Our study included 180 consecutive patients who had undergone SG (n=50), RYGB (n=53) and OAGB (n=77) at the Clinic for General and Visceral Surgery at the Kepler University Clinic Linz between January 1, 2016 and December 31, 2018. Retrospectively, the clinical and demographic data of the study participants, surgery duration and incidence of operative complications were taken from the hospital information system and collected in the obesity database Mazimoi ODS (Bariatric Patient Documentation and Data Analysis). 127 of the patients were interviewed by phone in October 2019; 53 patients could not be reached by phone. A standardized questionnaire was filled out for each patient, recording responses on postoperative dumping symptoms and their satisfaction with the procedure. The Sigstad Score was used to evaluate dumping. We did not perform any provocative tests in order to assess dumping symptoms in a way which is more relevant to daily practice.
Individual patient satisfaction was assessed by giving a school grade (1=very good, 2=good, 3=satisfactory, 4=sufficient, 5=not sufficient) and asking whether the patient would undergo the intervention again. Patient inclusion for bariatric surgery was based on the criteria of the National Institutes of Health Development Panel (Body Mass Index (BMI) >40 kg/m² or BMI >35 kg/m² with at least one obesity-associated comorbidity) [2]. Patients who have not met the NIH criteria or had incomplete preoperative clarification were excluded. Pregnancy was also not compatible with bariatric surgery. Patients who had previous abdominal surgery or had already undergone bariatric surgery before (e.g. gastric band), were also enrolled in the study. Further exclusion criteria for admission to the study were an incomplete follow-up protocol and any other bariatric procedure except SG, RYGB and OAGB. Preoperatively all candidates were evaluated by a multidisciplinary medical unit and underwent preoperative nutritional consultation and psychological, and comprehensive medical evaluations. A detailed assessment was performed of their general condition, comorbidities, risk factors, mental status, motivations for bariatric surgery, compliance and ability to adhere to a postoperative regimen. Biochemical and radiological studies (chest x-ray, upper GI series) as well as endocrine and cardiopulmonary assessment, were performed. The study was conducted after approval from the ethics committee and Institutional Review Board (Ethics committee protocol number: 1025/2020).
Variables
All variables were analysed at baseline (perioperative values) and included gender, age, BMI, preoperative weight, type of surgery, operative time and complications. Patients were evaluated 20.0 ± 12.1 months after SG, 20.0 ± 12.1 months after RYGB and 20.0 ± 10.7 months after OAGB during a phone interview which included filling out a standardized questionnaire after patient’s consent. Incidence of dumping syndrome and satisfaction were monitored and possible association between diabetes and dumping and satisfaction and dumping was tested. Complications were defined as minor in cases where no surgical reintervention was necessary (Clavien Dindo grade 1 or 2). Complications were defined as major when patients had to undergo surgical reexploration (Clavien Dindo grade 3 or higher). Operation time was defined as the beginning of the skin incision to completion of the surgical dressing. Sigstad Score was used to evaluate dumping (Table 1). A score of 7 and above was considered positive for dumping syndrome. Informed consent was obtained from all of the patients.
Statistical analysis
All data was evaluated using IBM SPSS Statistics 24.0 (SPSS Inc. Chicago, IL, USA) and then processed and graphically displayed in Microsoft Excel. By means of descriptive statistics, mean values, standard deviations, medians and ranges were calculated. An intention to treat approach as well as a per-protocol approach has been taken. All data of continuous variables were checked for normal distribution using Kolmogorov-Smirnov-test with Lilliefors significance correction. Comparisons (SG vs. RYGB vs. OAGB) of variables with normally distributed data without different variances were performed by a parametric analysis of variance (ANOVA). For comparisons of all other continuous variables and of variables measured on ordinal scales a non-parametric analysis of variance (Kruskal Wallis test) was used. Data of categorical variables were compared by the exact chi-square test. The absolute and relative frequencies of individual parameters were compared using frequency tables and displayed in crosstabs. Correlations were tested using Fisher's exact test. A significance level of 0.05 was applied to all statistical tests.
Operation techniques
In all patients standardized operation techniques were used and all procedures were performed laparoscopically under general anesthesia by the same surgeon. The operations were either primary bariatric surgery for 157 patients (87.2%) or reoperations following gastric banding in 23 patients (12.8%). Anastomosis integrity was verified after every procedure with an intraoperative endoscopic pneumatic-water test.
SG: Five ports were used. Three 12-mm ports were located in the epigastric, right hypochondriac and left lumbar region. A 5-mm port was placed subxiphoideal and a 15-mm port in the right lumbar region. A longitudinal resection from the angle of His to approximately 6cm orally to the pylorus was performed using a linear stapler (iDrive® with Tri-staple cartridges, Medtronic, USA). A 34-Fr bougie was used for calibration of the gastric tube and inserted along the lesser curvature.
RYGB: 5 ports were placed. One in the right hypochondriac region (12-mm), left hypochondriac region (12-mm), epigastrium (12-mm), subxiphoideal (5-mm), and left paraumbilical (12-mm). A gastric pouch was performed, calibrating it using a 34-Fr bougie, with a linear stapler (iDrive® with Tri-staple cartridges, Medtronic, USA). A 140-cm biliary limb and a 80-cm alimentary limb were performed. Gastrostomy was done using a 30-mm linear stapler. The enterotomies and gastrotomies were sutured with Vicryl 3/0 SH. Petersen’s space was always closed.
OAGB: 5 ports were placed in the same positions as in RYGB. A long and slim gastric pouch, calibrated with a 34-Fr bougie, was constructed. Termino-lateral gastrojejunal anastomosis with 30-mm linear stapler (iDrive® with Tri-staple cartridges, Medtronic, USA) was performed. The afferent loop was sutured up to the long stomach pouch with three Prolene sutures (anti-reflux sutures) and the draining loop is sutured to the antrum of the remnant stomach with another Prolene 2.0 suture with extracorporeal slip knots. The enterotomies and gastrotomies were closed with continuous barbed suture V-Loc 2/0 (Medtronic, USA). The biliopancreatic limb length ranged between 140 and 180 cm. We choose not to close the Petersen’s space when performing OAGB.