastomotic problems, gastrointestinal tract bleeding, and bowel obstruction appeared to be higher with the laparoscopic approach, though not markedly higher. Given these data, the overall benefit-risk profile for these 2 approaches appeared to be similar.

Observational Studies
Arterburn et al. (2021) published a retrospective, matched cohort study to investigate weight loss among patients with severe obesity undergoing RYGB, SG or nonsurgical treatment.⁴⁴ Among 17,258 RYGB, 13,900 SG, and 87,965 nonsurgical patients, the 5-year follow-up rate was 72.0%, 70.9%, and 64.5%, respectively. At 1, 5 and 10 years, RYGB patients had a %TWL of -28.35% (95% CI, -28.53 to -28.18), -21.74% (95% CI, -22.02 to -21.45), and -20.18% (95% CI, -21.00 to -19.34), respectively; at the same time points, nonsurgical patients had a %TWL of -0.22% (95% CI, -0.35 to -0.09), -2.24% (95% CI, -2.46 to -2.02), and -4.78% (95% CI, -5.51 to -4.04), respectively. At 1 and 5 years, SG patients had a %TWL of -22.98% (95% CI, -23.19 to -22.76) and -15.99% (95% CI, -16.58 to -15.40), respectively.

Wadden et al. (2019) reported on end-of-trial results from the Look AHEAD: Action for Health in Diabetes (Look AHEAD) trial, which evaluated outcomes in patients with T2D and obesity who had self-selected to receive bariatric surgery after failing an assigned intensive lifestyle intervention (ILI) or a diabetes support and education (DSE) control therapy.⁴⁵ Patients who received bariatric surgery were significantly more likely to be female (p < .001), younger (p < .001), and have higher BMI at randomization (p < .001). Patients underwent 127 RYGB, 58 LAGB, and 11 SG procedures, respectively. End-of-trial assessments were completed at 4.3 years post-surgery compared to 9.6 years post-randomization for the DSE and ILI participants. Patients undergoing RYGB, LAGB, or SG surgical procedures lost a mean of 22.4% ± 1.0%, 13.0% ± 1.5%, and 16.2% ± 3.3% of baseline weight, respectively. Twelve patients (6.1%) receiving bariatric surgery were randomized with a BMI < 35 kg/m2. The mean BMI was 37.0 ± 5.1, 37.1 ± 5.3, and 42.1 ± 5.8 for DSE, ILI, and surgery groups, respectively (p < .001). Overall, surgically-treated patients lost a mean of 19.3% of baseline weight, compared with 5.8% and 3.3% for the ILI and DSE participants. Full diabetes remission was achieved by 7.6% of bariatric surgery participants compared to 1.1% of ILI and 1.1% of DSE participants. Full remission was significantly more common in surgically treated participants in ILI (RR 6.72; 95% CI, 3.35 to 13.48; p < .001) or DSE (RR 7.07; 95% CI, 3.49 to 14.30; p < .001) groups. Significantly greater reductions in waist circumference (p < .001), triglyceride levels (ILI: p = .03; DSE: p = .02), and hemoglobin A1c (HbA1c) levels (p < .001) were observed in surgically-treated patients compared to ILI or DSE groups. The study was limited by heterogeneity in baseline characteristics and choice of surgical procedure. Results were not stratified by surgery type or BMI range.

Section Summary: Gastric Bypass for Adults With Morbid Obesity
Gastric bypass has been extensively studied. TEC Assessments and other systematic reviews have found that gastric bypass improved health outcomes, including weight loss and remission of T2D. A TEC Assessment also found similar weight loss with open and laparoscopic gastric bypass.

Laparoscopic Adjustable Gastric Banding for Adults with Morbid Obesity
Clinical Context and Therapy Purpose
The purpose of LAGB is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as standard medical care, in patients who are adults with morbid obesity.

The question addressed in this evidence review is: Does the LAGB procedure improve the net health outcome in adults who are obese?

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals who are adults with morbid obesity.

Interventions
The therapy being considered is LAGB.

Comparators
Comparators of interest include standard medical care. Treatment for adults with morbid obesity includes physical exercise, low carbohydrate dieting, and low-fat dieting.

Outcomes
The general outcomes of interest are OS, change in disease status, functional outcomes, health status measures, quality of life, treatment-related mortality, and treatment-related morbidity.

The existing literature evaluating LAGB as a treatment for morbid obesity has varying lengths of follow-up, ranging from 1 to 2 years. While studies described below all reported at least 1 outcome of interest, longer follow-up was necessary to fully observe outcomes. Therefore, 1-year follow-up is necessary to demonstrate weight loss efficacy. Longer follow-up of 5 to 10 years is desirable to assess maintenance of weight loss, impact on co-occurring conditions, and appearance of long-term complications.

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

• To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs;
• In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
• To assess longer-term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
• Consistent with a 'best available evidence approach,' within each category of study design, studies with larger sample sizes and longer durations were sought.
• Studies with duplicative or overlapping populations were excluded.

Review of Evidence
Systematic Reviews
A 2006 TEC Assessment updated the evidence on LAGB and compared outcomes with gastric bypass.46, This Assessment concluded that, for patients considering bariatric surgery, there was sufficient evidence to permit an informed choice between gastric bypass and LAGB. An informed patient might reasonably choose open gastric bypass or LAGB as the preferred procedure. Preoperative counseling should include education on the comparative risks and benefits (eg, extent of weight loss and frequency and timing of potential complications) of the 2 procedures to optimize choice based on preferences and shared decision making.

Weight loss outcomes from the studies reviewed in the Assessment confirmed that weight loss at 1 year was lower for LAGB than for open gastric bypass. The percentage of EWL at 1 year was approximately 40%, compared with 60% or higher for open gastric bypass. At time points beyond 1 year, some comparative studies have reported that the difference in weight loss between LAGB and open gastric bypass narrows, but other studies did not. Weight loss outcomes from the 9 single-arm series with the most complete follow-up did not support the hypothesis that the difference in weight loss shrinks after 1 to 2 years of follow-up. It appears more likely from the current data that attrition bias might have accounted for the diminution of the difference in weight loss over time, particularly when patients with bands removed or deflated were excluded from analysis.

These studies also confirmed that short-term (perioperative) complications are very low with LAGB and lower than with open gastric bypass or LAGB. Death was extremely rare, and serious perioperative complications probably occurred at rates less than 1%. The reported rates of long-term adverse events vary considerably. In the comparative trials, reoperations were reported in approximately 25% of patients, while, in the single-arm studies, the composite rate for reoperations were approximately 50% lower (11.9%). The rates of other long-term complications were also highly variable; eg, the range of rates for band slippage was 1% to 36%, and the range for port access problems was 2% to 20%. These data on long-term complications remain suboptimal. The reporting of long-term complications in these trials was not systematic or consistent. While impossible to determine the precise rates of long-term complications from these data, it is likely that complications have been underreported in many studies due to incomplete follow-up and lack of systematic surveillance. A publication by Ibrahim et al. (2017) reviewed 25,042 Medicare beneficiaries who underwent LAGB surgery; 18.5% (n = 4636) patients underwent 1 or more reoperation(s). Reoperation was prompted by the need for band removal (41.8%), band and port replacement (28.6%), and other requirements.47, The rates of long-term complications reported in some studies raise concern about the impact of these events on the overall benefit-risk profile for LAGB.

In comparing LAGB with open gastric bypass, there are tradeoffs in terms of risks and benefits. LAGB is a less invasive procedure associated with fewer procedural complications, decreased hospital stay and earlier return to usual activities. However, benefits defined by the amount of weight lost are lower for LAGB. The patterns of long-term complications also differ between the 2 procedures. For LAGB, longer-term adverse events related to the presence of a foreign body in the abdomen will occur and result in reoperations and removal of the band in a minority of patients. Patients who have their bands removed can later be offered an alternative bariatric surgery procedure, such as gastric bypass.

A systematic review by Chakravarty et al. (2012)48, comparing LAGB with other bariatric surgery procedures drew conclusions similar to the TEC Assessment. Reviewers included 5 RCTs. The RCTs found that patients using LAGB lost weight, but less weight than with other procedures (e.g., gastric bypass or SG). However, the short-term complication rate was lower with LAGB, and no difference was found in quality of life after LAGB versus other procedures.

Prospective Studies
Dixon et al. (2018) published a prospective, industry-sponsored study of morbidly obese patients who underwent implantation of the adjustable gastric banding system (LAP-BAND)49,. Between 2009 and 2013, 652 patients with a mean BMI of 45.4 kg/m2 were treated at 17 participating centers in the United States and Canada. At 5 years, the explant rate was 8.74% (95% CI, 6.6 to 10.9). Excluding explants, 100 (15.3%) reoperations were necessary during the follow-up period. A mean weight loss of 18.7% was achieved by 2 years and maintained through 5-year follow-up. The study was limited by the lack of control group.

Section Summary: Laparoscopic Adjustable Gastric Banding for Adults With Morbid Obesity
Systematic reviews of the literature have concluded that there is less weight loss with LAGB than gastric bypass.

Sleeve Gastrectomy for Adults With Morbid Obesity
Clinical Context and Therapy Purpose
The purpose of SG is to provide a treatment option that is an alternative to or an improvement of existing therapies, such as standard medical care, in patients who are adults with morbid obesity.

The question addressed in this evidence review is: Does SG improve the net health outcome in adults who are obese?

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals who are adults with morbid obesity.

Interventions
The therapy being considered is SG, an alternative approach to gastrectomy that can be performed on its own or in combination with malabsorptive procedures. In this procedure, the greater curvature of the stomach is resected from the angle of His to the distal antrum, resulting in a stomach remnant shaped like a tube or sleeve. This procedure can be done as an open or laparoscopic procedure.

Comparators
Comparators of interest include standard medical care. Treatment for adults with morbid obesity includes physical exercise, low carbohydrate dieting, and low-fat dieting.

Outcomes
The general outcomes of interest are OS, change in disease status, functional outcomes, health status measures, quality of life, treatment-related mortality, and treatment-related morbidity.

The existing literature evaluating SG as a treatment for morbid obesity has varying lengths of follow-up, ranging from 1 to 5 years. While studies described below all reported at least 1 outcome of interest, longer follow-up was necessary to fully observe outcomes. One-year follow-up is necessary to demonstrate weight loss efficacy. Longer follow-up of 5 to 10 years is desirable to assess maintenance of weight loss, impact on co-occurring conditions, and appearance of long-term complications.

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

• To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs.
• In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
• To assess longer-term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
• Consistent with a "best available evidence approach," within each category of study design, studies with larger sample sizes and longer durations were sought.
• Studies with duplicative or overlapping populations were excluded.

Review of Evidence
Systematic Reviews
Sleeve gastrectomy may be performed as a stand-alone procedure or in combination with a malabsorptive procedure, such as the BPD with BPD-DS. It has also been proposed as the first step in a 2-stage procedure, with gastric bypass or BPD as the second stage.

Numerous recent systematic reviews have compared SG and RYGB with regard to effects on weight, comorbidities, and complications.^{50,51,52,53,54,55}

Lee et al. (2021) performed a meta-analysis evaluating long-term (5 years) outcomes of laparoscopic RYGB versus SG (Table 4).⁵⁶ A total of 33 studies (N = 2,475) were included. Results demonstrated that RYGB resulted in a significantly greater decrease of BMI compared to SG at 1 and 3 years post-surgery; results at 5 years did not reach statistical significance (Table 5). A similar trend was seen for the resolution of dyslipidemia. Furthermore, neither RYGB nor SG was superior for the remission of T2D and hypertension at 5 years.

Gu et al. (2020) completed a meta-analysis of the medium- and long-term effects of laparoscopic SG and RYGB (Table 4).⁵⁰ The evaluation included 9,038 patients from 28 studies. Overall, 5 year follow-up results revealed that laparoscopic RYGB was associated with an improvement in percentage of EWL and remission of T2D, hypertension, and dyslipidemia as compared to laparoscopic SG. Han et al. (2020) also published a systematic review and meta-analysis involving 18 studies (N = 2,917) that compared weight loss and comorbidity resolution between laparoscopic SG and RYGB (Table 4).⁵¹ Results from this analysis revealed no significant difference in EWL or T2D resolution between the 2 procedures. Laparoscopic RYGB was found to be superior to SG with regard to dyslipidemia, hypertension, and GERD management; however, patients who underwent laparoscopic SG experienced fewer postoperative complications and reoperation rates.

Sharples et al. (2020) performed a systematic review and meta-analysis evaluating long-term (5 years) outcomes of RYGB and SG (Table 4).52, Overall, both RYGB and SG resulted in sustained weight loss and comorbidity control with RYGB associated with a greater percent EWL, improved dyslipidemia outcomes, and a reduced incidence of GERD (Table 5).

Shenoy et al. (2020) published a systematic review and meta-analysis of 9 studies that compared laparoscopic SG and RYGB in 2240 elderly (> 55 years) patients.⁵³ Results revealed no significant differences between the 2 bariatric procedures with regard to the rate of early complications (3.6% LSG versus 5.8% LRYGB; p = .15) and mortality (0.1% versus 0.8%; p = .27). Additionally, there was no difference in EWL between the procedures at 1 year (Table 5); however, the authors recommended SG for high-risk elderly patients due to the reduced mortality and complication rates with this procedure. Another systematic review and meta-analysis by Xu et al. (2020) involving 19 studies also concluded that SG was the preferable option for elder obese patients 60 years and older as it was found to be non inferior to RYGB with regard to efficacy, but overall had an improved safety profile.⁵⁷

Osland et al. (2017) published a systematic review and meta-analysis of RCTs comparing laparoscopic vertical SG with RYGB (Table 4).⁵⁸ The literature search, conducted from 2000 to November 2015, identified 9 RCTs for inclusion (N = 865 patients). Four trials were included in meta-analyses comparing percent EWL between the 2 groups. Results at both 6- and 12-month follow-ups showed that the procedures are comparable (Table 5). Osland et al. (2020) recently published a continuation of their work that focused exclusively on long-term (5 year) weight outcomes of laparoscopic vertical SG versus RYGB.⁵⁹ This systematic review and meta-analysis included 5 studies (SG = 520; RYGB = 508) and results revealed that a statistically significant BMI loss was seen with both SG: -11.37 kg/m2 (range: -6.3 to -15.7 kg/m2) and RYGB: -12.6 kg/m2 (range: -9.5 to -15.4 kg/m2) at 5 years. However, differences in reporting parameters limit the ability to reliably compare outcomes using statistical methods and the results may have been impacted by large dropout rates and per protocol analyses of the 2 largest included studies.

A systematic review by Juodeikis and Brimas (2017) summarized evidence on long-term results after SG (Table 3).⁶⁰ Reviewers included an RCT and 19 retrospective studies, with a total of 2,713 patients who received SG. Mean preoperative BMI was 46.9 kg/m². Mean duration of follow-up ranged from 5 to 11 years, and mean proportion of patients followed for 5 years was 68.5%. Seventeen studies (N = 1,501 patients) reported 5-year follow-up data. At 5 years, resolution of T2D, arterial hypertension, dyslipidemia, OSA, gastroesophageal reflux disease (GERD), and degenerative joint diseases also improved in most patients (Table 5). Two studies reported weight loss after 7 and 8 years; percent EWL rates were 56.6% and 54.8%, respectively.

In a meta-analysis of 21 randomized and nonrandomized studies (N = 18,766 patients) comparing SG with laparoscopic RYGB for morbid obesity, Zhang et al. (2015) reported no significant difference in percent EWL from 0.5- to 1.5-year follow-ups (Tables 3 and 4).⁶¹ However, after 1.5 years, RYGB was associated with higher percent EWL (2-year MD, 5.77; 95% CI, 4.29 to 7.25; p < .05). Adverse events were more frequent following RYGB (OR for major complication, 1.29; 95% CI, 1.22 to 3.22; p < .01).

Trastulli et al. (2013) conducted a systematic review of 15 RCTs (N = 1,191 patients) that compared SG with other bariatric procedures (Table 4).⁶² Summary statistics were provided; meta-analyses were not conducted (Table 5). Reviewers reported mean complication rates with SG of 12.1% (range, 10% to 13.2%) compared with 20.9% with LAGB (range, 10% to 26.4%). Percent EWL ranged from 49% to 81% with SG and from 62.1% to 94.4% with LAGB.

Brethauer et al. (2009) reviewed 36 studies (N = 2,570 patients) in a systematic review of SG as a staged and primary procedure, the largest trials coming from European centers (Table 4).⁶³ Thirteen studies (n = 821 patients) reported on high-risk patients having a staged approach and 24 studies (n = 1,749 patients) on SG as the primary procedure. Mean percent EWL, reported in 24 studies (n = 1,662 patients), was 55.4% overall. Mean postoperative BMI, reported in 26 studies (n = 1,940 patients), decreased from a baseline of 51.2 to 37.1 kg/m². Other studies reported weight loss in terms of BMI decrease, the percentage of BMI lost, or percentage of total weight lost; all had significant reductions from baseline. Rates of major postoperative complications ranged from 0% to 23.8% for all studies and from 0% to 15.3% in studies with more than 100 patients. Leaks (2.2%), bleeding episodes requiring reoperation (1.2%), and postoperative strictures requiring endoscopic or surgical intervention (0.6%) were reported in the 33 studies (n = 2,570 patients). All extracted studies reported mortality data, with 5 deaths within 30 days of surgery (overall mortality rate, 0.19%; 2 in the high-risk/staged group, 3 in the primary procedure group).

Table 4. Systematic Review Characteristics for Sleeve Gastrectomy

Study	Dates	Studies	Participants	Design	Duration
Lee et al. (2021) 56	Through Jan 2019	33	SG = 1252; RYGB = 1223	RCTs	1 to 5 y
Gu et al. (2020)50	Through Jan 2019	28	SG = 4597; RYGB = 4441	7 RCTs; 6 prospective; 15 retrospective	3 to 7 y
Han et al. (2020)51	Through Jan 2020	18	2917	9 RCTs; 9 nonrandomized studies of interventions	1 to 82.2 mo
Sharples et al. (2020)52	Through Dec 2018	5	729	RCTs	5 y
Shenoy et al. (2020)53	1991 to 2019	9	SG = 683; RYGB = 1557	RCTs; observational studies	Minimum follow-up: 1 y
Osland et al. (2017)58	2000 to Nov 2017	9	SG = 437; RYGB = 428	RCTs	3 mo to 5 y
Juodeikis et al. (2017)60	Through May 2016	20	1626	1 RCT; 19 retrospective	5 to 11 y
Zhang et al. (2015)61	Through Oct 2013	21	18,766	8 RCTs; 13 nonrandomized comparative	1 to 5 y
Trastulli et al. (2013)62	Through Nov 2012	15	1191	RCTs	6 mo to 3 y
Brethauer et al. (2009)63	1996 to 2009	36	2570	2 RCTs; 1 cohort; 33 case series	3 mo to 5 y

RCT: randomized controlled trial; RYGB: Roux-en-Y gastric bypass; SG: sleeve gastrectomy.

Table 5. Systematic Review Results for Sleeve Gastrectomy

Study	BMI mean difference (95% CI)	Comorbidities (95% CI)
Lee et al. (2021)56	Mean difference SG vs RYGB: 1 y (16 trials): -1.25 kg/m2 (-2.01 to -0.49) 3 y (5 trials): -1.71 kg/m2 (-2.68 to -0.74) 5 y (4 trials): -1.46 kg/m2 (-3.15 to 0.23)	Remission, SG vs RYGB: T2D (1 y): RR, 0.86 (0.71 to 1.04) T2D (3 y): RR, 0.88 (0.72 to 1.07) T2D (5 y): RR, 0.79 (0.57 to 1.10) Hypertension (5 y): RR, 0.86 (0.68 to 1.10) Dyslipidemia (5 y): RR, 0.68 (0.46 to 1.23)
	Percent EWL (95% CI)	Comorbidities (95% CI)
Gu et al. (2020)50	Weighted mean difference, RYGB and SG: 3 y (13 trials): -4.37 (-8.10 to -0.64) 5 y (9 trials): -2.20 (-3.83 to -0.57)	Remission, RYGB and SG: T2D (3 y): OR, 0.68 (0.48 to 0.95) T2D (5 y): OR, 0.63 (0.41 to 0.96) Hypertension (5 y): OR, 0.51 (0.38 to 0.68) Dyslipidemia (5 y): OR, 0.3 (0.19 to 0.48)
Han et al. (2020)51	Mean difference, RYGB and SG: RCTs: -0.16 (-0.52 to 0.19)	Resolution, RYGB and SG: T2D: RR, 1.07 (0.89 to 1.28) Dyslipidemia: RR, 1.36 (1.17 to 1.59) Hypertension: RR, 1.23 (1.04 to 1.45) GERD symptoms: RR, 0.16 (0.06 to 0.44)
Sharples et al. (2020)52	5 y: RYGB: 65.7% SG: 57.3%	RYGB vs. SG at 5 y: T2D resolution: 37.4% vs. 27.5% Diabetes improvement: 77.5% vs. 74% Hypertension resolution: 60.1% vs. 48.4% Hypertension improvement: 86.4% vs. 76.6% Dyslipidemia resolution: 68.6% vs. 55.2% GERD remission: 60.4% vs. 25%
Shenoy et al. (2020)53	Mean difference, RYGB and SG: -7.79 (-23.96 to 8.38)	Resolution, RYGB and SG: T2D (5 studies): OR, 1.02 (0.63 to 1.66) Hypertension (4 studies): OR, 0.57 (0.35 to 0.93) Obstructive sleep apnea (2 studies): OR, 1.14 (0.55 to 2.34)
Osland et al. (2017)58	Mean difference, SG and RYGB: 6 mo (3 trials): 0.5 (-5.0 to 6.0) 12 mo (2 trials): 7.6 (-0.1 to 15.3)	NR
Juodeikis et al. (2017)60	Mean rates for SG: 5 y (17 trials): 58.4% 7 y (2 trials): 56.6% 11 y (1 trial): 62.5%	Remission/improvement: T2D: 77.8% Hypertension: 68.0% Dyslipidemia: 65.9% Sleep apnea: 75.8%
Zhang et al. (2015)61	Mean difference, RYGB and SG: 6 mo (9 studies): 0.2 (-2.5 to 2.9) 12 mo (15 studies): 2.9 (-0.2 to 6.0) 4 y (3 studies): 2.7 (0.2 to 5.2)	Mean difference resolution, RYGB and SG: T2D (10 studies): 3.3 (2.0 to 5.5) Hypertension (10 studies): 1.3 (0.7 to 2.4) Dyslipidemia (5 studies): 1.1 (0.3 to 1.3) Sleep apnea (7 studies): 1.5 (0.8 to 2.6)
Trastulli et al. (2013)62	Mean by procedure: SG: 49% to 81% LGB: 62% to 94% LAGB: 29% to 48%	T2D: SG, 67% to 100% LGB, 80% to 100%
Brethauer et al. (2009)63	Mean rate overall for SG: 55% (range, 33% to 85%)	Remission/improvement: T2D: > 70% Significant reductions also seen in hypertension, hyperlipidemia, and sleep apnea

BMI: body mass index; CI: confidence interval; EWL: excess body weight loss; GERD: gastroesophageal reflux disease; LAGB: laparoscopic adjustable gastric banding; LGB: laparoscopic gastric bypass; NR: not reported; OR: odds ratio; RCT: randomized controlled trial; RR: relative risk; RYGB: Roux-en-Y gastric bypass; SG: sleeve gastrectomy; T2D: type 2 diabetes.

Randomized Controlled Trials
Hofso et al. (2019) published the results of a single-center, triple-blind RCT comparing the efficacy of RYGB (n = 54) versus SG (n = 55) on diabetes remission and ß-cell function in patients with obesity and T2D.⁶⁴ Inclusion criteria included previously verified BMI ≥ 35 kg/m2 and current BMI ≥ 33.0 kg/m2, HbA1c ≥6.5% or use of antidiabetic medications with HbA1c ≥6.1%, and age ≥18 years. One-year follow-up was completed by 107 (98%) of 109 patients, with 1 patient in each group withdrawing after surgery. In the intention-to-treat population, diabetes remission rates were superior in the gastric bypass group than in the SG group (risk difference 27%; 95% CI, 10 to 44; RR 1.57, 95% CI, 1.14 to 2.16; p = .0054). Results were similar in the per-protocol population (risk difference 27%; 95% CI, 10 to 45; RR 1.57; 95% CI, 1.14 to 2.15; p = .0036). The 2 procedures had a similar beneficial effect on ß-cell function.

Peterli et al. (2018) published a randomized study of adults with morbid obesity treated with either laparoscopic SG (LSG) or RYGB.65, Two hundred five patients (mean age, 45.5 years; mean BMI, 43.9; 72% women) treated at 4 Swiss bariatric centers were randomly assigned to receive SG (n = 101) or RYGB (n = 104) with 5-year follow-up. Excess BMI loss was 61.6% for SG and 68.3% for RYGB (95% CI, -14.30 to -0.06; p = .22). Gastric reflux remission was seen in 25.0% of SG and 60.4% of RYGB patients. Reoperations or interventions were necessary for 16/101 (15.8%) in the SG group and 23/104 (22.1%) of the RYGB group. The study was limited by the lack of analysis of diabetes remission information, and the results may not be generalizable.

Salminen et al. (2018) published a randomized trial, Laparoscopic Gastric Bypass vs. Laparoscopic Sleeve Gastrectomy in the Treatment of Morbid Obesity (SLEEVEPASS), comparing 5-year outcomes of morbidly obese patients (n = 240; mean age, 48 years; mean baseline BMI, 45.9; 69.6% women) who underwent either LSG (n = 121) or RYGB (n = 119).66, Five-year estimated mean percentage excess BMI losswas 49% (95% CI, 45 to 52) for SG and 57% (95% CI, 53 to 61) for gastric bypass. For SG and RYGB, respectively, rates of remission of T2D were 37% (n = 15/41) and 45% (n = 18/40; p>0.99). Medication for hypertension was discontinued in 20/68 (29%) SG patients and 37/73 (51%) RYGB patients (p = .02). Overall 5-year morbidity rate was 19% for SG and 26% for RYGB (p = .19), and there was no significant difference in quality of life between groups (p = .85). The study was limited by the following: (1) only a small number (n = 430) of bariatric procedures were performed in Finland at trial initiation in 2008, meaning a learning curve could account for some earlier technical complications, (2) the study had a higher reoperation rate for SG than other trials reported, (3) approximately 20% of patients were lost to follow-up, and (4) there was a lack of reliable information for diabetes duration at baseline.

Wolnerhanssen et al. (2021) pooled 5-year outcomes data from the 2018 studies by Peterli et al. and Salminen et al.67, Five-year follow-up was available for 199 of 228 patients after SG and 199 of 229 after RYGB. Patients who underwent SG had an estimated 7% greater excess BMI loss versus RYGB (p < .001). While remission rates for hypertension were better after RYGB versus SG (60.3% vs 44.9%; p < .049), between-group differences in rates of remission of T2D, OSA, or quality of life scores did not reach statistical significance. The rate of complications was higher after RYGB versus SG (37.2% vs 22.5%; p = .001), but there was no difference in mean Comprehensive Complication Index value (30.6 vs 31.0 points; p = .859).

An RCT comparing short-term outcomes of laparoscopic SG with gastric bypass was published in 2012.68, Trialists compared 30-day outcomes for 117 patients randomized to gastric bypass with 121 patients randomized to LSG. The rate of major complications (no deaths in either group) was 9.4% in the gastric bypass group compared with 5.8% in the LSG group (p = .29). Minor complications were more common in the gastric bypass group than in the LSG group (17.1% versus 7.4%, p = .02), as were combined major and minor complications (26.5% versus 13.2%, p = .01).

Karamanakos et al. (2008) carried out a double-blind RCT comparing outcomes of laparoscopic RYGB and LSG on body weight, appetite, fasting, and postprandial ghrelin and peptide YY (levels at 1, 3, 6, and 12 months after surgery).69, Thirty-two patients were randomized, half to each procedure. The decrease in body weight and BMI were marked and comparable in each group. EWL was greater after LSG than laparoscopic RYGB at 6 months (55.5% versus 50.2%; p = .04) and 12 months (69.7% versus 60.5%; p = .05), all respectively. Fasting peptide YY levels increased after both surgical procedures. Appetite decreased in both groups but decreased more after LSG.

Himpens et al. (2006) reported on a randomized trial comparing LAGB with isolated LSG in 80 patients and reported 3-year follow-up.70, Median baseline BMI was 37 kg/m2 (range, 30 to 47 kg/m2) in the LAGB group and 39 kg/m2 (range, 30 to 53 kg/m2) in the SG group. Outcomes of weight loss, feeling of hunger, sweet-eating, GERD, complications, and reoperations were recorded at 1- and 3-year follow-ups. Median decrease in BMI in the gastric bypass group was 15.5 kg/m2 (range, 5 to 39 kg/m2) after 1 year and 18 kg/m2 (range, 0 to 39 kg/m2) at 3 years after LAGB. One year after SG, decrease in BMI was 25 kg/m2 (range, 0 to 45 kg/m2) and 27.5 kg/m2 (range, 0 to 48 kg/m2) after 3 years. Median EWL in the LAGB group was 41.4% after 1 year and 48% at 3 years. Median EWL after SG was 58% and 66% at 1 and 3 years, respectively. More patients having SG than LAGB reported a loss of craving for sweets, but the difference was not statistically significant; GERD appeared de novo in more SG than LAGB patients at 1 year, and the relation reversed at 3 years; between-group differences were not statistically significant at either time point. Two SG patients required reoperation for complications. Seven late complications required reoperation after LAGB, including pouch dilations treated by band removal (n = 2) or conversion to RYGB (n = 1), 1 gastric erosion treated by conversion to RYGB, and 3 system disconnections that required reconnection. Four patients had reoperations for lack of efficacy (2 LAGB patients underwent conversion to RYGB, 2 SG patients underwent conversion to DS). The trialists noted that the number of reoperations was significant in both groups and that the severity of complications was greater in the SG group.

Section Summary: Sleeve Gastrectomy for Adults with Morbid Obesity
Systematic reviews of RCTs and observational studies, evaluating SG alone and comparing SG with RYGB, have found that SG results in substantial weight loss, comparable to RYGB and that this weight loss is durable for at least 5 years. A meta-analysis found that short-term weight loss was similar after SG or gastric bypass. Long-term weight loss was greater after gastric bypass, but SG is associated with fewer adverse events.

Biliopancreatic Diversion With Duodenal Switch for Adults with Morbid Obesity
Clinical Context and Therapy Purpose
The purpose of BPD-DS is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as standard medical care, in patients who are adults with morbid obesity.

The question addressed in this evidence review is: Does the BPD-DS procedure improve the net health outcome in adults who are obese?

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals who are adults with morbid obesity.

Interventions
The therapy being considered is BPD-DS. BPD may be performed with or without the DS procedure. In the BPD-DS, a SG is performed, preserving the pyloric sphincter. Preservation of the pyloric sphincter is intended to ameliorate dumping syndrome and to decrease the incidence of ulcers at the duodeno-ileal junction by providing a more physiologic transfer of stomach contents to the duodenum.

Comparators
Comparators of interest include standard medical care. Treatment for adults with morbid obesity includes physical exercise, low carbohydrate dieting, and low-fat dieting.

Outcomes
The general outcomes of interest are OS, change in disease status, functional outcomes, health status measures, quality of life, treatment-related mortality, and treatment-related morbidity.

The existing literature evaluating BPD-DS as a treatment for morbid obesity has varying lengths of follow-up, ranging from 1 to 15 years. While studies described below all reported at least 1 outcome of interest, longer follow-up was necessary to fully observe outcomes. One-year follow-up is necessary to demonstrate weight loss efficacy. Longer follow-up to 5 to 10 years is desirable to assess maintenance of weight loss, impact on co-occurring conditions, and appearance of long-term complications.

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

• To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs.
• In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
• To assess longer-term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
• Consistent with a "best available evidence approach," within each category of study design, studies with larger sample sizes and longer durations were sought.
• Studies with duplicative or overlapping populations were excluded.

Review of Evidence
Systematic Review
In an evidence-based review of literature, Farrell et al. (2009) summarized data on BPD with or without DS, RYGB (proximal), and LAGB, and reported that at a mean 1-year follow-up, EWL for BPD with or without DS (outcomes with and without DS not reported separately) was 72% (4 studies; n = 896 patients), 67% for RYGB (7 studies; n = 1,627 patients), and 42% for LAGB (11 studies; n = 4,456 patients).⁷¹ At mean follow-up of 5 years, EWL for BPD with or without DS was 73% (3 studies; n = 174 patients), 58% for RYGB (3 studies; n = 176 patients), and 55% for LAGB (5 studies; n = 640 patients). Reviewers noted that “given the marked paucity of prospectively collected comparative data among the different bariatric operations, it remains impossible to make definitive recommendations for one procedure over another.”

Nonrandomized Comparative Studies
Skogar et al. (2017) published results from a retrospective mail survey of patients undergoing BPD-DS (n = 113) or RYGB (n = 98) (Table 6).⁷² Reduction in BMI was statistically larger in patients receiving BPD-DS compared with patients receiving RYGB. Both groups experienced significant reductions in diabetes and OSA. Significant reductions in dyslipidemia were only seen in the group receiving BPD-DS. The overall complication rate was lower for patients undergoing RYGB.

Strain et al. (2007) published a comparative study of 72 patients who underwent RYGB (n = 50) or BPD (n = 22) (Table 6).⁷³ Choice of surgery was by the surgeon and/or patient, and the patient populations differed by age and time since surgery. Weight loss at 1 year was greater for BPD, with a reduction in BMI of 10.6 kg/m2 (23.3 lb) for BPD compared with 7.5 kg/m2 (16.5 lb) for RYGB (p < .001).

Prachand et al. (2006) published the largest comparative study of 350 super-obese patients with a BMI greater than 22.7 kg (50 lb) who underwent RYGB (n = 152) or Scopinaro BPD combined with the DeMeester BPD-DS (n = 198) (Table 6).⁷⁴ In this retrospective study, the decision for surgery was made by the surgeon and/or patient. The BPD-DS patients differed from RYGB patients on baseline weight and BMI; mean weight was 167 kg (368 lb; range, 267 to 597 lb) in BPD-DS patients and 157 kg (346 lb; range, 240 to 505 lb) in the RYGB group, and mean BMI was 27 kg/m2 (59 lb; range, 50 to 96 lb) in BPD-DS patients versus 26 kg/m2 (56 lb; range, 50 to 84 lb) in the RYGB group. At 1 year, data were reported for 143 BPD-DS patients and 81 RYGB patients (Table 7). EWL was greater for BPD (64.1%) versus RYGB (55.9%; p < .01), and the reduction in BMI was also greater with BPD (10.7 kg/m2 [23.6 lb]) versus RYGB (8.8 kg/m2 [19.4 lb]; p < .001). Complications and data on the resolution of comorbidities were not reported.

Table 6. Nonrandomized Comparative Study Characteristics for Biliopancreatic Diversion With Duodenal Switch

Study	Country	Dates	Participants	Follow-Up
Skogar et al. (2017)72	Sweden	2003 – 2012	BPD-DS: 113 RYGB: 98	4 y
Strain et al. (2007)73	U.S.	2002 – 2005	BPD-DS: 22 RYGB: 50	BPD-DS: 19 mo RYGB: 15 mo
Prachand et al. (2006)74	U.S.	2002 – 2005	BPD-DS: 198 RYGB: 152	3 y

BPD-DS: biliopancreatic diversion with duodenal switch; RYGB: Roux-en-Y gastric bypass.

Table 7. Nonrandomized Comparative Study Results for Biliopancreatic Diversion With Duodenal Switch

Study	Mean Reduction in BMI (SD)		Percent Achieving ≥50% EWL
	Presurgery, kg/m2	Postsurgery, kg/m2	pa	1 Year	2 Years	3 Years
Skogar et al. (2017)72 BPD-DS RYGB	56 (6.7) 52 (4.0)	31 (5.5) 36 (7.1)	< 0.01	NR NR
Strain et al. (2007)73 BPD-DS RYGB			< 0.01	NR NR
		Change in BMI
Prachand et al. (2006)74 BPD-DS RYGB	59 (6.7) 56 (6.8)	27.8 18.9	< 0.01	83.9 70.4b	89.2 79.3	84.2 59.3b

BMI: body mass index; BPD-DS: biliopancreatic diversion with duodenal switch; EWL: excess weight loss; NR: not reported; RYGB: Roux-en-Y gastric bypass; SD: standard deviation.
^a Between groups, difference in change.
^b p < .05.

Case Series
Strain et al. (2017) reported on the nutrient status of 190 patients receiving BPD-DS after 9 years of follow-up.⁷⁵ At baseline, the patients had a mean age of 43 years and mean BMI of 53 kg/m². All patients reported taking some supplements. Deficiencies in protein, iron, and calcium developed by year 3 and continued through the study. Zinc deficiencies developed by year 5. Folate levels increased during the study, probably due to the efficacy of the supplement. The authors warned that interventions need to be implemented to improve nutrient status in patients receiving BDP-DS.

The largest case series of this procedure is by Marceau et al. (2009), who reported on their 15-year experience with DS in 1423 patients from 1992 to 2005.⁷⁶ Follow-up evaluations were available for 97% of patients. Survival rate was 92%. After a mean of 7 years (range, 2 to 15 years), 92% of patients with an initial BMI of 50 kg/m² or less obtained a BMI of 35 kg/m² or less, and 83% of patients with a BMI greater than 50 kg/m2 achieved a BMI of less than 40 kg/m². Diabetes medication was discontinued in 92% and decreased in others. Use of continuous positive airway pressure was discontinued in 92% of patients, and the prevalence of cardiac risk index greater than 5 decreased by 86%. Operative mortality was 1%, the revision rate was 0.7%, and the reversal rate was 0.2%. Revision for failure to lose sufficient weight was needed in only 1.5% of patients. Severe anemia, vitamin deficiency, or bone damage were preventable or easily treated and without documented permanent damage.

Section Summary: Biliopancreatic Diversion With Duodenal Switch for Adults With Morbid Obesity
Nonrandomized comparative studies have found significantly higher weight loss after BPD-DS compared with gastric bypass at 1 year. A large case series found sustained weight loss after 7 years.

Biliopancreatic Diversion without Duodenal Switch for Adults With Morbid Obesity
Clinical Context and Therapy Purpose
The purpose of BPD without DS is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as standard medical care, in patients who are adults with morbid obesity.

The question addressed in this evidence review is: Does the BPD without DS procedure improve the net health outcome in adults who are obese?

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals who are adults with morbid obesity.

Interventions
The therapy being considered is BPD without DS.

Comparators
Comparators of interest include standard medical care. Treatment for adults with morbid obesity includes physical exercise, low carbohydrate dieting, and low-fat dieting.

Outcomes
The general outcomes of interest are OS, change in disease status, functional outcomes, health status measures, quality of life, treatment-related mortality, and treatment-related morbidity.

The existing literature evaluating BPD without DS as a treatment for morbid obesity has varying lengths of follow-up, ranging to 9 years. While studies described below all reported at least 1 outcome of interest, longer follow-up was necessary to fully observe outcomes. One-year follow-up is necessary to demonstrate weight loss efficacy. Longer follow-up of 5 to 10 years is desirable to assess maintenance of weight loss, impact on co-occurring conditions, and appearance of long-term complications.

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

• To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs.
• In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
• To assess longer-term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
• Consistent with a "best available evidence approach," within each category of study design, studies with larger sample sizes and longer durations were sought.
• Studies with duplicative or overlapping populations were excluded.

Review of Evidence
Systematic Review
The available evidence on BPD-DS was reviewed in the 2006 TEC Assessment, and BPD outcomes, with or without DS, were compared with those of gastric bypass.⁴⁶ One comparative trial and 7 single-arm series suggested that weight loss outcomes at 1 year were in the same range as for gastric bypass. While these data were not sufficient to distinguish small differences in weight loss between the 2 procedures, they did not support the hypothesis that BPD resulted in greater weight loss than open gastric bypass.

Randomized and Nonrandomized Studies
Complication rates have been poorly reported in these trials. The data have suggested that mortality is low (1%) and in the same range as for open gastric bypass. However, rates of other complications, especially long-term complications, cannot be determined from these data. Limited data have suggested that long-term nutritional and vitamin deficiencies occur at a high rate following BPD. Slater et al. (2004) focused specifically on vitamin and calcium deficiencies following BPD.⁷⁷ The authors reported high rates of vitamin and calcium abnormalities in their population over a 4-year period. By year 4, 48% of patients had low calcium, and 63% had low levels of vitamin D. Other fat-soluble vitamins showed similar patterns of abnormalities. Low vitamin A was found in 69% of patients at 4 years, low vitamin K in 68%, and low zinc in 50%. Dolan et al. (2004) reported similar data in a study that compared several technical variations of BPD.⁷⁸ The authors reported low calcium levels in 12% to 34% of patients, low vitamin D in 22.2% to 70.6%, low vitamin A in 53% to 67%, and low vitamin K in 44% to 59%. Also, this study reported high rates of iron deficiency (11% to 47%) and anemia (11% to 40%).

Skroubis et al. (2006) randomized 130 patients with a BMI of 35 to 50 kg/m² to RYGB or BPD without DS using a variant of BPD that included RYGB in place of SG.⁷⁹ All patients were followed for at least 2 years. Weight loss outcomes were superior for the BPD group at every interval examined up to 2 years. EWL at 1 year was 73.7% for RYGB and 83.1% for BPD (p < .001); at 3 years, EWL was 72.6% for RYGB and 83.1% for BPD (p < .001). There were more early complications in the RYGB group, but this difference was not statistically significant (6 complications versus 1, respectively; p = .12). Late complications also did not differ significantly between the RYGB group (16 complications) and BPD groups (22 complications; p = .46).

Case Series
Numerous clinical series of BPD have been published but high-quality trials directly comparing outcomes of this procedure with gastric bypass are lacking. The largest experience with BPD (N = 1,217 patients) was reported by Scopinaro et al. (1996), who developed the procedure.⁸⁰ With a follow-up of up to 9 years, the authors reported a durable EWL of 75%, suggesting that weight loss is greater with this procedure than with gastric restrictive procedures. Also, most patients reported disappearance or improvement of complications such as OSA, hypertension, hypercholesterolemia, and diabetes. The authors considered protein malnutrition to be the most serious metabolic complication, occurring in almost 12% of patients and responsible for 3 deaths. This complication could require inpatient treatment with total parenteral nutrition. To address protein malnutrition, 4% of patients underwent reoperation to elongate the common limb (thus increasing protein absorption) or to have the operation reversed, restoring normal intestinal continuity. The authors also found that protein malnutrition was strongly related to ethnicity and, presumably, patient eating habits, with an increased incidence among those from southern Italy where the diet contains more starch and carbohydrates than the north. Peripheral neuropathy may occur in the early postoperative period due to excessive food limitation but may be effectively treated with large doses of thiamine. Bone demineralization, due to decreased calcium absorption, was seen in about 33% of patients during the first 4 postoperative years. All patients were encouraged to maintain an oral calcium intake of 2 g/d, with monthly vitamin D supplementation.

Section Summary: Biliopancreatic Diversion without Duodenal Switch for Adults With Morbid Obesity
A TEC Assessment reviewed the available observational studies and concluded that weight loss was similar after BPD without the DS and gastric bypass. However, BPD without DS leads to complications, especially long-term nutritional and vitamin deficiencies.

Vertical-Banded Gastroplasty for Adults With Morbid Obesity
Clinical Context and Therapy Purpose
The purpose of VBG is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as standard medical care, in patients who are adults with morbid obesity.

The question addressed in this evidence review is: Does the VBG procedure improve the net health outcome in adults who are obese?

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals who are adults with morbid obesity.

Interventions
The therapy being considered is VBG. In this procedure, the stomach is segmented along its vertical axis, a plug of the stomach is removed, and a propylene collar is placed through this hole and then stapled to itself. It can be performed using an open or laparoscopic approach.

Comparators
Comparators of interest include standard medical care. Treatment for adults with morbid obesity includes physical exercise, low carbohydrate dieting, and low-fat dieting.

Outcomes
The general outcomes of interest are OS, change in disease status, functional outcomes, health status measures, quality of life, treatment-related mortality, and treatment-related morbidity.

Negative outcomes associated with VBG include complications such as esophageal reflux, dilation, or obstruction of the stoma.

The existing literature evaluating VBG as a treatment for morbid obesity has varying lengths of follow-up, ranging from 3 to 10 years. While studies described below all reported at least 1 outcome of interest, longer follow-up was necessary to fully observe outcomes. Therefore, 3 to 10 years of follow-up is considered necessary to demonstrate efficacy.

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

• To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs.
• In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
• To assess longer-term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
• Consistent with a "best available evidence approach," within each category of study design, studies with larger sample sizes and longer durations were sought.
• Studies with duplicative or overlapping populations were excluded.

Review of Evidence
Systematic Reviews
Hseih et al. (2014) conducted a systematic review of studies reporting greater than 10-year follow-up for VBG, which included 3 studies with extractable data.⁸¹ Mean EWL was 61.4% from baseline to follow-up in the 3 studies, but reviewers noted a lack of long-term evidence related to outcomes following VBG.

A number of nonrandomized, comparative studies of open gastric bypass versus VBG were included in the 2003 TEC Assessment (N = 8 studies, total N = 3,470 patients).⁸² All 8 studies reported greater amounts of weight loss with open gastric bypass. These studies reported a 44% to 70% improvement in TWL, a 28% to 43% improvement in the percent EWL, and 19% to 36% more patients with more than 50% EWL for those undergoing gastric bypass compared with VBG. Comparison of adverse events was difficult because the data did not permit rigorous assessment. Nevertheless, the data suggested that the mortality rate for both surgeries was low overall. Serious perioperative adverse events were also infrequently reported, but somewhat higher for gastric bypass. Long-term adverse events were inconsistently reported, although it appeared that revision rates were higher for VBG.

Randomized Controlled Trials
A small body of literature has compared outcomes between VBG and open gastric bypass. The most rigorous of these comparative trials, the Adelaide Study (1990), randomized 310 morbidly obese patients to gastric bypass, VBG, or horizontal gastroplasty.⁸³ The percentage of patients with greater than 50% EWL at 3-year follow-up was 67% for gastric bypass, 48% for VBG, and 17% for horizontal gastroplasty (p < .001). There were no demonstrable differences in adverse events across groups.

A second, smaller RCT by Sugerman et al. (1987) randomized 40 patients to a VBG or a gastric bypass procedure.⁸⁴ After 9 months, the gastric bypass patients had significantly greater weight loss that was maintained at 3-year follow-up. The gastric bypass patients lost approximately 64% of excess weight, whereas the gastroplasty patients lost 37% of excess weight.

Case Series
Relatively high rates of complications, revisions, and reoperations led to the abandonment of VBG as a bariatric surgery procedure in the United States. An example of these results is a large case series with long-term follow-up by MacLean et al. (1990), who reported on 201 patients undergoing VBG followed for a minimum of 2 years.⁸⁵ Staple line perforation occurred in 48% of patients, and 36% underwent reoperation either to repair the perforation or to repair a stenosis at the rate-limiting orifice. However, the more than 50% of patients who maintained an intact staple line had a durable weight loss of 75% to 100% of excess weight.

Section Summary: Vertical-Banded Gastroplasty for Adults With Morbid Obesity
A TEC Assessment identified 8 nonrandomized comparative studies evaluating VBG with gastric bypass. The assessment found that weight loss was significantly greater with open gastric bypass than with VBG. Also, VBG has relatively high rates of complications, revisions and reoperations.

Two-Stage Bariatric Surgery Procedures for Adults with Morbid Obesity
Clinical Context and Therapy Purpose
The purpose of 2-stage bariatric surgery procedures is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as standard medical care, in patients who are adults with morbid obesity.

The question addressed in this evidence review is: Do 2-stage bariatric surgery procedures improve the net health outcome in adults who are obese?

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals who are adults with morbid obesity.

Interventions
The therapy being considered is 2-stage bariatric surgery. Bariatric surgeries performed in 2 stages have been proposed as a treatment option, particularly for patients with “super-obesity” defined as a BMI greater than 50 kg/m². The rationale for a 2-stage procedure is that the risk of an extensive surgery is prohibitive in patients who are extremely obese. Therefore, a procedure with low-risk (usually an SG) is performed first. After the patient loses some weight, thus lowering the surgical risk, a second more extensive procedure (e.g., BPD) is performed.

Comparators
Comparators of interest include standard medical care. Treatment for adults with morbid obesity includes physical exercise, low carbohydrate dieting, and low-fat dieting.

Outcomes
The general outcomes of interest are OS, change in disease status, functional outcomes, health status measures, quality of life, treatment-related mortality, and treatment-related morbidity.

The existing literature evaluating 2-stage bariatric surgery as a treatment for morbid obesity has varying lengths of follow-up, ranging from 1 to 5 years. While studies described below all reported at least 1 outcome of interest, longer follow-up was necessary to fully observe outcomes. Therefore, 1 to 5 years of follow-up is considered necessary to demonstrate efficacy.

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

• To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs.
• In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
• To assess longer-term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
• Consistent with a "best available evidence approach," within each category of study design, studies with larger sample sizes and longer durations were sought.
• Studies with duplicative or overlapping populations were excluded.

Review of Evidence
Randomized Controlled Trial
Coffin et al. (2017) published results on the use of intragastric balloon (IGB) prior to a laparoscopic gastric bypass in patients with super-obesity.⁸⁶ Patients with a BMI greater than 45 kg/m² were randomized to an IGB (n = 55) or standard medical care (n = 60) during the 6 months prior to a planned laparoscopic gastric bypass procedure. Five patients had the IGB removed earlier than 6 months due to complications (n = 3) or patient request (n = 2). Patients receiving IGBs during the first 6 months of the study experienced significantly more BMI reduction (2.8 kg/m²; range 1.7 to 6.2 kg/m²) than patients receiving standard care (0.4 kg/m²; range 0.3 to 2.2 kg/m²). Weight loss during months 6 through 12, after the laparoscopic gastric bypass procedure, was greater in the patients who received standard of care before the procedure. Duration of hospitalization after laparoscopic gastric bypass and quality of life did not differ between groups.

Case Series
Most of the evidence on 2-stage procedures consists of case series of patients undergoing SG as the initial procedure. Many do not report on the second-stage surgery. A minority of patients undergoing first-stage surgery proceed to second-stage surgery. Cottam et al. (2006) reported on 126 patients with a mean BMI of 65 kg/m² who underwent LSG as the first phase of a planned 2-stage procedure.⁸⁷ The incidence of major perioperative complications for LSG was 13%. After 1 year, mean EWL was 46%. Thirty-six (29%) patients proceeded to the second-stage procedure, which was laparoscopic gastric bypass. The incidence of major complications following the second procedure was 8%.

In a similar study, Alexandrou et al. (2012) reported on 41 patients who underwent SG as the first-stage of a planned 2-stage procedure.⁸⁸ After 1-year of follow-up, 12 (29%) patients achieved a BMI of less than 35 kg/m² and were ineligible for the second-stage procedure. Of the remaining 28 patients, 10 (24%) underwent the second-stage procedure. The remaining 18 (44%) patients were eligible for but had not undergone, the second-stage procedure at the last follow-up.

Patients who undergo 2-stage procedures are at risk for complications from both procedures. Silecchia et al. (2009) described the complication rates in 87 patients who underwent a stage 1 SG followed by BPD in 27 patients.⁸⁹, For the first stage, 16.5% of patients had complications of bleeding, fistula, pulmonary embolism, acute renal failure, and abdominal abscess. For the 27 patients who underwent the second-stage BPD, 29.6% had major complications, including bleeding, duodeno-ileal stenosis, and rhabdomyolysis.

Section Summary: Two-Stage Bariatric Surgery Procedures for Adults With Morbid Obesity
The evidence from an RCT and several case series does not support a 2-stage bariatric surgery procedure for improving outcomes in patients with extreme levels of obesity. There is no evidence to suggest that weight loss is improved or that complications are reduced by this approach. Most patients who receive SG as the initial procedure lose sufficient weight during the first year so that a second procedure is no longer indicated. Also, patients undergoing a 2-stage procedure are at risk for complications from both procedures; therefore, it is likely that overall complications are increased by this approach.

Laparoscopic Gastric Plication for Adults With Morbid Obesity
Clinical Context and Therapy Purpose
The purpose of laparoscopic gastric plication is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as standard medical care, in patients who are adults with morbid obesity.

The question addressed in this evidence review is: Does laparoscopic gastric plication improve the net health outcome in adults who are obese?

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals who are adults with morbid obesity.

Interventions
The therapy being considered is laparoscopic gastric plication. Laparoscopic gastric plication is a bariatric procedure that involves laparoscopic placement of sutures over the greater curvature (laparoscopic greater curvature plication) or anterior gastric region (laparoscopic anterior curvature plication) to create a tube-like stomach. To achieve gastric restriction the procedure requires 2 main steps, mobilization of the greater curvature of the stomach and suture plication of the stomach.

Comparators
Comparators of interest include standard medical care. Treatment for adults with morbid obesity includes physical exercise, low carbohydrate dieting, and low-fat dieting.

Outcomes
The general outcomes of interest are OS, change in disease status, functional outcomes, health status measures, quality of life, treatment-related mortality, and treatment-related morbidity.

The existing literature evaluating laparoscopic gastric plication as a treatment for morbid obesity has varying lengths of follow-up, ranging from 1 to 12 years. While studies described below all reported at least 1 outcome of interest, longer follow-up was necessary to fully observe outcomes. One-year follow-up is necessary to demonstrate weight loss efficacy. Longer follow-up of 5 to 10 years is desirable to assess maintenance of weight loss, impact on co-occurring conditions, and appearance of long-term complications.

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

• To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs.
• In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
• To assess longer-term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
• Consistent with a "best available evidence approach," within each category of study design, studies with larger sample sizes and longer durations were sought.
• Studies with duplicative or overlapping populations were excluded.

Review of Evidence
Systematic Reviews
Li et al. (2021) reported on a systematic review of 18 studies (N = 1,329) comparing outcomes after laparoscopic SG versus laparoscopic greater curvature gastric plication.⁹⁰ Results demonstrated that SG is superior to greater curvature gastric plication with regard to providing effective weight loss through 24 months; statistical significance was not reached at 36 months. The difference in the improvement of comorbidities such as T2D, hypertension, and OSA did not reach statistical significance between groups, nor did the risks of major complications or mortality.

Ji et al. (2014) reported on a systematic review of 14 studies reporting outcomes after laparoscopic gastric plication.⁹¹ Reviewers included a nonrandomized matched cohort analysis, 10 uncontrolled case series, and 3 case reports. The nonrandomized cohort study was small (N = 19). Only 3 studies identified included more than 100 patients. Mean preoperative BMI ranged from 31.2 to 44.5 kg/m². Mean percent EWL after the procedure was reported in 9 studies (n = 1,407 patients), and ranged from 31.8% to 74.4% at follow-up times ranging from 6 to 24 months. One study reported weight loss in terms of percent decrease in BMI, with a reported decrease at 6 and 12 months of 66.4% and 60.2%, respectively. One study compared anterior plication with greater curvature plication and reported increased weight loss with greater curvature plication (percent EWL, 53.7% versus 23.3%, respectively). Reporting of complications was heterogeneous across studies, but no deaths were reported, and the rate of major postoperative complications requiring reoperation ranged from 0% to 15.4% (average, 3.7%), most commonly due to gastric obstruction or gastric perforation. Surgical techniques were not standardized.

In a systematic review, Abdelbaki et al. (2012) summarized outcomes from 7 studies of laparoscopic gastric plication, 2 of which enrolled more than 100 patients (N = 307 patients).⁹² All studies reported some incidence of nausea and vomiting, most of which were mild. Twenty (6.5%) patients were readmitted, of whom 14 (4.6%) patients required reoperation, most commonly for gastric obstruction (8/14 [57%]). Table 8 provides a comparison of the studies included in these systematic reviews. Tables 9 and 10 discuss characteristics and results, respectively.

Table 8. Comparison of Trials/Studies Included in SR & M-A

Study	Li et al. (2021)	Ji et al. (2014)	Abdelbaki et al. (2021)
Abdelbaki et al. (2014)	⚫
Abdelnazer et al. (2016)	⚫
Abouzeid et al. (2015)	⚫
Atlas et al. (2013)		⚫
Brethauer et al. (2011)		⚫	⚫
Buzga et al. (2017)	⚫
Casajoana et al. (2017)	⚫
Chouillard et al. (2015)	⚫
Fried et al. (2012)		⚫
Grubnik et al. (2015)	⚫
Hi et al. (2012)		⚫
Li et al. (2018)	⚫
Lopeznava et al. (2020)	⚫
Morshed et al. (2011)	⚫
Miu et al. (2013)		⚫
Nabil et al. (2018)	⚫
Neagoe et al. (2019)	⚫
Niazi et al. (2013)		⚫
Park et al. (2017)	⚫
Pujol Gebelli et al. (2011)		⚫	⚫
Ramos et al. (2010)		⚫	⚫
Sharma et al. (2014)	⚫
Skrekas et al. (2011)		⚫	⚫
Shen et al. (2013)	⚫	⚫
Taha et al. (2012)		⚫
Talebpour et al. (2007)			⚫
Talebpour et al. (2017)	⚫	⚫
Toprak et al. (2015)	⚫
Tsang et al. (2012)		⚫	⚫
Verdi et al. (2015)	⚫
Watkins et al. (2012)		⚫	⚫

M-A: meta-analysis; SR: systematic review. 

Table 9. Systematic Review Characteristics for Laparoscopic Gastric Plication

Study	Dates	Studies	Participants	Design	Duration
Li et al. (2021)90	Dec 2020	18	1,329	6 retrospective cohort; 7 prospective cohort; 5 RCTs	1 mo to 3 y
Ji et al. (2014)91	Jun 2013	14	1,450	1 matched cohort; 10 case series; 3 case reports	6 mo to 10 y
Abdelbaki et al. (2012)92	NR	7	307	5 case series; 2 case reports	3 y

NR: not reported; RCT: randomized controlled trial.

Table 10. Systematic Review Results for Laparoscopic Gastric Plication

Study	% Excessive Weight Loss	Complications	Conclusions
Li et al. (2021)90	MD (95% CI) between SG and gastric plication: 6 mo: 5.37 (1.59 to 9.16) 12 mo: 13.23 (9.93 to 16.54) 24 mo: 19.62 (1.15 to 38.08) 36 mo: 24.63 (-1.94 to 51.21)	OR (95% CI) between SG and gastric plication: Bleeding: 1.37 (0.61 to 3.09) Stenosis: 0.57 (0.23 to 1.38) Leak: 1.58 (0.61 to 4.15) Mortality: 1.39 (0.09 to 22.55)	SG is superior to gastric plication with regard to providing effective weight loss in the short- and mid-term. The procedures are similar in terms of major complications.
		Rate % (range)
Ji et al. (2014)91	31.8 to 74.4	3.7 (0 to 15.4)	Favorable short-term efficacy and safety profile; long-term follow-up and prospective trials needed
Abdelbaki et al. (2012)92	6 mo: 51 to 54 12 mo: 53 to 67	8 (7 to 15.3)	Prospective randomized trials vs. gastric plication with established bariatric procedures needed

CI: confidence interval; MD: mean difference; OR: odds ratio; SG: sleeve gastrectomy.

Randomized Controlled Trials
In additional to the studies included in the above-summarized systematic reviews, Sullivan et al. (2017) published results from the randomized, subject and evaluator-blinded, parallel-group, multicenter clinical trial using an endoscopic suturing device (G-CATH EZ™ suture anchor delivery catheter) for primary weight loss (ESSENTIAL), a randomized sham-controlled trial evaluating the efficacy and safety of endoscopic gastric plication (Table 11).⁹³ Patients (N = 332) were randomized 2:1 to the active or sham procedure. All patients were provided low-intensity lifestyle therapy (LT). The primary endpoint was total body weight loss (TBWL) at 12-month follow-up. The MD in TBWL for patients receiving the procedure compared with patients receiving the sham procedure was 3.6% (95% CI, 2.1 to 5.1). Significant differences between the active and sham groups were also reported in a change in weight from baseline, percent excess weight loss, BMI, and improvement in diabetes (Table 12). No significant differences were detected in improvements in hyperlipidemia or hypertension between the treatment groups.

Table 11. RCT Characteristics for Laparoscopic Gastric Plication

					Interventions
Interventions	Countries	Sites	Dates	Participants	Active	Comparator
Sullivan et al. (2017)93	U.S.	11	2013 – 2014	Patients 22 to 60 y, BMI ≥ 30 kg/m2, and ≥ 1 obesity-related comorbidity, or BMI ≥ 35 kg/m2, and with or without obesity-related comorbidity Race (active, sham): White: 71%, 64.8% Indian: 0%, 0.9% Black: 28.1%, 31.5% Mixed: 0.9%, 2.8% Ethnicity (active, sham) Not Hispanic/Latino: 93.7%, 92.8% Hispanic/Latino: 6.3%, 7.2%	Endoscopic gastric plication (n = 221)	Sham procedure (n = 111)

BMI: body mass index; RCT: randomized controlled trial.

Table 12. RCT Results for Laparoscopic Gastric Plication

Study; Trial Name	BMI Reduction	Weight Lossa
	Mean Change (SD)b	Difference (95% CI)	Mean (SD)b	Difference (95% CI)
Sullivan et al. (2017);93 ESSENTIAL		1.2 (0.6 to 1.9)		3.6 (2.1 to 5.1)
Endoscopic gastric plication	1.7		4.9 (7.0)
Sham	0.5		1.4 (5.6)

BMI: body mass index; CI: confidence interval; ESSENTIAL: The randomized, subject and evaluator-blinded, parallel-group, multicenter clinical trial using an endoscopic suturing device (G-CATH EZ™ suture anchor delivery catheter) for primary weight loss; RCT: randomized controlled trial; SD: standard deviation.
^a For Sullivan et al. (2017), percent total body weight loss at 12 months. 
^b At 12-month follow-up.

Study relevance, design, and conduct limitations are summarized in Tables 13 and 14.

Table 13. Study Relevance Limitations

Study	Populationa	Interventionb	Comparatorc	Outcomesd	Duration of Follow-upe
Sullivan et al. (2017);93 ESSENTIAL	4. Majority White, not Hispanic/Latino patients.	4. Low-intensity lifestyle therapy used with procedure.	2. Low-intensity lifestyle therapy used.

The study limitations stated in this table are those notable in the current review; this is not a comprehensive gaps assessment. 
^a Population key: 1. Intended use population unclear; 2. Study population is unclear; 3. Study population not representative of intended use; 4, Enrolled populations do not reflect relevant diversity; 5. Other.
^b Intervention key: 1. Not clearly defined; 2. Version used unclear; 3. Delivery not similar intensity as comparator; 4. Not the intervention of interest (e.g., proposed as an adjunct but not tested as such); 5: Other.
^c Comparator key: 1. Not clearly defined; 2. Not standard or optimal; 3. Delivery not similar intensity as intervention; 4. Not delivered effectively; 5. Other.
^d Outcomes key: 1. Key health outcomes not addressed; 2. Physiologic measures, not validated surrogates; 3. Incomplete reporting of harms; 4. Not establish and validated measurements; 5. Clinically significant difference not prespecified; 6. Clinically significant difference not supported; 7. Other.
^e Follow-Up key: 1. Not sufficient duration for benefit; 2. Not sufficient duration for harms; 3. Other.

Table 14. Study Design and Conduct Limitations

Study	Allocationa	Blindingb	Selective Reportingc	Data Completenessd	Powere	Statisticalf
Sullivan et al. (2017);93 ESSENTIAL	5. Lead-in cohort of 34 subjects was not randomized but underwent the active treatment procedure for the purposes of investigator training.	1. Evaluator-blinded only.			4. Weight loss results were lower in both the active and sham control groups than estimated in the power analysis.

The study limitations stated in this table are those notable in the current review; this is not a comprehensive gaps assessment.
^a Allocation key: 1. Participants not randomly allocated; 2. Allocation not concealed; 3. Allocation concealment unclear; 4. Inadequate control for selection bias; 5. Other.
^b Blinding key: 1. Participants or study staff not blinded; 2. Outcome assessors not blinded; 3. Outcome assessed by treating physician; 4. Other.
^c Selective Reporting key: 1. Not registered; 2. Evidence of selective reporting; 3. Evidence of selective publication; 4. Other.
^d Data Completeness key: 1. High loss to follow-up or missing data; 2. Inadequate handling of missing data; 3. High number of crossovers; 4. Inadequate handling of crossovers; 5. Inappropriate exclusions; 6. Not intent to treat analysis (per protocol for noninferiority trials); 7. Other.
^e Power key: 1. Power calculations not reported; 2. Power not calculated for primary outcome; 3. Power not based on clinically important difference; 4. Other.
^f Statistical key: 1. Analysis is not appropriate for outcome type: (a) continuous; (b) binary; (c) time to event; 2. Analysis is not appropriate for multiple observations per patient; 3. Confidence intervals and/or p values not reported; 4. Comparative treatment effects not calculated; 5. Other.

Observational Study
Pattanshetti et al. (2013) published the results of a study that described the evolution of an LAGB plication procedure, a hybrid procedure involving both LAGB and greater curvature plication developed by the authors.⁹⁴ Eighty patients were included, with a baseline mean BMI of 38.05 kg/m2. At 6, 12, 18 and 24 months postsurgery, mean percent EWL was 42.6%, 56.4%, 57.6% and 65.8%, respectively. Five postoperative complications required reoperation.

Section Summary: Laparoscopic Gastric Plication for Adults with Morbid Obesity
There is a shortage of comparative studies, especially RCTs, comparing the safety and efficacy of laparoscopic gastric plication with other bariatric surgery procedures. A 2021 systematic review demonstrated that SG is superior to greater curvature gastric plication with regard to providing effective weight loss through 24 months; statistical significance was not reached at 36 months. The difference in the improvement of comorbidities and risk of major complications or mortality did not reach statistical significance between groups. One RCT compared endoscopic gastric plication with a sham procedure, reporting 1-year follow-up results in favor of the intervention. Longer-term follow-up and additional comparative studies are needed.

Single Anastomosis Duodeno-ileal Bypass With Sleeve Gastrectomy for Adults with Morbid Obesity
Clinical Context and Therapy Purpose
The purpose of single anastomosis duodeno-ileal bypass with sleeve gastrectomy (SADI-S) is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as standard medical care, in patients who are adults with morbid obesity.

The question addressed in this evidence review is: Does SADI-S improve the net health outcome in adults who are obese?

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals who are adults with morbid obesity.

Interventions
The therapy being considered is SADI-S.

Comparators
Comparators of interest include standard medical care. Treatment for adults with morbid obesity includes physical exercise, low carbohydrate dieting, and low-fat dieting.

Outcomes
The general outcomes of interest are OS, change in disease status, functional outcomes, health status measures, quality of life, treatment-related mortality, and treatment-related morbidity.

The existing literature evaluating SADI-S as a treatment for morbid obesity has varying lengths of follow-up, ranging from 3 to 5 years. While studies described below all reported at least 1 outcome of interest, longer follow-up was necessary to fully observe outcomes. One-year follow-up is necessary to demonstrate weight loss efficacy. Longer follow-up to 5 to 10 years is desirable to assess maintenance of weight loss, impact on co-occurring conditions, and appearance of long-term complications.

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

• To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs.
• In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
• To assess longer term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
• Consistent with a "best available evidence approach," within each category of study design, studies with larger sample sizes and longer durations were sought.
• Studies with duplicative or overlapping populations were excluded.

Review of Evidence
No controlled trials of SADI-S were identified. Some case series have reported on weight loss and other clinical outcomes up to 5 years postsurgery.

Systematic Review
In 2018, Shoar et al. published a systematic review of 12 studies, comprising 5 cohorts, 4 case series, and 3 case reports, that reviewed the efficacy and safety of SADI-S.⁹⁵ The studies included 581 patients who underwent SADI-S. These patients were between 18 and 71 years of age with a BMI between 33 to 71.5 kg/m2. Of the total surgeries, 508 (87.4%) were primary and 73 (12.6%) were revisional. Follow-up was available between 6 and 60 months after the procedure. Results revealed the average percent EWL was 30% at 3 months, 55% at 6 months, 70% at 1 year, and 85% at 2 years. The comorbidity resolution rate was 74.1% for T2D, 96.3% for hypertension, 68.3% for dyslipidemia, 63.3% for OSA, and 87.5% for GERD. The most common complication was diarrhea (1.2%) and vitamin A, selenium, and iron deficiency were the most common nutritional deficiencies. There was also the possibility of protein malnutrition in up to 34% of patients when measured. The authors concluded that SADI-S was associated with a promising short-term weight loss outcome and comorbidity resolution rate; however, RCTs are warranted to compare this procedure to more commonly performed bariatric procedures.

Observational Studies
Torres et al. (2017) published a retrospective chart review of patients from their center receiving bariatric procedures, evaluating outcomes at 3-year follow-up.⁹⁶ Outcomes were evaluated separately for patients with and without diabetes. For patients without diabetes, comparisons were made among patients who underwent RYGB (n = 149) or SADI-S (n = 106). For patients with diabetes, comparisons were made among patients who underwent RYGB (n = 97), BPD-DS (n = 77), or SADI-S (n = 97). Among the patients without diabetes, significant differences favoring SADI-S over RYGB were found in: percent EWL; systolic blood pressure; total, high-density lipoprotein and low-density lipoprotein cholesterol; and insulin. Significant differences were not found in diastolic blood pressure or fasting glucose. Among the patients with T2D, remission rates using American Diabetic Association criteria were: 55%, 70%, and 76% for patients receiving RYGB, BPD-DS and SADI-S, respectively. Patients with diabetes who underwent BPD-DS or SADI-S achieved significantly lower total cholesterol and triglyceride levels compared with those undergoing RYGB after 3 years of follow-up.

Case Series
One larger series, by Sanchez-Pernaute et al. (2015), reported on 97 patients with obesity and T2D.⁹⁷ The authors reported that control of diabetes, defined as HbA1C levels less than 6.0%, was achieved by between 70% and 84% of patients at different time points. Remission rates were higher for patients on oral therapy than those on insulin and were higher in patients with a shorter duration of diabetes.

Section Summary: Single Anastomosis Duodeno-ileal Bypass With Sleeve Gastrectomy for Adults With Morbid Obesity
A systematic review of 12 observational studies concluded that SADI-S was associated with promising weight loss and comorbidity resolution. No published controlled trials have evaluated SADI-S. A comparative chart review found that patients without diabetes experienced significantly better weight loss and lipid profiles with SADI-S than with RYGB and patients who had diabetes experienced significantly higher rates of remission with SADI-S than with RYGB. Long-term safety and efficacy outcomes and comparative RCTs are still needed.

Duodenojejunal Sleeve for Adults With Morbid Obesity
Clinical Context and Therapy Purpose
The purpose of the duodenojejunal sleeve procedure is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as standard medical care, in patients who are adults with morbid obesity.

The question addressed in this evidence review is: Does the duodenojejunal sleeve procedure improve the net health outcome in adults who are obese?

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals who are adults with morbid obesity.

Interventions
The therapy being considered is the duodenojejunal sleeve procedure.

Comparators
Comparators of interest include standard medical care. Treatment for adults with morbid obesity includes physical exercise, low carbohydrate dieting, and low-fat dieting.

Outcomes
The general outcomes of interest are OS, change in disease status, functional outcomes, health status measures, quality of life, treatment-related mortality, and treatment-related morbidity.

The existing literature evaluating duodenojejunal sleeve as a treatment for morbid obesity has varying lengths of follow-up, ranging from 3 to 6 months. While studies described below all reported at least 1 outcome of interest, longer follow-up was necessary to fully observe outcomes. One-year follow-up is necessary to demonstrate weight loss efficacy. Longer follow-up to 5 to 10 years is desirable to assess maintenance of weight loss, impact on co-occurring conditions, and appearance of long-term complications.

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

• To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs.
• In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
• To assess longer term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
• Consistent with a "best available evidence approach," within each category of study design, studies with larger sample sizes and longer durations were sought.
• Studies with duplicative or overlapping populations were excluded.

Review of Evidence
Systematic Review
The EndoBarrier (GI Dynamics) is a fluoropolymer sleeve that is reversibly fixated to the duodenal bulb and extends 80 cm into the small bowel, usually terminating in the proximal jejunum. A systematic review of the effect of EndoBarrier on weight loss and diabetes control outcomes was published in 2016.98, It included 5 small RCTs (N = 235 patients; range, 18 to 77 patients), with follow-up ranging from 12 to 24 weeks. Comparators were diet and/or other lifestyle modifications, and 2 studies had sham controls. All studies were judged to be at high-risk of bias using the Cochrane risk of bias tool. Combined results demonstrated that the EndoBarrier group had 12.6% greater EWL (95% CI, 9.0 to 16.2) than medical therapy. For diabetes control outcomes, trends toward greater improvement in the EndoBarrier group were not statistically significant. MD in HbA1C level was -0.8% (95% CI, -1.8 to 0.3) and the RR of reducing or discontinuing diabetic medications was 3.28 (95% CI, 0.54 to 10.73).

Randomized Controlled Trial
The largest single trial was a multicenter RCT published in 2014; it included 77 patients with T2D and a BMI greater than 30 kg/m2.99, Patients were treated for 6 months with EndoBarrier or medical therapy. At 6 months, the EndoBarrier was removed, and patients were followed for an additional 6 months. Thirty-eight patients were randomized to the EndoBarrier group, and 31 (82%) of 38 completed 12 months of treatment. Thirty-nine patients were randomized to medical treatment, and 35 (90%) of 39 completed 12 months of treatment. At 6 months, the decrease in BMI was significantly greater in the EndoBarrier group than in the medical therapy group (3.3 kg/m2 versus 1.8 kg/m², p < .05), and at 12 months the difference in BMI was of marginal statistical significance (2.2 kg/m² versus 1.3 kg/m², p = .06), respectively. HbA1C level was significantly lower in the EndoBarrier group at 6 months (7.0% versus 7.9%, p < .05), but at 12 months the difference between groups did not differ significantly (7.3% versus 8.0%, p = .95).

Observational Study
Obermayer et al. (2021) evaluated outcomes after treatment with EndoBarrier in 10 patients with T2D and an average BMI of 43.3 kg/m2.¹⁰⁰ Results demonstrated that EndoBarrier reduced mean body weight from 121.2 ± 18.5 kg to 116.3 ± 18.2 kg (p = .006) 4 weeks after the start of therapy, and to 115.1 ± 21.4 kg (p = .075 vs. baseline) until explantation of the device after 36 weeks. There was an increase in weight to 117.2 ± 20.8 kg (p = 0.117 vs. baseline) 24 weeks after explanation.

Section Summary: Duodenojejunal Sleeve for Adults With Morbid Obesity
A systematic review of evidence on a duodenojejunal sleeve included 5 RCTs and found significantly greater short-term weight loss (12 to 24 weeks) with duodenojejunal sleeves compared with medical therapy. There was no significant difference in symptom reduction associated with diabetes. However, all RCTs had small sample sizes and were judged by the systematic reviewers to be at high-risk of bias.

Intragastric Balloon Devices for Adults With Morbid Obesity
Clinical Context and Therapy Purpose
The purpose of IGB devices is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as standard medical care, in patients who are adults with morbid obesity.

The question addressed in this evidence review is: Do IGB devices improve the net health outcome in adults who are obese?

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals who are adults with morbid obesity.

Interventions
The therapy being considered is IGB devices. IGB devices are placed in the stomach via endoscope or swallowing to act as space-occupying devices to induce satiety. As of 2017, 2 IGB devices have U.S. Food and Drug Administration (FDA) approval, each designed to stay in the stomach for no more than 6 months. Obalon is a swallowable 3-balloon system. The OBERA Intragastric Balloon System (previously marketed outside of the United States as BioEnterics) is a saline-inflated silicone single balloon system.

Comparators
Comparators of interest include standard medical care. Treatment for adults with morbid obesity includes physical exercise, low carbohydrate dieting, and low-fat dieting.

Outcomes
The general outcomes of interest are OS, change in disease status, functional outcomes, health status measures, quality of life, treatment-related mortality, and treatment-related morbidity.

The existing literature evaluating IGB devices as a treatment for morbid obesity has varying lengths of follow-up, ranging from 5 to 10 years. While studies described below all reported at least 1 outcome of interest, longer follow-up was necessary to fully observe outcomes. One-year follow-up is necessary to demonstrate weight loss efficacy. Longer follow-up of 5 to 10 years is desirable to assess maintenance of weight loss, impact on co-occurring conditions, and appearance of long-term complications.

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

• To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs.
• In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
• To assess longer term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
• Consistent with a "best available evidence approach," within each category of study design, studies with larger sample sizes and longer durations were sought.
• Studies with duplicative or overlapping populations were excluded.

Review of Evidence
Systematic Reviews
Several systematic reviews of RCTs evaluating IGB devices for the treatment of obesity have been published; none was limited to FDA-approved devices.^{101,102,103,104}

Kotinda et al. (2020) published a systematic review and meta-analysis that evaluated the efficacy of IGB devices in comparison to sham or lifestyle interventions in overweight and obese adults.¹⁰⁴ Thirteen RCTs with 1523 patients were included. Results revealed that the mean percent EWL difference between the IGB and control groups was 17.98% (95% CI, 8.37 to 27.58; p < .00001), significantly favoring IGB. IGB was also significantly favored when evaluating the mean percent TWL difference between the groups: 4.40% (95% CI, 1.37 to 7.43; p < .00001). Similarly, the difference in actual weight loss and BMI loss was 6.12 kg and 2.13 kg/m², respectively. Overall, IGB was found to be more effective than lifestyle intervention alone for weight loss; however, the majority of included RCTs used 1 fluid-filled IGB and there was significant heterogeneity between the included studies.

The systematic review by Tate et al. (2017) focused on RCTs, published between 2006 and 2016.¹⁰⁵ Additional inclusion criteria were: sham, lifestyle modification, or pharmacologic agent as a comparator; at least 1 outcome of body weight change; and study duration of 3 or more months. Eight RCTs were included in the review, with 4 contributing to the meta-analysis. The meta-analysis included 777 patients and showed a significant improvement in percent TBWL with IGB compared with control (5.5%; 95% CI, 4.3 to 6.8). However, there was significant heterogeneity among the trials (I2 = 62%), so interpretation of results is limited. The percent TBWL with IGB is lower than expected with RYGB (reported 27%) or with the most efficacious pharmacologic agent (reported 9%).

Saber et al. (2017) identified 20 RCTs reporting weight loss outcomes after IGB implantation or a non-IGB control intervention.¹⁰¹ IGB was compared with sham in 15 trials, behavioral modification in 4 trials, and pharmacotherapy in 1 trial. In 17 trials, patients received LT in addition to other interventions. Studies were published between 1987 and 2015 and sample sizes varied from 21 to 326 participants. Outcomes were reported between 3 and 6 months. In a meta-analysis of 7 RCTs reporting BMI loss as an outcome, there was a significantly greater BMI loss in the IGB group than in the control group (mean effect size [ES], 1.59 kg/m2; 95% CI, -0.84 to 4.03; p < .001). Findings on other outcomes were similar. A meta-analysis of 4 studies reporting percent EWL favored the IGB group (ES = 14.25%; 95% CI, 2.09 to 26.4; p = .02). Also, a meta-analysis of 6 studies reporting absolute weight loss favored the IGB group (ES = 4.6 kg; 95% CI, 1.6 to 7.6; p = .003).

Although the review was not limited to FDA approved devices, older devices were air-filled and newer devices, including the 2 approved by FDA in 2015, are fluid-filled. Sufficient data were available to conduct a sensitivity analysis of 3-month efficacy data. A meta-analysis of 4 studies did not find a significant difference in weight loss with air-filled IGB devices or a control intervention at 3 months (ES = 0.26; 95% CI, -0.12 to 0.64; p = .19). In contrast, a meta-analysis of 8 studies of fluid-filled devices found significantly better outcomes with the IGB than with control (ES = 0.25; 95% CI, 0.05 to 045; p = .02).

Randomized Controlled Trials
Pivotal trials on FDA-approved devices have been published.

Courcoulas et al. (2017) published a multicenter, pivotal RCT evaluating the Obera IGB in the United States (as noted, the device has been used in other countries).¹⁰⁶ A total of 317 patients were randomized and initiated 6 months of treatment with an IGB plus LT (n = 137) or LT only (n = 136). Patients were followed for an additional 6 months. Key eligibility criteria were age 18 to 65 years, baseline BMI between 30 and 40 kg/m2, a history of obesity for at least 2 years, and having failed previous weight loss attempts. Nineteen patients in the IGB group and 121 in the control group completed the 6-month treatment period.

Coprimary effectiveness outcomes, assessed at 9 months, were mean percent EWL and difference in mean weight loss. Mean percent EWL at 9 months was 26.4% in the IGB group and 10.1% in the control group (difference, 16.2%; 95% CI, 12.3 to 20.2; p < .001). Mean weight loss at 9 months was -8.8 kg (-19.4 lb) in the IGB group and -3.2 kg (-7.1 lb) in the control group (p < .001). There were also significant between-group differences in mean weight loss and mean percent EWL at 6 and 12 months.

Most adverse events in the Obera pivotal trial were anticipated accommodative symptoms. A total of 139 (87%) patients reported nausea, 121 (76%) reported vomiting, and 92 (58%) reported abdominal pain. Fewer than 5% of these adverse events were serious; most were mild or moderate. Thirty patients in the device group had the IGB removed before month 6 because of an adverse event (n = 15) or patient request (n = 15). There were no deaths and 9 serious adverse events unrelated to device accommodation; among others, these included a case of gastric outlet obstruction and a case of gastric perforation with sepsis.

The Courcoulas et al. (2017) pivotal trial was not blinded or sham-controlled; however, a double-blind sham-controlled randomized trial evaluating the BioEnterics IGB (now called the Obera device) was published by Genco et al. (2006).¹⁰⁷ This crossover trial included 32 obese patients ages 25 to 50 years with a mean BMI of 47.3 kg/m². Patients received, in random order, 3 months of an IGB and 3 months of sham. (Both groups underwent upper gastrointestinal endoscopy, but no device was placed in the sham group.) Patients who initially received the IGB had a mean BMI reduction of 5.8 kg/m² after 3 months; after crossover to sham, these patients had a mean additional BMI reduction of 1.1 kg/m². Patients initially in the sham group had an initial mean BMI reduction of 0.4 kg/m2; after crossover to an active device, these patients had a mean BMI reduction of 5.1 kg/m². The between-group difference in BMI reductions was statistically significant (p < .001). Findings on other outcomes (mean percent EWL, mean weight loss) were similar.

Case Series
A case series of patients treated with an IGB with up to 60-month follow-up was published by Kotzampassi et al. (2012),¹⁰⁸ A total of 500 patients were treated with the BioEnterics IGB. Twenty-six patients did not complete the initial 6 months of treatment, and another 77 patients did not comply with dietary restrictions and did not have satisfactory weight loss at 6 months. Among 352 patients with data available, BMI was 44.5 kg/m² at baseline, 35.7 kg/m² at device removal, 38.8 kg/m² 12 months after device removal, and 40.1 kg/m² 24 months after device removal. Mean percent EWL was 43.9% at device removal, 27.7% 12 months after device removal, and 17% 24 months after device removal. Among the 195 patients with available 5-year data, mean baseline BMI was 43.3 kg/m², mean BMI at device removal was 33.8 kg/m², and mean BMI at 5 years was 40.1 kg/m². Mean percent EWL at 5 years was 13.0%. Overall, patients who initially complied with 6 months of IGB device use and lost weight slowly gained weight over time but weighed less at final follow-up than at baseline.

Section Summary: Intragastric Balloon Devices for Adults With Morbid Obesity
Evidence includes RCTs, a case series with long-term follow-up on 1 of the devices, and systematic reviews on various IGB devices. RCTs have found significantly better weight loss outcomes with IGB devices compared with sham treatment or LT alone. One RCT followed patients for an additional 6 months after IGB removal and found sustained weight loss. A large case series with follow-up up to 5 years has suggested that patients regain weight over time. Additional long-term follow-up data are needed. There are some adverse events, and in a minority of cases, these adverse events can be severe. The FDA wrote 2 letters in 2017 to health care providers, 1 warning of spontaneous balloon inflation and pancreatitis and the other reporting 5 unanticipated deaths occurring in 2016 to 2017 following the IGB procedure. In June 2018, the FDA reported that, since 2016, a total of 12 deaths occurred in patients with liquid-filled intragastric balloons worldwide; 7 of these deaths were in patients in the U.S. Health care providers are encouraged to monitor patients receiving IGBs.

Aspiration Therapy Device for Adults With Morbid Obesity
Clinical Context and Therapy Purpose
The purpose of the aspiration therapy (AT) device is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as standard medical care, in patients who are adults with morbid obesity.

The question addressed in this evidence review is: Does the AT device improve the net health outcome in adults who are obese?

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals who are adults with morbid obesity.

Interventions
The therapy being considered is the AT device.

Comparators
Comparators of interest include standard medical care. Treatment for adults with morbid obesity includes physical exercise, low carbohydrate dieting, and low-fat dieting.

Outcomes
The general outcomes of interest are OS, change in disease status, functional outcomes, health status measures, quality of life, treatment-related mortality, and treatment-related morbidity.

The existing literature evaluating AT device as a treatment for morbid obesity has varying lengths of follow-up, ranging from 1 to 2 years. While studies described below all reported at least 1 outcome of interest, longer follow-up was necessary to fully observe outcomes. Therefore, 1 to 2 years of follow-up is considered necessary to demonstrate efficacy.

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

• To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs.
• In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
• To assess longer term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
• Consistent with a "best available evidence approach," within each category of study design, studies with larger sample sizes and longer durations were sought.
• Studies with duplicative or overlapping populations were excluded.

Review of Evidence
Randomized Controlled Trials
AT involves an FDA-approved device (AspireAssist) that allows patients to drain a portion of the stomach contents after meals via an implanted tube connected to an external skin port. One RCT has been published. The Pivotal Aspiration Therapy with Adjusted Lifestyle (PATHWAY) trial, by Thompson et al. (2017), randomized 207 participants to 52 weeks of AspireAssist therapy plus lifestyle counseling (n = 127) or lifestyle counseling alone (n = 70).¹⁰⁹ Participants were between 21 and 65 years of age, with a BMI ranging from 35 to 55 kg/m². Coprimary outcomes were mean EWL at 52 weeks and the proportion of patients with 25% or more EWL at 52 weeks. Investigators did a modified intention-to-treat analysis including all patients in the AspireAssist group who attempted tube placement (n = 111) and all patients in the lifestyle counseling group who attended at least 1 therapy session (n = 60). Mean EWL at 52 weeks was 31.5% in the AspireAssist group and 9.8% in the lifestyle counseling group. The difference between groups was 21.7% (95% CI, 15.3 to 28.1), which was greater than the 10% difference needed to meet the a priori definition of success. The proportion of patients with 25% or more EWL at 52 weeks was 58.6% in the AspireAssist group and 22% in the lifestyle counseling group (p < .001). Bulimia or binge eating disorder were exclusion criteria and, during the study, there was no evidence that patients developed bulimia or that devices were overused (i.e., used > 3 times a day). Most of the adverse events (83.8%) in the AspireAssist group were associated with placement of a percutaneous endoscopic gastric tube. All 5 serious adverse events occurred in the AspireAssist group (mild peritonitis, severe abdominal pain and a case of product malfunction). Product malfunction was related to malfunction of the A-tube, typically occurring within the first week of implantation and seen in 90% of adverse events seen with the AspireAssist. The durability of a treatment effect beyond 1 year was not reported.

Thompson et al. (2019) published 4-year outcomes from the PATHWAY trial.¹¹⁰ AT patients were permitted to continue the study beyond 1 year up to a maximum of 5 years provided they maintained at least 10% TWL from baseline at each year end. Out of 82 AT patients who completed year 1, 58 continued in the next phase, 43 completed year 2, 22 completed year 3, and 15 completed year 4 in the trial. Of 58 AT participants continuing in the study, 43 withdrew before completion of year 4, with 25/43 meeting their weight loss goal or losing > 10% of their baseline weight. Forty of 58 patients (69%) achieved at least 10% TWL at 4 years or at time of study withdrawal. Out of 60 patients treated in the LT control group, only 31 completed the full initial study year. Two serious adverse events were reported in years 2 to 4. One patient developed a secondary fistula superior to the A-tube fistula, which resolved following A-tube removal. The second patient experienced an A-tube malfunction, which was replaced. A total of 57 adverse events, including the 2 serious adverse events, were recorded. The adverse events with the greatest frequency were peristomal irritation (12 events), persistent fistulas (12 events), and peristomal granulation tissue (8 events). A total of 27 A-tubes required replacement over the 4 years of the study. Reasons for replacement include tube defects (~ 50%) and tube leaks (~ 30%). According to the study survival analysis, one can expect 50% of A-tubes to be replaced within approximately 3.5 years postgastrostomy. No clinically significant metabolic disorders were observed. No evidence for the development of any eating disorders was noted. Study results are summarized in Table 15. Study relevance, design and conduct limitations are summarized in Tables 16 and 17.

Table 15. Results of PATHWAY Trial

		> 25% EWL1	% TWL	ΔHbA1C2	IW Quality of Life Total Score2,3
Thompson et al. (2017); PATHWAY109	Year 1, n	% [95% CI]	% (SD) [95% CI]	% (SD)	Mean (SD)
AT	mITT: 111 PP: 82	mITT: 56.8 [49.0 to 64.5] PP: 68.3 [NR]	mITT: 12.1 (9.6) [NR] PP: 14.2 (9.8) [12.1 to 16.4]	mITT: -0.36 (0.45) PP: NR	mITT: 6.2 (13.4) PP: NR
LT	mITT: 60 PP: 31	mITT: 22.0 [15.3 to 28.1] PP: 25.8 [NR]	mITT: 3.6 (6.0) [NR] PP: 4.9 (7.0) [NR]	mITT: -0.22 (0.27) PP: NR	mITT: 3.3 (10.0) PP: NR
Mean Difference [95% CI]		NR	8.6 [6.2 to 10.9]2	-1.4 [-0.28 to 0.0]2	2.9 (SD: 12.5)2
P-Value		mITT: < .001 PP: < .001	NR	P = .052	P = .034
Thompson et al. (2019); PATHWAY110	AT5	> 25% EWL1		% TWL
		% (SD)		% (SD) [95% CI]
Year 1	82	68.3 (NR)		14.2 (9.8) [12.1 to 16.4]
Year 2	43	72.1 (NR)		15.3 (8.8) [12.6 to 18.0]
Year 3	22	63.6 (NR)		16.6 (10.5) [12.0 to 21.3]
Year 4	15	73.3 (NR)		18.7 (11.7) [12.2 to 25.2]

AT: aspiration therapy; CI: confidence interval; EWL: excess weight loss; HbA1c: hemoglobin A1c; IWquality of life: Impact of Weight of Quality of Life survey; LT: lifestyle therapy; mITT: modified intent-to-treat; NR: not reported; PATHWAY: 
Pivotal Aspiration Therapy with Adjusted Lifestyle; PP: per protocol; SD: standard deviation; TWL: total weight loss. 
¹ Primary outcome measure. 
² Based on the modified intent-to-treat analysis. 
³ Improvement in quality-of-life measures is reflected by increasing IWquality of life scores. 
⁴ Treatment differences in individual IWquality of life component scores did not reach statistical significance.
⁵ Based on the per-protocol analysis.

Table 16. Study Relevance

Study; Trial	Populationa	Interventionb	Comparatorc	Outcomesd	Follow-Upe
Thompson et al. (2017; 2019); PATHWAY109,110	4. The majority of enrolled patients were white, non-Hispanic.		2: No active comparator for years 2 to 4.

The study limitations stated in this table are those notable in the current review; this is not a comprehensive gaps assessment.
PATHWAY: Pivotal Aspiration Therapy with Adjusted Lifestyle.
^a Population key: 1. Intended use population unclear; 2. Study population is unclear; 3. Study population not representative of intended use; 4. Enrolled populations do not reflect relevant diversity; 5. Other.
^b Intervention key: 1. Not clearly defined; 2. Version used unclear; 3. Delivery not similar intensity as comparator; 4.Not the intervention of interest.
^c Comparator key: 1. Not clearly defined; 2. Not standard or optimal; 3. Delivery not similar intensity as intervention; 4. Not delivered effectively.
^d Outcomes key: 1. Key health outcomes not addressed; 2. Physiologic measures, not validated surrogates; 3. No CONSORT reporting of harms; 4. Not establish and validated measurements; 5. Clinical significant difference not prespecified; 6. Clinical significant difference not supported.
^e Follow-Up key: 1. Not sufficient duration for benefit; 2. Not sufficient duration for harms.

Table 17. Study Design and Conduct Limitations

Study	Allocationa	Blindingb	Selective Reportingc	Data Completenessd	Powere	Statisticalf
Thompson et al. (2019); PATHWAY109,110		2: Blinding to outcome assessment unclear. 3: Blinding and identity of outcome assessors unclear.		1: High loss to follow-up or missing data. High loss to pre- and post-enrollment withdrawals. 2: Multiple strategies utilized for handling of missing data. 5: Inappropriate exclusion of patients with TWL < 10% during years 2 – 4 from analysis. 6: Modified intent to treat analysis not carried through.	1: Study not powered beyond 1 y of follow-up. Study underpowered for completers at 1 y. 3: Rationale for clinically important difference not provided.	2: Not all sensitivity analyses are statistically significant for primary effectiveness outcome (at least 50% of participants achieving at least 25% EWL); unclear if analysis is appropriate for multiple observations per patient or extent of missing data.

The study limitations stated in this table are those notable in the current review; this is not a comprehensive gaps assessment.
EWL: excess weight loss; PATHWAY: Pivotal Aspiration Therapy with Adjusted Lifestyle; TWL: total weight loss.
^a Allocation key: 1. Participants not randomly allocated; 2. Allocation not concealed; 3. Allocation concealment unclear; 4. Inadequate control for selection bias.
^b Blinding key: 1. Not blinded to treatment assignment; 2. Not blinded outcome assessment; 3. Outcome assessed by treating physician.
^c Selective Reporting key: 1. Not registered; 2. Evidence of selective reporting; 3. Evidence of selective publication.
^d Data Completeness key: 1. High loss to follow-up or missing data; 2. Inadequate handling of missing data; 3. High number of crossovers; 4. Inadequate handling of crossovers; 5. Inappropriate exclusions; 6. No intent to treat analysis (per protocol for noninferiority trials).
^e Power key: 1. Power calculations not reported; 2. Power not calculated for primary outcome; 3. Power not based on clinically important difference.
^f Statistical key: 1. Analysis is not appropriate for outcome type: (a) continuous; (b) binary; (c) time to event; 2. Analysis is not appropriate for multiple observations per patient; 3. Confidence intervals and/or p values not reported; 4. Comparative treatment effects not calculated.

Case Series
In addition to the RCT, a case series by Noren and Forssell (2016) evaluated AspireAssist use by 25 obese patients.¹¹¹ Patients had 1 year of AT and also participated in a cognitive-behavioral therapy weight loss program for the initial 3 months. Patients were instructed to aspirate 3 times a day after meals. Twenty (80%) patients completed the 1-year intervention period. Mean baseline weight was 107.4 kg. In a per-protocol analysis, the mean EWL was 54.5% at 12 months. Data on 15 (60%) patients were available at 24 months; mean EWL was 61.5%.

Section Summary: Aspiration Therapy Device for Adults With Morbid Obesity
The evidence consists of an RCT with 4 years of follow-up and a small case series with up to 2 years of follow-up. The RCT found significantly greater weight loss (measured several ways) with AT compared with LT at 1 year. Forty of 58 patients (69%) achieved at least 10% TWL at 4 years or at time of study withdrawal; however, only 15/111 initial AT patients completed the study through 4 years. In addition to a high degree of missing data, the PATHWAY study noted a potentially high degree of adverse events related to A-tube malfunction, an element of the therapy which is expected to require replacement within approximately 3.5 years postgastrostomy in 50% of cases. The impact of this on health outcomes compared to existing surgical approaches is unknown. The case series followed only 15 patients more than 1 year; at 2 years, study completers had not regained weight and instead had lost additional excess weight. The total amount of data on AT remains limited and additional studies need to be conducted before conclusions can be drawn about the long-term effects of treatment on weight loss, metabolism, safety and nutrition.

Revision Bariatric Surgery for Adults With Morbid Obesity Who Failed Bariatric Surgery
Clinical Context and Therapy Purpose
The purpose of revision bariatric surgery is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as standard medical care, in patients who are adults with morbid obesity and who failed bariatric surgery

The question addressed in this evidence review is: Does revision bariatric surgery improve the net health outcome in adults who are obese?

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals who are adults with morbid obesity and failed bariatric surgery.

Interventions
The therapy being considered is revision bariatric surgery.

Comparators
Comparators of interest include standard medical care. Treatment for adults with morbid obesity includes physical exercise, low carbohydrate dieting, and low-fat dieting.

Outcomes
The general outcomes of interest are OS, change in disease status, functional outcomes, health status measures, quality of life, treatment-related mortality, and treatment-related morbidity.

The existing literature evaluating revision bariatric surgery as a treatment for morbid obesity has varying lengths of follow-up, ranging from 1 to 3 years. While studies described below all reported at least 1 outcome of interest, longer follow-up was necessary to fully observe outcomes. One-year follow-up is necessary to demonstrate weight loss efficacy. Longer follow-up of 5 to 10 years is desirable to assess maintenance of weight loss, impact on co-occurring conditions, and appearance of long-term complications.

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

• To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs.
• In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
• To assess longer term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
• Consistent with a "best available evidence approach," within each category of study design, studies with larger sample sizes and longer durations were sought.
• Studies with duplicative or overlapping populations were excluded.

Review of Evidence
Systematic Reviews
Matar et al. (2021) published a systematic review of 556 patients (n = 17 studies) who underwent RYGB for SG-related complications, including GERD (30.4% cases) and insufficient weight loss and weight regain (52% of cases).¹¹² The mean BMI at the time of conversion ranged from 33.3 to 48.3 kg/m². The pooled baseline BMI at conversion was 38.5 kg/m² (95% CI, 36.49 to 40.6), at 6 months was down to 28.6 kg/m² (95% CI, 16.1 to 41.0), and after 1 year was up to 32.1 kg/m² (95% CI, 25.50 to 38.7). The pooled mean %TWL after completion of treatment was 25.2% (95% CI, 12.8 to 37.5) at 6 months and 22.8% (95% CI, 13.5 to 32.1) at 1 year. There was a 16.4% complication rate at 30 days, which decreased to 11.4% after 30 days. At 1-year post RYGB, the rate of resolution for common comorbidities was as follows: GERD, 79.7% (95% CI, 59.6 to 91.3); T2D, 57.7% (95% CI, 36.9 to 76.1); hypertension, 49.4% (95% CI, 25.8 to 73.3).

Parmar et al. (2020) published a systematic review of 1,075 patients (n = 17 studies) who underwent one anastomosis/mini gastric bypass (OABG-MGB) as a revisional bariatric procedure after failure of a primary LAGB and SG.¹¹³ No RCTs were available on this topic and no meta-analyses were performed as part of this systematic review. The most commonly reported reason for revisional surgery was poor response (81%) followed by gastric band failure (35.9%), GERD (13.9%), intolerance (12.8%), staple line disruption (16.5%), pouch dilatation (17.9%), and stomal stenosis (10.3%). Results revealed that after the revisional OABG-MGB, the mean percent EWL was 50.8% at 6 months, 65.2% at 1 year, 68.5% at 2 years, and 71.6% at 5 years. Resolution of comorbidities after OAGB-MGB was significant with 80.5% of patients with T2D, 63.7% of patients with hypertension, and 79.4% of patients with GERD reporting resolution. The overall readmission rate following OAGB-MGB was 4.73%, the mortality rate was 0.3%, and the leak rate was 1.54%. Although the authors concluded that OAGB-MGB is a safe and effective choice for revisional bariatric surgery, RCTs on this topic are needed as currently only retrospective cohort studies with heterogenous data are available.

Brethauer et al. (2014) conducted a systematic review of reoperations after primary bariatric surgery for the American Society for Metabolic and Bariatric Surgery that included 175 studies, most of which were single-center retrospective reviews.¹¹⁴ The review is primarily descriptive, but made the following conclusions:

“The current evidence regarding reoperative bariatric surgery includes a diverse group of patient populations and procedures. The majority of the studies are single institution case series reporting short- and medium-term outcomes after reoperative procedures. The reported outcomes after reoperative bariatric surgery are generally favorable and demonstrate that additional weight loss and co-morbidity reduction is achieved with additional therapy. The risks of reoperative bariatric surgery are higher than with primary bariatric surgery and the evidence highlights the need for careful patient selection and surgeon expertise.”

Nonrandomized Studies
Petrucciani et al. (2021) published a retrospective analysis of 215 patients who underwent revisional OAGB with a biliopancreatic limb of 150 cm after failing LAGB at a single center between 2010 and 2016.¹¹⁵ The indication for surgery was weight loss failure in 30.7% of cases and long-term complications in the remaining cases. The mean BMI at the time of OAGB was 42 kg/m². At 2 years after OAGB, 9.7% of patients were lost to follow-up, BMI was down to 28 ± 5.5 kg/m², %EWL was 88.2 ± 23.9, and %TWL was 38.7 ± 9.3. At 5 years after OAGB, 16.6% of patients were lost to follow-up, BMI was slightly up to 29.2 ± 5.8 kg/m², %EWL was 82.4 ± 25, and %TWL was 36.1 ± 10. Overall postoperative morbidity was 13.5% with a 5.9% rate of postoperative abscess with or without staple line leak. Treatment-resistant GERD occurred in 21.3% of patients; conversion to RYGB was required in 4.2% of cases.

Almalki et al. (2018) published a retrospective analysis of patients diagnosed with failed restrictive procedure who underwent revision bariatric surgery.¹¹⁶ One hundred sixteen patients between 2001 and 2015 had revision RY gastric bypass (R-RYGB; n = 35) or revision single anastomosis- (mini-) gastric bypass (R-SAGB; n = 81); the primary indications for revisional procedures were weight regain (50.9%), inadequate weight loss (31%), and intolerance (18.1%). Major complications occurred in 12 (10%) patients without significant difference between groups (R-SAGB, n = 9; R-RYGB, n = 3). At 1 year after revision surgery, the R-SAGB group (76.8% EWL) showed better weight loss than R-RYGB (32.9% EWL; p = .001). In the 37.1% of patients available for follow-up at 5 years, R-SAGB had significantly lower hemoglobin levels than R-RYGB (8.2 ± 3.2 g/dl versus 12.8 ± 0.5 g/dl; p = .03). The study was limited by its retrospective nature, relatively short follow-up time, and lack of consideration of data related to patient compliance.

Sudan et al. (2015) reported on safety and efficacy outcomes for reoperative bariatric surgeries using data from a national registry, the Bariatric Outcomes Longitudinal Database.117, The Bariatric Outcomes Longitudinal Database was a large, multi-institutional bariatric surgery-specific database to which data were submitted from 2007 through 2012 by 1029 surgeons and 709 hospitals participating in the Bariatric Surgery Centers of Excellence program. Surgeries were classified as primary or reoperative bariatric. Reoperations were further divided into corrective surgeries (when complications or incomplete treatment effect of a previous bariatric operation was addressed, but the initial operation was not changed) or conversions (when an index bariatric operation was changed to a different type of bariatric operation or a reversal restored original anatomy.) Of 449,473 bariatric operations in the database, 420,753 (93.6%) operations had no further reoperations (primary operations) while 28,270 (6.3%) underwent reoperations. Of the reoperations, 19,970 (69.5%) were corrective and 8,750 (30.5%) were conversions. The primary bariatric operations were RYGB (n = 204,705 [49.1%]), LAGB (n = 153,142 [36.5%]), SG (n = 42,178 [10%]), and BPD-DS (n = 4,260 [1%]), with the rest classified as miscellaneous. LAGB was the most common primary surgery among conversions (57.5% of conversions; most often [63.5%] to RYGB). Compared with primary operations, mean hospital length of stay was longer for corrections (2.04 days versus 1.8 days, p < .001) and for conversions (2.86 days versus 1.8 days, p < .001). Mean percent EWL at 1 year was 43.5% after primary operation, 39.3% after conversions, and 35.9% after corrective operations (statistical comparison not reported). One-year mortality was higher for conversions (0.31%) than for primary surgeries (0.17%; p < .001), with no statistically significant difference for corrections (0.24%) compared with primary surgeries (0.17%; p = NS). One-year serious adverse event rates were higher for conversions (3.61%) than for primary operations (1.87%; p < .001), with no statistically significant difference for corrections (1.9%) compared with primary operations (1.87%; p = NS). The authors concluded that reoperation after primary bariatric surgery is relatively uncommon, but generally safe and efficacious when it occurs.

Endoscopic Revision Procedures
While bariatric surgery revision or correction can be conducted using standard surgical approaches, novel endoscopic procedures are being developed. Some procedures use devices also being evaluated for the endoscopic treatment of GERD (see evidence review 2.01.38). The published data on the use of these devices for treatment of regained weight is limited. Published case series have reported results using a number of devices and procedures (including sclerosing injections) as a treatment for this condition. The largest series (2007) found involved 28 patients treated with a sclerosing agent (sodium morrhuate).¹¹⁸ Reported trials that used 1 of the suturing devices had fewer than 10 patients. For example, Herron et al. (2008) reported on a feasibility study in animals.¹¹⁹ Thompson et al. (2006) reported on a pilot study with changes in anastomotic diameter and weight loss in 8 patients who regained weight and had dilated gastrojejunal anastomoses after RYGB.¹²⁰ No comparative trials were identified; comparative trials are important because of the known association between an intervention and short-term weight loss.

The StomaphyX device, which has been used in this approach, was cleared by FDA through the 510(k) process. It was determined to be equivalent to the EndoCinch system, which has 510(k) marketing clearance for endoscopic suturing for gastrointestinal tract surgery. Eid et al. (2014) reported on results from a single-center RCT that compared the StomaphyX device with a sham procedure for revisions in patients with prior weight loss after RYGB at least 2 years earlier.¹²¹ Enrollment was initially planned for 120 patients, but the trial was stopped prematurely after 1-year follow-up was completed by 45 patients in the StomaphyX group and 29 patients in the sham control group because preliminary analysis failed to achieve the primary efficacy endpoint in at least 50% of StomaphyX patients. The primary 12-month efficacy endpoint (reduction in pre-RYGB excess weight by ≥ 15%, excess BMI loss, and BMI < 35 kg/m²) was achieved by 10 (22.2%) of 45 in the StomaphyX group and 1 (3.4%) of 29 in the sham control group (p < .01).

A 2009 survey of American Society for Metabolic and Bariatric Surgery members (bariatric surgeons) indicated different risk tolerance and weight loss expectations for primary and revisional endoscopic procedures.⁶³ The surgeons were “willing to accept less weight loss and more risk for revisional endoluminal procedures than for primary endoluminal procedures.” The durability of the procedures was a concern, and most surgeons were unwilling to consider the procedures until their efficacy has been proven. A 2013 systematic review of studies reporting outcomes after endoluminal revision of primary bariatric surgery conducted by the American Society for Metabolic and Bariatric Surgery concluded: “The literature review shows the procedures on the whole to be well tolerated with limited efficacy. The majority of the literature is limited to small case series. Most of the reviewed devices are no longer commercially available."¹²²

Cohen et al. (2019) conducted a systematic review evaluating the safety and efficacy of endoscopic gastroplasty for medically uncontrolled obesity.¹²³ Nine observational studies and a single RCT were identified by the authors. Follow-up duration in the majority of studies was limited to 6 to 12 months with several studies reporting high rates of loss to follow-up. Percent total body weight loss ranged from -15.1% to 19.5%. Reduction in BMI ranged from -1.69 to -7.5 kg/m². Serious adverse events ranged from 2% to 10%. The quality of the current evidence was graded very low to moderate, with limited long-term data on weight loss durability and procedure safety.

Section Summary: Revision Bariatric Surgery for Adults With Morbid Obesity Who Failed Bariatric Surgery
For surgical revision of bariatric surgery after failed treatment, evidence from systematic reviews and nonrandomized studies suggests that revisions are associated with improvements in weight similar to those seen in primary surgery. However, the published scientific literature on the use of endoscopic devices and procedures in patients who regain weight after bariatric surgery is very limited.

Bariatric Surgery as a Treatment for Type 2 Diabetes for Adults With Diabetes Who Are Not Morbidly Obese
Clinical Context and Therapy Purpose
The purpose of gastric bypass, SG, BPD, and adjustable gastric banding is to provide treatment options that are alternatives to or improvements on existing therapies, such as standard medical care, in patients who are diabetic and not morbidly obese.

The question addressed in this evidence review is: Do various bariatric surgery procedures improve the net health outcome in those with diabetes who are not obese?

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals who are diabetic and not morbidly obese.

Interventions
The therapy being considered is gastric bypass, SG, BPD, and adjustable gastric banding. Current indications for bariatric surgery view poorly or uncontrolled diabetes as a comorbidity whose presence supports the need for surgery in patients with a BMI of 35 to 40 kg/m². There also is growing interest in gastrointestinal surgery to treat patients with T2D in patients with lower BMI.

Comparators
Comparators of interest include standard medical care. Treatment for patients who are diabetic includes blood sugar regulation and insulin therapy.

Outcomes
The general outcomes of interest are OS, change in disease status, functional outcomes, health status measures, quality of life, treatment-related mortality, and treatment-related morbidity.

The existing literature evaluating gastric bypass, SG, BPD, and adjustable gastric banding as a treatment for diabetes has varying lengths of follow-up, ranging from 1 to 5 years.

While studies described below all reported at least 1 outcome of interest, longer follow-up was necessary to fully observe outcomes. One-year follow-up is necessary to demonstrate weight loss efficacy. Longer follow-up to 5 to 10 years is desirable to assess maintenance of weight loss, impact on co-occurring conditions, and appearance of long-term complications.

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

• To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs.
• In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
• To assess longer term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
• Consistent with a "best available evidence approach," within each category of study design, studies with larger sample sizes and longer durations were sought.
• Studies with duplicative or overlapping populations were excluded.

This section focuses on RCTs and systematic reviews of RCTs comparing bariatric surgery with medical therapy.

Review of Evidence
Type 2 Diabetes and Body Mass Index 30 to 34.9 kg/m²
Systematic Reviews
Wu et al. (2016) published a meta-analysis of studies comparing bariatric surgery with nonsurgical interventions for patients who had T2D.¹²⁴ Eight RCTs with 619 patients were included. RCTs addressed RYGB (6 studies), LAGB (3 studies), LSG (1 study), and BPD (1 study). Mean BMI across studies was 29 kg/m² or higher; in 6 of 8 studies, mean BMI was 35 kg/m² or higher. One study had a 5-year follow-up, and the others had 1 to 3 years of follow-up. The study with a 5-year follow-up, by Mingrone et al. (2015), was limited to patients with a BMI of at least 35 kg/m².¹²⁵ All 8 studies reported remission of T2D as an efficacy endpoint. A pooled analysis found a significantly higher rate of T2D remission in the bariatric surgery versus the nonsurgical treatment group (RR, 5.76; 95% CI, 3.15 to 10.55; p < .001). Another diabetes-related outcome (mean reduction in HbA1C levels) was significantly greater after bariatric surgery than nonsurgical treatment (MD, -1.29; 95% CI, -1.70 to -0.87). Also, there was a significantly greater reduction in BMI with bariatric surgery than with nonsurgical treatment (MD, -5.80; 95% CI, -6.95 to -4.64; p < .001).

Since the publication of the Wu et al. (2016) meta-analysis, 5-year follow-up has been reported for the Schauer et al. (2017) RCT, which is shown in Table 18. When the Wu et al. (2016) meta-analysis was published, only 3-year findings of the Schauer etal (2017) study were available. The study included patients with T2D who had a BMI of 27 to 43 kg/m². The RCTs evaluating bariatric surgery in patients with T2D, including the 5-year follow-up of the Schauer et al. (2017) study, are summarized in Table 18.

Muller-Stich et al. (2015) published a systematic review of RCTs and observational studies on bariatric surgery in patients with T2D and a BMI less than 35 kg/m².¹²⁶ Eleven comparative trials of medical therapy versus bariatric surgery were included, with 5 RCTs and 6 nonrandomized comparative studies identified. Follow-up was between 1 and 3 years. The primary outcome reported was remission of diabetes. On combined analysis, bariatric surgery was associated with a higher remission rate than medical therapy (OR, 14.1; 95% CI, 6.7 to 29.9; p < .001). On secondary outcomes, surgery was associated with a greater decrease in BMI (MD, -5.5 kg/m2; 95% CI, -6.7 to -4.3; p < .001), a lower HbA1C level (MD, -1.4%; 95% CI, -1.9 to -0.9; p < .001), lower rates of hypertension (OR, 0.25; 95% CI, 0.12 to 0.50; p < .001), and lower rates of dyslipidemia (OR, 0.21; 95% CI, 0.10 to 0.44; p < .001).

Also, Rao et al. (2015) published a meta-analysis of short-term outcomes for patients with T2D and a BMI of 35 kg/m² or less who underwent RYGB.¹²⁷ Nine articles were included (N = 343 patients). After 12 months, patients with T2D had a significant decrease in BMI (weighted MD, -7.42; 95% CI, -8.87 to -5.97; p < .001) and improvements in HbA1C levels (weighted MD, -2.76; 95% CI, -3.41 to -2.11; p < .000). Reviewers reported that longer term follow-up would be needed.

Previously, a 2012 TEC Assessment evaluated bariatric surgery in diabetic patients with a BMI less than 35 kg/m².¹²⁸ The evidence consisted mainly of case series. The Assessment identified only observational studies. Based on the data, the Assessment concluded that gastric bypass met TEC criteria as a treatment for diabetes in patients with a BMI less than 35 kg/m² but that other procedures did not meet the TEC criteria for this indication:

• There were no randomized trials comparing bariatric surgery with medical treatment for diabetic subjects with a BMI less than 35 kg/m². There was only 1 randomized trial comparing 2 bariatric procedures. Therefore, studies were categorized by procedure type and presented as case series, regardless of the underlying study type.
• Nine studies reported diabetes remission rates and other outcomes in subjects undergoing gastric bypass. Diabetes remission rates varied between 48% and 100% at follow-up times of 1 year and beyond. One study was a randomized trial of gastric bypass versus SG; in it, diabetes remission associated with gastric bypass was 93% versus 47% for SG at 1 year.
• Two studies reported outcomes of SG. Diabetes remission rates were 55% and 47% at 1 year.
• One study reported outcomes of ileal interposition. The diabetes remission rate at a mean follow-up time of 39.1 months was 78.3%.
• Two studies reported outcomes of gastric banding. The outcomes reported were not considered to be rigorous, because the only measure of diabetes outcome was the withdrawal of diabetes medication. Reported remission rates were 27.5% and 50% at variable follow-up times.
• One study of BPD reported a remission rate of 67% for subjects with a BMI between 30 and 35 kg/m² and 27% for subjects with a BMI between 25 and 30 kg/m² at 12-month follow-up.
• One study reported outcomes of duodenojejunal exclusion. Subjects in this study had more severe diabetes than subjects enrolled in other studies; 100% were on insulin treatment, and the duration of diabetes was between 5 and 15 years. The diabetes remission rate was 17% at 6 months.

Summaries of various systematic reviews and meta-analyses on the use of bariatric or metabolic surgery in patients with a BMI < 35 kg/m² are available, and report efficacy in achieving weight loss, glycemic control, T2D remission, and mitigation of various cardiovascular disease factors through 1 to 5 years of follow-up.^129,130 However, current studies are limited by heterogeneity in applied surgical intervention and threshold definitions for T2D remission. Longer-term (> 5 years) RCTs evaluating the use of metabolic surgery in lower BMI patients for the treatment of T2D are pending (NCT02328599).

Observational studies evaluating patients undergoing bariatric surgery in patients with T2D with a follow-up to 3 or more years are shown in Table 18.

Table 18. Randomized Controlled Trials of Bariatric Surgery Procedures Comparing Patients Who Had Type 2 Diabetes With Controls

Study (Country)	N	BMI Range, kg/m2	Patients With BMI ≤35 kg/m2	Length of FU, years	Definition of Diabetes Remission	Diabetes Remission Rate, n/N (%)			P-Value
Simonson et al. (2019)131 (U.S.)	40	30 to 45	39% LAGB; 36% DWM	3	FBS < 126 mg/dL AND HbA1C < 6.5%	13%		5%	.601
						Surgery (LAGB)		Control (ILI/A1C-R)
Dixon et al. (2008)49 (U.S.)	60	30 to 40	22%	2	% achieving FBS < 126mg/dL HbA1C < 6.2% (off meds)	22/30 (93%)		4/30 (13%)	< .001
						Surgery (RYGB)		Control (HILI/A1C-R)
Ikramuddin et al. (2015)132 (U.S.)	120	30 to 40	59%	2	% achieving all 3 ADA goals: HbA1C < 7.0% LDL < 2.59 mmol/L SBP < 130 mm Hg	26/60 (43%)		8/59 (14%)	< .001
						Surgery (RYGB)	Control 1 (GCP/A1C-R)	Control 2 (GCP/A1C-S)
Liang et al. (2013)133 (China)	108	>28g		1	T2D remissionb	28/31 (90%)	0%	0%	< .05
						Surgery (RYGB)	Surgery (LAGB)	Control (HILI/A1C-S)
Courcoulas et al. (2015)134 (U.S.)	61	30 to 40	43%	3	Partial: HbA1C < 6.5% Full: HbA1C < 5.7% (off meds)	8/20 (40%) Full: 3/20 (15%)	6/21 (29%) Full: 1/21 (5%)	0%	.004
Courcoluas et al. (2020)135, (U.S.)				5	Partial: HbA1c < 6.5% Full: HbA1C < 5.7% (off meds)	6/20 (30%) Full: 1/20 (5%)	4/21 (19%) Full: 0	0%	.0208
						Completers
						Surgery (RYGB)	Surgery (LSG)	Control (ILI/A1C-S)
Schauer et al. (2017)136, (U.S.)	150	27 to 43	37%	5h	% HbA1C < 6.0% (meds)	14/49 (29%)	11/49 (23%)	2/38 (5%)	0.01c/0.03d
						Intention-to-Treat
						26.4%	20.4%	7.3%	0.08e/0.17f
						Surgery (RYGB)	Surgery (BPD)	Control (GCP/A1C-S)
Mingrone et al. (2015)125, (Italy)	60	35+	0%	5	%HbA1C ≤ 6.5% (% meds × 1 y)	8/19 (42%)	13/19 (68%)	0%	< 0.001
						Surgery (LAGB)		Control (ILI/A1C-R)
Wentworth et al. (2014)137, (Australia)	51	25 to 30	100%	2	< 125 mg/dL or 200 mg/dL 2-h OGTT (off meds x 2 d)	12/23 (52%)		2/25 (8%)	0.001
						Surgery (RYGB)		Control (HILI/A1C-S)
Halperin et al. (2014)138, (U.S.)	43	30 to 42	30%	1	% HbA1C < 6.5%	11/19 (58%)		3/19 (16%)	.03

ADA: American Diabetes Association; A1C-R: HbA1c reasonable goal of < 7%; A1C-S: HbA1c stringent goal of < 6.5%; BMI: body mass index; BPD: biliopancreatic diversion; DWM: diabetes and weight management; FBS: fasting blood sugar; FU: follow-up; GCP: good clinical practice; HbA1c: hemoglobin A1c; HILI: 
highly intensive lifestyle intervention; ILI: intensive lifestyle intervention; LAGB; laparoscopic adjustable gastric banding; 
LDL: low-density lipoprotein; LSG: laparoscopic sleeve gastrectomy; OGTT: Oral Glucose Tolerance Test; RYGB: Roux-en-Y gastric bypass; SBP: systolic blood pressure; T2D: type 2 diabetes.
^a All RCTs in this table are in the Wu et al. (2016) meta-analysis; 7 of the 8 (except Mingrone et al) are in the Muller-Stich et al. (2015) meta-analysis; the Rao et al. (2015) meta-analysis and the TEC Assessment did not include RCTs. 
No additional RCTs comparing bariatric surgery with nonsurgical treatment in patients who had T2D were identified.
^b Used as a secondary outcome. Primary outcome was change in left ventricular mass index.
^c Unadjusted (RYGB vs. control).
^d Unadjusted (LSG vs. control).
^e RYGB vs. control.
^f LSG vs. control.
^g WHO Asia-Pacific Obesity Classification.
^h Through February 2017.

Table 19. Observational Studies Assessing Bariatric Surgery in Patients With Type 2 Diabetes With Follow-Up ≥3 Years

Study (Country)	N	BMI Range, kg/m2	Patients With BMI ≤35 kg/m2	Length of FU	Interv	Mean HbA1c	Mean BMI, kg/m2	Diabetes Remission Rate
Group 1						Base	FU	Base	FU
Scopinaro et al. (2014)80 (Italy)	20 treated; 27 matched diabetic controls	30 to 34.9	100%	3 y	RYGB	9.5%	7.0%a	32.9	26.0a	5/20 (25%)
					Control	9.3%	7.7%a	33.0	32.6
Lanzarini et al. (2013)139 (Chile)	31	30 to 35	100%	30 moc	RYGB	7.9%	5.5%a	33.1	24.7a	29/31 (94%)
Boza et al. (2011)140 (Chile)	30	< 35	100%	2 y	RYGB	8.1%	≈6.2%a,b	33.5	23.9a	12 mo: 25/30 (83.3%) 2 y: 13/20 (65%)
DePaula et al. (2012)141 (Brazil)	202	< 35	100%	39 moc	SG