Abstract
As the prevalence of morbid obesity continues to climb in America, so does the popularity of the Roux-en-Y gastric bypass (RYGB) to achieve weight loss goals; however, a long-term risk of RYGB is marginal ulceration, which requires urgent surgery if perforated. We sought to identify characteristics associated with elective vs urgent presentation for marginal ulcer following RYGB. Retrospective data for consecutive cases with marginal ulcers that required surgical intervention from May 2016 to February 2021 were queried from our institution’s bariatric database, and differences in patient characteristics and clinical course were assessed according to presentation. Forty-three patients underwent surgery for marginal ulcer during the study timeframe. Twenty-four (56%) patients presented electively and were treated with resection of the gastroenterostomy and reanastomosis; the remaining 19 (44%) presented urgently with perforation and were treated with omental patch repair. Demographics, comorbidities, and medications were similar between groups. Patients with urgent presentations were less likely to have bleeds (0% vs. 33%, P = 0.0056) and strictures (16% vs. 46%, P = 0.0368), but were more likely to require admission to the intensive care unit (32% vs. 4%, P = 0.0325) and have a longer median length of stay (2 vs. 5 days, P < 0.0001). Bariatric surgeons must properly counsel patients about the risk of marginal ulcer development to prevent dangerous perforation, intensive care unit stays, and long hospitalizations.
Keywords: Anastomotic ulcer, bariatric surgery, gastrojejunal ulcer, marginal ulcer
The prevalence of obesity (body mass index [BMI] > 30 kg/m2) in American adults has increased from 31% in 1999 to 42% in 2018, and the prevalence of morbid obesity (BMI ≥40 kg/m2) has nearly doubled in that same timeframe (5% to 9%).1 In response, the popularity of bariatric surgical procedures has also increased. The Roux-en-Y gastric bypass (RYGB) remains one of the most commonly performed procedures in the United States for weight loss, with over 450,000 cases since 2011.2 As with any surgical procedure, there are associated risks; the most common long-term complications are anastomotic stricture, internal herniation, and marginal ulceration (also known as gastrojejunal ulcer, anastomotic ulcer, or ischemic ulcer).3
The reported rates of marginal ulcer development vary from 0.6% to 25%, depending on risk factors, which can involve both mechanical and functional outcomes of the surgery. An increased size of the gastric pouch, the presence of a gastrogastric fistula,4,5 and the use of nonabsorbable suture6 are all mechanical risk factors. Other risk factors include, but are not limited to, use of nonsteroidal antiinflammatory drugs (NSAIDs), tobacco use, alcohol use, Helicobacter pylori infection, diabetes mellitus, immunosuppression,7 and gastroesophageal reflux disease. Use of a low-dose selective serotonin reuptake inhibitor has been associated with decreased risk of marginal ulcer development.8
Patients can present electively with a nonperforated ulcer, having classic symptoms of epigastric pain, nausea, and emesis, or less commonly with melena or hematemesis.4 Alternatively, the ulcer may perforate and cause an urgent presentation. The timing and severity of presentation often determine the course of treatment. Omental patch repair has been established as an effective surgical option for perforated marginal ulcers,9 while gastrojejunostomy revision is preferred in an elective setting or urgently without hemodynamic instability.10
To more clearly understand ulcer development, we reviewed our academic tertiary center’s cases of marginal ulcers that required surgical intervention and sought to identify characteristics associated with elective vs urgent presentation, as well as the ulcer time to development and subsequent requirement of intensive care unit (ICU) admission.
METHODS
Institutional review board approval was granted by the Baylor Scott & White Research Institute prior to the conduct of this study. Retrospective data for consecutive cases with marginal ulcers that required surgical intervention from May 2016 to February 2021 were queried from our institution’s bariatric database. The cases selected were those who underwent surgery for marginal ulcer disease; patients managed nonoperatively were excluded from this study. Charts were reviewed for preoperative characteristics including age, sex, initial BMI, BMI at the time of surgery, history of other revisions, and the time (in years) from RYGB to surgery for marginal ulcer. Patient comorbidities examined included diabetes mellitus, gastroesophageal reflux disease, NSAID use, alcohol use, tobacco use, immunosuppression, and selective serotonin reuptake inhibitor use. Information on the nature of presentation, type of operation performed, location and size of the perforation or ulcer, and presence of other complications such as stricture, gastrogastric fistula, and bleeding was also collected. Postoperative length of stay, intensive care requirements, and mortality were recorded. NSAID, alcohol, and tobacco use was counted if there was any mention in the documented history of present illness of repeated use, or if there were any prior diagnoses of alcohol or tobacco use or abuse disorders.
The technique used for reconstruction and reanastomosis was similar for all patients. This was done with initial firing of a linear stapler across the gastric pouch. This was placed just below the left gastric pedicle a full 2 cm cephalad to the gastroenterostomy. A second firing of a linear stapler was made in the cephalad direction up to the angle of His approximately 1 cm medial to the prior pouch staple line, completing the creation of the gastric pouch. An enterotomy was made on the end of the Roux limb through which the end-to-end anastomosis (EEA) stapler was inserted. The OrVil 25 mm anvil device was then used transorally to mate with the 25 mm EEA stapler. A spike was floored through the intermesenteric border, mated with the anvil, closed, and fired to create the gastroenterostomy. After removal of the EEA stapler, the distal end of the Roux limb was resected with a linear stapler to close the defect.
Categorical variables are presented as frequencies and percentages. Continuous variables are presented as mean ± standard deviation, or median [quartile 1, quartile 3], if skewed. We tested for differences in patient characteristics and clinical courses for patients according to presentation (elective vs urgent), surgical timing (within 5 years of index procedure vs ≥5 years), and ICU admission using two-sample t-tests, chi-square tests, or Wilcoxon rank sum tests and Fisher’s exact tests, as appropriate. We considered a multivariable Cox model for the outcome of time to ulcer using variables identified as significant in bivariate analyses.
RESULTS
Over the 4 years of data examined, 43 patients underwent surgery for marginal ulcer disease (Table 1). The average age was 51 ± 13 years, and 91% were women. The average initial BMI was 49 ± 10 kg/m2, and the average BMI at presentation of surgery for marginal ulcer was 30 ± 10 kg/m2 (median = 29 [23, 34]). RYGB was the index operation for 38 patients (88%); 3 (7%) patients received laparoscopic vertical sleeve gastrectomies, 1 (2%) underwent laparoscopic adjustable gastric banding, and 1 (2%) had a Nissen fundoplication; all converted to RYGB prior to presentation. Eleven patients (26%) had a history of prior gastric bypass revisions, due to gastrogastric fistula repair, internal hernia repair, or prior perforations.
Table 1.
Patient characteristics by presentation type (n = 43)
| Presentation |
||||
|---|---|---|---|---|
| Overall (n = 43) |
Elective (n = 24) |
Urgent (n = 19) |
P value | |
| Women | 39 (91%) | 21 (88%) | 18 (95%) | 0.618 |
| Age (years) | 51 ± 13 | 51 ± 13 | 51 ± 13 | 0.888 |
| Age at index procedure (years) | 44 ± 14 | 44 ± 14 | 44 ± 13 | 0.925 |
| Initial body mass index (kg/m2)* | 49 ± 10 | 48 ± 10 | 51 ± 9 | 0.279 |
| Current body mass index (kg/m2) | 29 [23, 34] | 29 [23, 32] | 31 [24, 42] | 0.124 |
| Paired reduction in body mass index (kg/m2)* | 18 ± 10 | 20 ± 11 | 15 ± 8 | 0.145 |
| Diabetes | 7 (16%) | 5 (21%) | 2 (11%) | 0.437 |
| Gastroesophageal reflux disease | 37 (86%) | 22 (92%) | 15 (79%) | 0.380 |
| Alcohol consumption >1 glass per day | 5 (12%) | 2 (8%) | 3 (16%) | 0.640 |
| Tobacco | 0.112 | |||
| Current | 8 (19%) | 2 (8%) | 6 (32%) | |
| Former | 15 (35%) | 8 (33%) | 7 (37%) | |
| Never | 20 (47%) | 14 (58%) | 6 (32%) | |
| Oral steroid | 2 (5%) | 1 (4%) | 1 (5%) | 1 |
| Nonsteroidal antiinflammatory drug use | 4 (9%) | 1 (4%) | 3 (16%) | 0.306 |
| Antidepressant use | 21 (49%) | 12 (50%) | 9 (47%) | 1 |
| Selective serotonin reuptake inhibitor use | 13 (30%) | 8 (33%) | 5 (26%) | 0.743 |
| Index procedure | 0.198 | |||
| Sleeve gastrectomy | 3 (7%) | 3 (13%) | 0 (0%) | |
| Nissen fundoplication | 1 (2%) | 1 (4%) | 0 (0%) | |
| Roux-en-Y gastric bypass | 38 (88%) | 20 (83%) | 18 (95%) | |
| Laparoscopic band | 1 (2%) | 0 (0%) | 1 (5%) | |
| Index approach | 1 | |||
| Laparoscopic | 37 (86%) | 20 (83%) | 17 (89%) | |
| Laparoscopic-to-open | 1 (2%) | 1 (4%) | 0 (0%) | |
| Open | 5 (12%) | 3 (13%) | 2 (11%) | |
| Prior revision | 16 (37%) | 11 (46%) | 5 (26%) | 0.189 |
| Days from consultation to elective revision | 90 [33, 136] | 90 [33, 136] | – | – |
| Time to ulcer (years) | 6 [2, 10] | 6 [2, 9.5] | 6 [4, 10] | 0.451 |
*Missing 9 observations for initial and paired body mass index.
Of the 43 patients, 24 (56%) presented on an elective basis, all of whom were treated with resection of the gastroenterostomy and reanastomosis (Table 1). These patients underwent this surgery a median of 90 [33, 136] days following initial consultation for the ulcer. The remaining 19 patients (44%) presented with perforation in an urgent setting, all of whom were treated with omental patch repair of the perforated ulcer. Only 2 (11%) of these 19 patients subsequently underwent elective resection of the gastroenterostomy and reanastomosis. Forty (93%) of the 43 patients were managed laparoscopically. An open approach was used for the other 3 (7%) patients, all of whom presented with perforation. The location of the ulcers varied among patients. A total of 24 (56%) were located at the gastrojejunal anastomosis, 14 (33%) were on the jejunal side, and only 1 (2%) on the gastric side. There were 12 (28%) located anteriorly, and only 1 (2%) posteriorly. Additionally, there was 1 (2%) duodenal ulcer and 1 (2%) pyloric ulcer on the gastric remnant. In the other 2 patients, the location of the ulcer was not reported. Other intraoperative findings included stricture (14, 33%), gastrogastric fistula (7, 16%), and bleeding (8, 19%). The median time to ulcer was 6 [2, 10] years, with 15 (35%) patients requiring surgery within 5 years of their index procedure. Table 2 lists findings related to the clinical course. The median length of stay for all patients was 3 [2, 5] days. Postoperatively, 7 patients (16%) required admission to the ICU. There were no deaths within 30 days of surgery; however, one patient expired at a skilled nursing facility after discharge more than 30 days from surgery, and one patient expired over a year later due to complications from alcoholic cirrhosis.
Table 2.
Clinical course by presentation type (n = 43)
| Characteristic | Overall (n = 43) | Presentation |
P value | |
|---|---|---|---|---|
| Elective (n = 24) |
Urgent (n = 19) |
|||
| Current approach | 0.079 | |||
| Laparoscopic | 40 (93%) | 24 (100%) | 16 (84%) | |
| Laparoscopic-to-open | 2 (5%) | 0 (0%) | 2 (11%) | |
| Open | 1 (2%) | 0 (0%) | 1 (5%) | |
| Ulcer size (mm)* | 7 [3, 15] | 15 [10, 20] | 4 [3, 5] | 0.005 |
| Length of stay postprocedure (days) | 3 [2, 5] | 2 [2, 3] | 5 [4, 8] | <0.0001 |
| Intraoperative finding: bleed | 8 (19%) | 8 (33%) | 0 (0%) | 0.006 |
| Intraoperative finding: stricture | 14 (33%) | 11 (46%) | 3 (16%) | 0.037 |
| Intraoperative finding: fistula | 8 (19%) | 5 (21%) | 3 (16%) | 1 |
| Intensive care unit admission | 7 (16%) | 1 (4%) | 6 (32%) | 0.033 |
| Death | 2 (5%) | 1 (4%) | 1 (5%) | 1 |
*Missing 11 observations for ulcer size.
After stratifying by presentation type (Table 1), the median ulcer size for those presenting urgently was significantly lower than that of patients presenting electively (3.5 [3, 5] mm vs. 15 [10, 20] mm, P = 0.005). Bleeding and stricture were observed less frequently among urgent patients (0% vs. 33%, P = 0.006; 16% vs. 46%, P = 0.037, respectively) (Table 2). Patients with elective resections had shorter lengths of stay than those with perforations (2 [2, 3] days vs. 5 [4, 8] days, P < 0.0001) and were less likely to require ICU admission (1 [4%] vs. 6 [32%], P = 0.033).
Most patient characteristics were similar between those requiring intervention within 5 years of index compared to those requiring intervention at 5 or more years (Table 3); however, patients requiring earlier intervention had significantly lower BMI at the index procedure and current presentation (initial: 42 [40, 49] vs. 50 [47, 57], P = 0.0394; current: 24 [22, 30] vs. 31 [27, 36], P = 0.032) and were less likely to have had RYGB as their index procedure (11 [73%] vs. 27 [96%], P = 0.019). A multivariable Cox model examining initial BMI and a dichotomized index procedure variable failed to identify either of these factors as being independently associated with the outcome (P > 0.05).
Table 3.
Patient characteristics according to timing of surgery (n = 43)
| Characteristic | Early surgery (<5 years from index procedure) |
P value | |
|---|---|---|---|
| No (n = 28) | Yes (n = 15) | ||
| Gender (females) | 26 (93%) | 13 (87%) | 0.602 |
| Age (years) | 52 ± 12 | 50 ± 14 | 0.685 |
| Age at index procedure (years) | 42 ± 13 | 48 ± 15 | 0.173 |
| Initial body mass index (kg/m2)* | 50 [47, 57] | 42 [40, 49] | 0.039 |
| Current body mass index (kg/m2) | 31 [26.5, 36] | 24 [22, 30] | 0.032 |
| Paired reduction in body mass index (kg/m2)* | 18 [9, 23] | 16 [12, 25] | 0.532 |
| Diabetes | 6 (21%) | 1 (7%) | 0.391 |
| Gastroesophageal reflux disease | 23 (82%) | 14 (93%) | 0.403 |
| Alcohol consumption >1 glass per day | 5 (18%) | 0 (0%) | 0.145 |
| Tobacco use | 0.303 | ||
| Current | 5 (18%) | 3 (20%) | |
| Former | 12 (43%) | 3 (20%) | |
| Never | 11 (39%) | 9 (60%) | |
| Oral steroid use | 2 (7%) | 0 (0%) | 0.535 |
| Nonsteroidal antiinflammatory drug use | 2 (7%) | 2 (13%) | 0.602 |
| Antidepressant use | 15 (54%) | 6 (40%) | 0.396 |
| Selective serotonin reuptake inhibitor use | 8 (29%) | 5 (33%) | 0.742 |
| Index procedure | 0.019 | ||
| Sleeve gastrectomy | 0 (0%) | 3 (20%) | |
| Nissen fundoplication | 0 (0%) | 1 (7%) | |
| Roux-en-Y gastric bypass | 27 (96%) | 11 (73%) | |
| Laparoscopic band | 1 (4%) | 0 (0%) | |
| Index procedural approach | 0.201 | ||
| Laparoscopic | 22 (79%) | 15 (100%) | |
| Laparoscopic-to-open | 1 (4%) | 0 (0%) | |
| Open | 5 (18%) | 0 (0%) | |
| Prior revision | 8 (29%) | 8 (53%) | 0.109 |
| Days from consultation to elective revision (n = 24) | 38 [17, 122] | 97 [50, 218] | 0.201 |
| Years from index procedure for ulcer | 9 [6, 13] | 2 [1, 3] | <0.001 |
| Elective presentation | 14 (50%) | 10 (67%) | 0.294 |
| Ulcer size (mm)* | 7 [3, 20] | 8 [3, 15] | 0.512 |
| Length of stay postprocedure (days) | 3 [2, 5] | 3 [2, 4] | 0.166 |
*Missing 9 observations for initial and paired body mass index; missing 11 observations for ulcer size.
The two characteristics associated with ICU admission (Table 4) were consuming >1 serving of alcohol per day (43% vs. 6%, P = 0.024) and presentation for perforation (86% vs. 36%, P = 0.033). Similarly, those without ICU stays had shorter length of stay than those with ICU admission (3 [2, 4] vs. 8 [5, 28], P = 0.0008).
Table 4.
Patient characteristics according to intensive care unit admission (n = 43)
| Characteristic | Intensive care unit admission |
P value | |
|---|---|---|---|
| No (n = 36) | Yes (n = 7) | ||
| Females | 32 (89%) | 7 (100%) | 1 |
| Age (years) | 51 ± 12 | 49 ± 16 | 0.683 |
| Age at index procedure (years) | 45 ± 13 | 42 ± 15 | 0.653 |
| Initial body mass index (kg/m2)* | 49 ± 10 | 51 ± 6 | 0.581 |
| Current body mass index (kg/m2) | 29 [23, 34] | 31 [30, 45] | 0.134 |
| Paired reduction in body mass index (kg/m2)* | 18 ± 11 | 14 ± 5 | 0.321 |
| Diabetes | 5 (14%) | 2 (29%) | 0.318 |
| Gastroesophageal reflux disease | 31 (86%) | 6 (86%) | 1 |
| Alcohol consumption >1 glass per day | 2 (6%) | 3 (43%) | 0.024 |
| Tobacco use | 0.213 | ||
| Current | 5 (14%) | 3 (43%) | |
| Former | 13 (36%) | 2 (29%) | |
| Never | 18 (50%) | 2 (29%) | |
| Oral steroid | 1 (3%) | 1 (14%) | 0.302 |
| Nonsteroidal antiinflammatory drug use | 4 (11%) | 0 (0%) | 1 |
| Antidepressant use | 17 (47%) | 4 (57%) | 0.698 |
| Selective serotonin reuptake inhibitor use | 12 (33%) | 1 (14%) | 0.412 |
| Index procedure | 0.608 | ||
| Sleeve gastrectomy | 2 (6%) | 1 (14%) | |
| Nissen fundoplication | 1 (3%) | 0 (0%) | |
| Roux-en-Y gastric bypass | 32 (89%) | 6 (86%) | |
| Laparoscopic band | 1 (3%) | 0 (0%) | |
| Index procedural approach | 1 | ||
| Laparoscopic | 31 (86%) | 6 (86%) | |
| Laparoscopic-to-open | 1 (3%) | 0 (0%) | |
| Open | 4 (11%) | 1 (14%) | |
| Prior revision | 14 (39%) | 2 (29%) | 0.695 |
| Time from index to procedure for ulcer (years) | 6 [2, 10] | 6 [5, 10] | 0.680 |
| Surgery within 5 years of index procedure | 14 (39%) | 1 (14%) | 0.391 |
| Urgent presentation | 13 (36%) | 6 (86%) | 0.033 |
| Ulcer size (mm)* | 8 [3, 15] | 5 [5, 10] | 0.891 |
| Length of stay postprocedure (days) | 3 [2, 4] | 8 [5, 25] | <0.001 |
*Missing 9 observations for initial and paired body mass index; missing 11 observations for ulcer size.
DISCUSSION
In this study of 43 patients with prior RYGB who subsequently presented for surgery for marginal ulcer, patients with urgent presentations were less likely to have bleeds and strictures, but were more likely to require ICU admission and have a prolonged length of stay. We found that lower BMI was associated with ulcer development within 5 years of index procedure, but that the timing of ulcer development was not associated with outcomes. Early ulcer development was also associated with prior conversions to RYGB, with a sleeve gastrectomy, laparoscopic gastric banding, or Nissen fundoplication as the index case. Additionally, we found that consuming >1 alcoholic beverage per day and presenting with a perforation increased the risk of an ICU admission.
The presentation, either electively or urgently (with perforation), clearly determined the operation to be performed. When elective, the hemodynamic stability of the patient as well as the anatomic findings lead to a more definitive reconstruction. In contrast, patients with perforation are often hemodynamically compromised, contaminated, and friable, an environment more favorable to a patch repair. Among patients with elective presentation, the observed larger ulcer size shows that it requires a larger surface area to become clinically relevant without perforation. Stricture and bleeding were more common in the elective cases, likely attributable to chronic inflammation and scarring with an indolent course.11 The ulcers in patients with perforation would have likely progressed in size and inflammation had the perforation not occurred.
Patients requiring earlier intervention (<5 years from index) had significantly lower BMI both at the index operation and the current presentation. With an average BMI of 42 vs. 50 at the index operation, these patients may have had easier access to medical care and may have been more likely to return for care at the early onset of postoperative symptoms.12 The lower BMI at presentation for reoperation could also be partially explained by food avoidance due to exacerbation of symptoms. When evaluating a patient with chronic abdominal pain after RYGB, persistent weight loss from food avoidance and malnutrition is an alarming symptom. It is possible that the patients without food avoidance may take longer to return to clinic for elective reconstruction.13 Another explanation for the difference in BMI at early (<5 years) reoperation is the weight nadir at 1 to 2 years post-RYGB, with consistent weight regain as time since the index operation passes.14
The association of alcohol consumption with ICU admission is a clear example of the importance of healthy lifestyle choices after bariatric surgery. Similarly, the association between tobacco use and ICU admission trended toward statistical significance, with 43% of those admitted to the ICU being current tobacco users, compared to only 14% of those who did not require ICU admission. Although reported in the literature, the higher incidence of marginal ulcers among users of tobacco, alcohol, or NSAIDs was not seen directly with our patient population. This is most likely due to a smaller sample size, as well as potential variance in reported use. Nonetheless, these behavioral factors speak to the importance of education and counseling after RYGB.15 Bariatric surgeons carry the responsibility to convey to patients the increased risk of marginal ulcer development with modifiable choices such as use of tobacco, alcohol, and/or NSAIDs.
Our patients’ characteristics (i.e., age, gender, and initial BMI) were similar to a recent meta-analysis. The average age of our cohort was 51 compared to 46 in the meta-analysis. Both groups were overwhelmingly female, 91% vs. 80%. Lastly, the average BMI of 49 kg/m2 is just slightly higher than the reported average of 46 kg/m2 in the study.16 Interestingly, this meta-analysis also demonstrated a relationship of antidepressant use with weight regain. The frequency of antidepressant use in our cohort of patients at 49% is much higher than the national average in the United States, reported to be 13.2%.17 This suggests that mental health, in addition to physical health, is important when evaluating patients for RYGB, both in avoiding weight recidivism and marginal ulcer development.
This study has all the limitations inherent to a small observational retrospective design. Additionally, because this study only included patients who required surgical intervention for marginal ulcer, we could not delineate risk factors for ulcer development (i.e., we did not have data on controls). Further, we did not provide care at the time of index procedure, so we cannot account for center variability of procedural protocols. Initial BMI was missing for some patients and may have contributed to a loss of power in the multivariable Cox analysis. We attempted to create a multivariable (Firth) logistic regression for the outcome of ICU admittance; however, we observed instabilities within the models attributable to small case counts.
In this small observational study of 43 cases with prior RYGB who subsequently presented for surgery for marginal ulcer, we found that urgent presentations were associated with higher rates of postoperative ICU admissions and longer lengths of stay. We also found that consuming >1 alcoholic beverage per day was associated with ICU admission. As such, it is critically important for bariatric surgeons to properly counsel and educate patients regarding the possibility of marginal ulcers as well as associated modifiable risk behaviors. Future work in this field is warranted to decrease the risk of perforation.
Disclosure statement/Funding
The authors report no funding or conflicts of interest.
References
- 1.Hales CM, Carroll MD, Fryar CD, Ogden CL.. Prevalence of obesity and severe obesity among adults: United States, 2017–2018. NCHS Data Brief. 2020;360:1–8. [PubMed] [Google Scholar]
- 2.American Society for Metabolic and Bariatric Surgery . Estimate of bariatric surgery numbers, 2011–2019. https://asmbs.org/resources/estimate-of-bariatric-surgery-numbers. Accessed 2021.
- 3.Al Harakeh AB. Complications of laparoscopic Roux-en-Y gastric bypass. Surg Clin North Am. 2011;91(6):1225–1237, viii. doi: 10.1016/j.suc.2011.08.011. [DOI] [PubMed] [Google Scholar]
- 4.Coblijn UK, Goucham AB, Lagarde SM, Kuiken SD, van Wagensveld BA.. Development of ulcer disease after Roux-en-Y gastric bypass, incidence, risk factors, and patient presentation: a systematic review. Obes Surg. 2014;24(2):299–309. doi: 10.1007/s11695-013-1118-5. [DOI] [PubMed] [Google Scholar]
- 5.Capella JF, Capella RF.. Gastro-gastric fistulas and marginal ulcers in gastric bypass procedures for weight reduction. Obes Surg. 1999;9(1):22–27; discussion 28. doi: 10.1381/096089299765553674. [DOI] [PubMed] [Google Scholar]
- 6.Sacks BC, Mattar SG, Qureshi FG, et al. Incidence of marginal ulcers and the use of absorbable anastomotic sutures in laparoscopic Roux-en-Y gastric bypass. Surg Obes Relat Dis. 2006;2(1):11–16. doi: 10.1016/j.soard.2005.10.013. [DOI] [PubMed] [Google Scholar]
- 7.Di Palma A, Liu B, Maeda A, Anvari M, Jackson T, Okrainec A.. Marginal ulceration following Roux-en-Y gastric bypass: risk factors for ulcer development, recurrence and need for revisional surgery. Surg Endosc. 2021;35(5):2347–2353. doi: 10.1007/s00464-020-07650-0. [DOI] [PubMed] [Google Scholar]
- 8.Sverdén E, Mattsson F, Sondén A, et al. Risk factors for marginal ulcer after gastric bypass surgery for obesity: a population-based cohort study. Ann Surg. 2016;263(4):733–737. doi: 10.1097/SLA.0000000000001300. [DOI] [PubMed] [Google Scholar]
- 9.Wendling MR, Linn JG, Keplinger KM, et al. Omental patch repair effectively treats perforated marginal ulcer following Roux-en-Y gastric bypass. Surg Endosc. 2013;27(2):384–389. doi: 10.1007/s00464-012-2492-0. [DOI] [PubMed] [Google Scholar]
- 10.Chau E, Youn H, Ren-Fielding CJ, Fielding GA, Schwack BF, Kurian MS.. Surgical management and outcomes of patients with marginal ulcer after Roux-en-Y gastric bypass. Surg Obes Relat Dis. 2015;11(5):1071–1075. doi: 10.1016/j.soard.2014.12.011. [DOI] [PubMed] [Google Scholar]
- 11.Sidani S, Akkary E, Bell R.. Catastrophic bleeding from a marginal ulcer after gastric bypass. JSLS. 2013;17(1):148–151. doi: 10.4293/108680812X13517013318274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Jackson TD, Zhang R, Glockler D, et al. Health inequity in access to bariatric surgery: a protocol for a systematic review. Syst Rev. 2014;3:15. doi: 10.1186/2046-4053-3-15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Conceição EM, Goldschmidt A.. Disordered eating after bariatric surgery: clinical aspects, impact on outcomes, and intervention strategies. Curr Opin Psychiat. 2019;32(6):504–509. doi: 10.1097/YCO.0000000000000549. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Thomas DD, Anderson WA, Apovian CM, et al. Weight recidivism after Roux-en-Y gastric bypass surgery: an 11-year experience in a multiethnic medical center. Obesity (Silver Spring). 2019;27(2):217–225. doi: 10.1002/oby.22360. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Marcotte E, Chand B.. Management and prevention of surgical and nutritional complications after bariatric surgery. Surg Clin North Am. 2016;96(4):843–856. doi: 10.1016/j.suc.2016.03.006. [DOI] [PubMed] [Google Scholar]
- 16.King WC, Belle SH, Hinerman AS, Mitchell JE, Steffen KJ, Courcoulas AP.. Patient behaviors and characteristics related to weight regain after Roux-en-Y gastric bypass: a multicenter prospective cohort study. Ann Surg. 2020;272(6):1044–1052. doi: 10.1097/SLA.0000000000003281. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Brody DJ, Gu Q.. Antidepressant use among adults: United States, 2015–2018. NCHS Data Brief. 2020;377:1–8. [PubMed] [Google Scholar]