Knotless tissue control devices for hand-sewn gastrojejunostomy in robotic-assisted Roux-en-Y gastric bypass Authors: Stephen Johnston, Najmuddin Gunja, Aakash Jha, Walter Danker, Shanthi Krishnaswami, Zhun Cao, Barbara Johnson, Jörg Tomaszewski Published: 2026 ABSTRACT BackgroundLiterature on clinical and economic outcomes with hand-sewn gastrojejunostomy using Spiral Antibacterial Knotless Tissue Control Devices (KTCD) in Roux-en-Y gastric bypass (RYGB) is scant.ObjectivesAssess clinical and economic outcomes of patients receiving KTCD for hand-sewn gastrojejunostomy vs. alternative means of anastomosis in robotic-assisted RYGB. SettingUS hospital.MethodsPatients who received gastrojejunostomy using KTCD in robotic-assisted RYGB between 01/01/2020-08/31/2023 identified in the Premier Healthcare Database (PHD) were evaluated via medical chart review (RYGB=index). Patients with alternative means of anastomosis, including non-KTCD sutures/staplers, in robotic-assisted RYGB in the PHD during the same timeframe were utilized as controls. Optimization-based stable balance weighting balanced the cohorts on baseline characteristics. Outcomes included a composite complication measure (anastomotic leak, bleeding, dilation of small intestine, infection, pelvic abscess, ileus/bowel obstruction) evaluated at index, index through 30 days and 90 days, anastomotic leak through 90 days, length of stay (LOS), and index costs.Results145 patients in two hospitals met KTCD selection criteria; 6,491 met control criteria. Incidences of complications at index, 30 days, and 90 days were lower for KTCD vs. controls (0.7% vs. 4.5%, 0.7% vs. 7.3%, 2.1% vs. 8.3%; all p<0.01). LOS was shorter for KTCD vs. controls (1.34 vs. 1.62 days; p=0.002). Anastomotic leaks (1.4% vs. 2.0%; p=0.143) and index costs ($17,042 vs. $16,431; p=0.494) did not differ between KTCD vs. controls.ConclusionsThis retrospective analysis of gastrojejunostomy in robotic-assisted RYGB found fully hand-sewn gastrojejunostomy with KTCD was associated with lower complication rates and shorter LOS vs. alternative means of anastomosis.Supplementary InformationThe online version contains supplementary material available at 10.1007/s11695-026-08782-3. Introduction Barbed sutures may reduce localized stress on approximated tissues and may be more effective in preserving tissue tension [1]. Hemostasis and leakage prevention are critical for suture performance, and the utilization of barbed sutures may securely reapproximate tissues with less time, cost, and aggravation [2]. STRATAFIX Spiral Antibacterial Knotless Tissue Control Devices (KTCD; Ethicon, Inc.) have barbs in a spiral pattern that allow for tissue approximation without the need to tie surgical knots [3]. KTCD are intended to be used in a variety of open, laparoscopic, and robotic procedures across various medical specialties [3]. The utilization of KTCD has the potential to improve patient outcomes in minimally-invasive Roux-en-Y gastric bypass (RYGB) given that RYGB involves multiple suturing tasks and dissection in challenging tissue and confined spaces. In RYGB, the gastrojejunostomy is a critical anastomosis that may be performed with either mechanical staplers, sutures or the combination of both techniques [4–7]. Although prior research has examined the use of staplers and conventional sutures for the gastrojejunostomy, there is only limited data on the use of barbed sutures for a hand-sewn gastrojejunostomy in RYGB [4, 8, 9]. Furthermore, at the time of this study, the instructions for use (IFU) for KTCD include a statement that the safety and effectiveness have not been established for use in gastrointestinal anastomoses, highlighting a gap in evidence that presents a potential barrier to the confident use of KTCD for this purpose [3]. To address this evidence gap, the present study was undertaken to evaluate STRATAFIX Spiral KTCD in hand-sewn gastrojejunostomy in RYGB; however, candidate hospitals in which fully hand-sewn gastrojejunostomies with KTCD in RYGB were performed only utilized robotic RYGB. Hence, the objectives of the study were amended to assess the post-operative outcomes of patients where KTCD was used for hand-sewn gastrojejunostomy in robotic-assisted RYGB and to compare KTCD vs. alternative means of anastomosis (including conventional sutures, mechanical stapling or the combination of both, “non-KTCD”) for gastrojejunostomy in robotic-assisted RYGB. Materials and Methods Data Sources This retrospective, real-world study used data from the all-payer, geographically diverse Premier Healthcare Database (PHD), which contains standardized clinical coding, including diagnosis and procedure codes, as well as hospital chargemaster descriptions from approximately 25% of US hospital admissions (> 1,300 facilities and > 300 million unique patients) [10]. Candidate hospitals in which KTCD was used during an RYGB between January 1, 2020-August 31, 2023 were identified in the PHD. The hospitals were invited to participate in a retrospective electronic health record (EHR) chart review of unstructured surgical notes to verify that KTCD was indeed used for fully hand-sewn anastomosis; these data were documented in a Case Report Form (CRF). Data captured by the CRF were linked back to the PHD for analysis. In addition, data from the PHD during the same timeframe were utilized to identify a comparative control cohort of patients in which no KTCD were used (i.e., who received alternative means of anastomosis, including stapling and/or sutures). The first RYGB was designated as index. The medical chart review received IRB approval from ADVARRA (approval # MOD01539925). The use of the PHD was reviewed by the WCG Institutional Review Board (WCG® IRB) and was determined to be exempt from broad IRB approval, as this study did not involve human subjects’ research and used data from an anonymous, de-identified, database compliant with the Health Insurance Portability and Accountability Act of 1996. Patient Population Patients aged ≥ 18 years with robotic-assisted RYGB performed with use of KTCD or with alternative means of anastomosis from January 1, 2020-August 31, 2023 and whose hospital contributed data to the PHD for at least 90 days post-discharge were considered for inclusion; although the initial search period for the KTCD cohort extended back to 2016, the earliest year of surgery for patients who were eligible for inclusion in the KTCD cohort was 2020). For the KTCD cohort, the sites confirmed use of KTCD for full/total/complete anastomosis as a single means to create the gastrojejunostomy during the index RYGB. For the comparator cohort, there could be no evidence of KTCD use (including all STRATAFIX variants) during the index RYGB, as measured using catalog numbers and text string searches in the hospital charge master table of the PHD. Although the study was not designed as a “robotic only” study, the participating hospitals for the KTCD cohort only utilized robotic-assisted RYGB; therefore, the non-KTCD control cohort was also restricted to robotic-assisted RYGB. Also, although either STRATAFIX Spiral PDS/PDO or Monocryl sutures could be used for the procedures, the sites indicated that all cases used STRATAFIX Spiral PDS/PDO sutures. Study Measures The following baseline patient demographic and clinical characteristics were evaluated: age, sex, race (i.e., White, Black, Other), ethnicity (i.e., Hispanic, not Hispanic, Other), primary insurance payor type (i.e., Commercial, Medicare, Medicaid, Other), Charlson Comorbidity Index (CCI) [6–8] score (i.e., 0, 1 or 2, ≥ 3), body mass index (BMI) category (i.e., normal weight: 20 to < 25, overweight: 25 to < 30, obesity class 1: 30 to < 35, and obesity class 2 and higher: ≥35.0 distributed across 10 BMI categories), discharge year (i.e., 2020–2023), teaching hospital status, and hospital bed size (i.e., 100–500 beds or 500 + beds). The CCI is a composite measure of comorbidity created by using 17 dimensions associated with chronic disease and overall health conditions [11–13]. Select CCI comorbidities thought to potentially impact RYGB complications based on clinical expert opinion (i.e., myocardial infarction [MI], congestive heart failure [CHF], chronic pulmonary disease, rheumatoid arthritis [RA], liver disease, diabetes, diabetes with complications, and renal disease) were also evaluated to determine if their presence differed between patients with KTCD and non-KTCD. Clinical outcomes were measured using International Classification of Diseases, 10th Revision, Clinical Modification/Procedure (ICD-10-CM) Coding System codes. A composite of the following complications was assessed: anastomotic leak, bleeding, dilation of small intestine, infection, pelvic abscess, and ileus/bowel obstruction at index, index through 30 days and index through 90 days; and anastomotic leak as a standalone alone through 90 days. Because there are no ICD-10-CM diagnosis codes that are specific to anastomotic leak, previously published surrogate diagnoses [14] were used as a proxy to identify anastomotic leaks; specifically, the diagnosis codes used were based on ICD-9-CM to ICD-10-CM forward mapping and omitting code K91.3: ‘Postprocedural intestinal obstruction,’ resulting in: K63.2 ‘Fistula of intestine’, K65.1 ‘Peritoneal abscess’, and K91.89 ‘Other postprocedural complications and disorders of digestive system’. Additional outcomes included index hospital length of stay (LOS) and total index hospital costs. In the Premier Healthcare Database, hospital costs for each encounter are reported directly by data-contributing hospitals and represent costs from the hospital’s perspective (i.e., how much the encounter cos the hospital in terms of staff time, supplies, etc.), as opposed to charges or reimbursement. Given that hematoma may be a common and sometimes benign event, a sensitivity analysis excluding hematoma from the overall composite of complications through 90 days was also conducted. Statistical Analyses Continuous variables were summarized using means and standard deviations, while categorical/binary variables were presented as frequencies and percentages. Stable balancing weighting, an optimization-based analogue to propensity score weighting, was used to balance the cohorts on multiple patient and hospital baseline characteristics. Optimization-based stable-balancing weight methods have been shown to provide ample flexibility with respect to pre-specification of covariate balance goals, and to result in better post-weighting covariate balance with larger effective sample sizes compared to propensity score weighting [15, 16]. The threshold for absolute standardized mean differences (SMD) was set at |0.1 [17]. Variables for adjustment were selected empirically based on what was available in the PHD database, clinical expert opinion, and observed statistical associations. Covariates used for weighting included age, sex, race, ethnicity, payor type, CCI score, select individual comorbidities (i.e., MI, CHF, chronic pulmonary disease, RA, liver disease, diabetes, diabetes with complications, and renal disease), BMI category, discharge year, teaching hospital status, and hospital bed size. Generalized linear models (GLM) with a log link and gamma error distribution were used to compare index costs; a logit link and binomial error distribution was used for binary outcomes; a log link and negative binomial error distribution was used for LOS. All GLMs accounted for clustering at the hospital level with robust standard errors. A p-value of < 0.05 was set as the threshold for statistical significance. Statistical analyses were performed using StataSE 16 (64 bit). As the study spanned multiple years, during which healthcare cost inflation would influence the magnitude of the hospital cost outcomes, all costs were adjusted to 2022 constant US dollars using the Medical Care component of the Bureau of Labor Statistics (BLS) Consumer Price-Index (CPI) measure of inflation [18]. Results A total of 145 patients in two hospitals from one health system met the study selection criteria for the KTCD robotic-assisted hand-sewn gastrojejunostomy cohort; 6,491 patients met the criteria for the non-KTCD cohort. After stable balancing weighting, the effective sample size for the control cohort was 3,050 patients. Table 1 presents the baseline characteristics of patients after stable-balancing weighting (see Table S1 for baseline characteristics before weighting). All covariates were balanced after weighting, with a SMD |≤0.1|. Most patients were aged 35–44 (KTCD 35.2%, non-KTCD 30.6%) or 45–54 (KTCD 30.3%, non-KTCD 31.2%) years, female (KTCD 81.4%, non-KTCD 84.9%), White (KTCD 49.7%, non-KTCD 51.6%) or Black (KTCD 42.1%, non-KTCD 38.1%), not Hispanic (KTCD 87.6%, non-KTCD 83.4%), had commercial health insurance (KTCD 76.6%, non-KTCD 71.6%), and had a CCI score of 0 (KTCD 53.8%, non-KTCD 49.9%) or 1–2 (KTCD 40.7%, non-KTCD 43.9%). Most patients had a BMI between 40.0 and 59.9 mg/kg2 (KTCD 69.7%, non-KTCD 70.5%; overall, 28.7% with BMI 40.0–44.9 mg/kg2, 20.9% with BMI 45.0–49.9 mg/kg2, and 20.9% with BMI 50.0–59.9 mg/kg2). Table 1Patient baseline demographic and clinical characteristics of patients with KTCD vs. non-KTCD gastrojejunostomy in robotic-assisted RYGB (after weighting)KTCD (n = 145)Non-KTCD (n = 3050*)Standardized Mean Difference**Age Category, n (%)18–3417.2%19.9%0.06634–4435.2%30.6%0.10045–5430.3%31.2%0.01855–6413.1%14.8%0.04765–744.1%3.4%0.03375+0.0%0.1%0.023% Female81.4%84.9%0.100Race, n (%)White49.7%51.6%0.041Black42.1%38.1%0.100Other/Unknown8.3%10.3%0.058Ethnicity, n (%)Hispanic7.6%9.2%0.044Not Hispanic87.6%83.4%0.100Other/Unknown4.8%7.4%0.100Payor Type, n (%)Commercial76.6%71.6%0.100Medicare9.0%10.2%0.036Medicaid13.8%14.9%0.028Other0.7%3.3%0.100CCI Category, n (%)053.8%49.9%0.0771 or 240.7%43.9%0.0643+5.5%6.2%0.024CCI Diagnoses, n (%)Myocardial infarction1.4%0.9%0.046Congestive heart failure0.7%2.3%0.100Chronic pulmonary disease18.6%22.6%0.096Rheumatoid arthritis2.1%2.2%0.011Liver disease2.1%5.6%0.100Diabetes20.7%23.6%0.067Diabetes with complications8.3%6.3%0.079Renal disease2.8%3.5%0.040BMI Category, n (%)Normal weight: 20 to < 250.0%0.0%0.014Overweight: 25 to < 301.4%1.1%0.100Obesity: 30 to < 356.2%4.7%0.100Obesity: 35.0 to 35.92.1%2.5%0.027Obesity: 36.0 to 36.92.8%2.5%0.017Obesity: 37.0 to 37.92.8%3.4%0.033Obesity: 38.0 to 38.95.5%3.5%0.100Obesity: 39.0 to 39.93.4%4.0%0.026Obesity: 40.0 to 44.928.3%28.7%0.010Obesity: 45.0 to 49.920.0%20.9%0.021Obesity: 50.0 to 59.921.4%20.9%0.011Obesity: 60.0 to 69.95.5%5.5%0.001Obesity: ≥700.7%0.8%0.018Obesity: ≥30***0.0%1.4%0.100Year of Discharge, n (%)202017.2%21.4%0.100202142.8%38.2%0.100202240.0%36.9%0.06520230.0%3.5%0.100Hospital Teaching Status, n (%)Teaching32.4%37.2%0.100Hospital Bed Size, n (%)100–500 Beds67.6%62.9%0.100500 + Beds32.4%37.1%0.100BMI body mass index, CCI Charlson Comorbidity Index, CHF, KTCD Knotless Tissue Control Devices, MI, RA, RYGB Roux-en-Y gastric bypass* Effective sample size after weighting** The threshold for absolute standardized mean differences was set at 0.1. No differences exceeded the threshold after weighting*** Some patients receive a coarser classification based on the diagnosis code used Table 1Patient baseline demographic and clinical characteristics of patients with KTCD vs. non-KTCD gastrojejunostomy in robotic-assisted RYGB (after weighting)KTCD (n = 145)Non-KTCD (n = 3050*)Standardized Mean Difference**Age Category, n (%)18–3417.2%19.9%0.06634–4435.2%30.6%0.10045–5430.3%31.2%0.01855–6413.1%14.8%0.04765–744.1%3.4%0.03375+0.0%0.1%0.023% Female81.4%84.9%0.100Race, n (%)White49.7%51.6%0.041Black42.1%38.1%0.100Other/Unknown8.3%10.3%0.058Ethnicity, n (%)Hispanic7.6%9.2%0.044Not Hispanic87.6%83.4%0.100Other/Unknown4.8%7.4%0.100Payor Type, n (%)Commercial76.6%71.6%0.100Medicare9.0%10.2%0.036Medicaid13.8%14.9%0.028Other0.7%3.3%0.100CCI Category, n (%)053.8%49.9%0.0771 or 240.7%43.9%0.0643+5.5%6.2%0.024CCI Diagnoses, n (%)Myocardial infarction1.4%0.9%0.046Congestive heart failure0.7%2.3%0.100Chronic pulmonary disease18.6%22.6%0.096Rheumatoid arthritis2.1%2.2%0.011Liver disease2.1%5.6%0.100Diabetes20.7%23.6%0.067Diabetes with complications8.3%6.3%0.079Renal disease2.8%3.5%0.040BMI Category, n (%)Normal weight: 20 to < 250.0%0.0%0.014Overweight: 25 to < 301.4%1.1%0.100Obesity: 30 to < 356.2%4.7%0.100Obesity: 35.0 to 35.92.1%2.5%0.027Obesity: 36.0 to 36.92.8%2.5%0.017Obesity: 37.0 to 37.92.8%3.4%0.033Obesity: 38.0 to 38.95.5%3.5%0.100Obesity: 39.0 to 39.93.4%4.0%0.026Obesity: 40.0 to 44.928.3%28.7%0.010Obesity: 45.0 to 49.920.0%20.9%0.021Obesity: 50.0 to 59.921.4%20.9%0.011Obesity: 60.0 to 69.95.5%5.5%0.001Obesity: ≥700.7%0.8%0.018Obesity: ≥30***0.0%1.4%0.100Year of Discharge, n (%)202017.2%21.4%0.100202142.8%38.2%0.100202240.0%36.9%0.06520230.0%3.5%0.100Hospital Teaching Status, n (%)Teaching32.4%37.2%0.100Hospital Bed Size, n (%)100–500 Beds67.6%62.9%0.100500 + Beds32.4%37.1%0.100BMI body mass index, CCI Charlson Comorbidity Index, CHF, KTCD Knotless Tissue Control Devices, MI, RA, RYGB Roux-en-Y gastric bypass* Effective sample size after weighting** The threshold for absolute standardized mean differences was set at 0.1. No differences exceeded the threshold after weighting*** Some patients receive a coarser classification based on the diagnosis code used Patient baseline demographic and clinical characteristics of patients with KTCD vs. non-KTCD gastrojejunostomy in robotic-assisted RYGB (after weighting) BMI body mass index, CCI Charlson Comorbidity Index, CHF, KTCD Knotless Tissue Control Devices, MI, RA, RYGB Roux-en-Y gastric bypass * Effective sample size after weighting ** The threshold for absolute standardized mean differences was set at 0.1. No differences exceeded the threshold after weighting *** Some patients receive a coarser classification based on the diagnosis code used Table S2 shows the unadjusted number and percentage of patients with each individual complication before weighting. After adjustment, the proportion of patients with complications was lower among patients with KTCD vs. non-KTCD at index (KTCD 0.7%, non-KTCD 4.5%; p = 0.004), within index through 30 days (KTCD 0.7%, non-KTCD 7.3%; p < 0.001), and within index through 90 days (2.1% vs. 8.3%; p < 0.001) (Fig. 1). Findings were consistent when removing hematoma from the 90-day complication composite (sensitivity analysis), with a lower proportion of patients with KTCD having complications vs. non-KTCD (KTCD 1.4%, non-KTCD 7.7%; p < 0.001). The difference between the groups in the proportion of patients with anastomotic leak within 90 days was not statistically significant (KTCD 1.4%, non-KTCD 2.0%; p = 0.143) (Fig. 1). Fig. 1Adjusted proportions of patients with complications with KTCD vs. non-KTCD gastrojejunostomy in robotic-assisted RYGB Adjusted proportions of patients with complications with KTCD vs. non-KTCD gastrojejunostomy in robotic-assisted RYGB Adjusted index LOS was lower for patients with KTCD vs. alternative means of anastomosis (1.34 vs. 1.62 days; p = 0.002) (Fig. 2). The difference in adjusted index costs of patients with KTCD vs. non-KTCD was not statistically significant (KTCD $17,042, non-KTCD $16,431; p = 0.494) (Fig. 3). Fig. 2Adjusted mean LOS of patients with KTCD vs. non-KTCD gastrojejunostomy in robotic-assisted RYGB Adjusted mean LOS of patients with KTCD vs. non-KTCD gastrojejunostomy in robotic-assisted RYGB Fig. 3Adjusted mean index costs of patients with KTCD vs. non-KTCD gastrojejunostomy in robotic-assisted RYGB Adjusted mean index costs of patients with KTCD vs. non-KTCD gastrojejunostomy in robotic-assisted RYGB Conclusion This study evaluated and compared fully hand-sewn gastrojejunostomy with KTCD vs. alternative means of anastomosis in RYGB. The study focused on real-world complications and LOS outcomes since, at the time of this study, the IFU for STRATAFIX™ Spiral Antibacterial KTCD included a statement that the safety and effectiveness had not been established for use in gastrointestinal anastomoses [3]. To our knowledge, this is the first study comparing fully hand-sewn gastrojejunostomy with KTCD vs. alternative means of anastomosis in RYGB. In an increasingly cost-conscious healthcare environment, this study highlighted that KTCD may result in a lower rate of complications and shorter LOS vs. other means of anastomosis in RYGB. While the reduction in adjusted index length of stay associated with KTCD use was statistically significant, the absolute difference of 0.28 days may be of limited clinical relevance for individual patients. However, in the context of the U.S. healthcare system, small reductions in length of stay can translate into meaningfully improved hospital efficiency when applied across large patient populations. RYGB surgeries can lead to serious complications including anastomotic leaks, bleeding, infection, obstructions, and subsequent anastomotic strictures [19–21]. Using KTCD in these procedures may facilitate the preservation of tissue tension and maintenance of approximation of the incised or dissected tissues without the need for tying knots and time-consuming insertion [22, 23]. Combining the benefits of KTCD barbed sutures with hand-sewn anastomosis may offer the best results and least complications [24–26], aiming to achieve optimal outcomes in complex procedures like RYGB. The results of the current study are consistent with the few previous studies that have described the use of barbed sutures in RYGB. A case report of hand-sewn gastrojejunostomy using bidirectional KTCD found no leaks or stenoses at 6-month follow-up [27]. An evaluation of barbed sutures for 96 totally hand-sewn anastomoses during laparoscopic gastric bypass (48 patients) found the suture-related complication rate was comparable with data in the literature (n = 2 leaks from the jejunojejunal anastomosis, n = 0 leaks from gastrojejunostomy, n = 2 stenoses of the gastrojejunal anastomosis, and n = 0 bleeding complications) [28]. A retrospective review of patients who underwent laparoscopic RYGB with a hand-sewn gastrojejunostomy using knotless unidirectional barbed absorbable suture (n = 46) or standard monofilament absorbable suture (n = 38) found no significant difference in the overall complication rate between the two groups (standard 18.4% vs. barbed 13%; p = 0.77) [29]. No complications were secondary to the jejunojejunostomy closure or gastrojejunostomy and no anastomotic leak or stenosis occurred in either group. A strength of this study is that it used observational healthcare utilization data from diverse clinical settings which reflect real-world conditions. Data from a large hospital administrative database were supplemented by clinical details collected from the unstructured surgical notes by medical chart review to improve the rigor of the study. Additionally, this study included comprehensive variables around patient and hospital characteristics. Another strength is that the study used stable balancing weights, an optimization-based analogue to propensity score weighting, to balance the cohorts on the numerous baseline characteristics. This statistical methodology preserves patient data allowing for inclusion of outliers and more representative outcomes (compared to matching which can result in a loss of the sample) [15–17]. A limitation of the study is that the KTCD cohort comprised only 145 patients from 2 hospitals. There are also limitations inherent to database analyses, including potential coding errors or misclassifications within the database, which could affect accuracy. Real-world administrative data lack granular clinical variables, and unmeasured variables may lead to residual confounding after adjusted analyses and may impact study outcomes; therefore, as this was not a randomized clinical trial, findings should be interpreted as associations rather than causation. The study is also limited by the nature of the PHD, as it is only able to capture hospital readmissions to the same hospital, possibly underestimating actual complication rates and costs, though this is unlikely to differ across the study comparators. The present study evaluated fully hand-sewn gastrojejunostomies with KTCD vs. alternative means of anastomosis, including sutures or stapling; however, it did not compare alternative suture materials against KTCD – which remains an important aim for future research. Finally, the study results may not necessarily be generalizable to all hospitals in the US or globally. Despite these limitations, the study provides valuable insights into the complications and healthcare resource utilization associated with the fully hand-sewn gastrojejunostomies using KTCD in robotic-assisted RYGB. In this retrospective analysis of patients in the PHD undergoing robotic-assisted RYGB, we found that the fully hand-sewn gastrojejunostomy using KTCD was associated with a lower incidence proportion of 90-day complications, lower LOS, and similar costs as compared with alternative means of anastomosis such as non-KTCD sutures or stapling. The findings indicate that KTCD barbed sutures may offer a safe and effective alternative to conventional suturing and mechanical stapling techniques in gastrojejunal anastomoses without introducing additional risks or costs. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1