Complications in Bariatric Surgery with Focus on Gastric Bypass

Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine 1676

Complications in bariatric surgery with focus on gastric bypass

BJARNI VIDARSSON

ACTA UNIVERSITATIS UPSALIENSIS ISSN 1651-6206 ISBN 978-91-513-0991-0 UPPSALA urn:nbn:se:uu:diva-417549 2020 Dissertation presented at Uppsala University to be publicly examined in H:son Holmdahlsalen, Entrance 100/101, Akademiska Sjukhuset, Uppsala, Friday, 9 October 2020 at 09:00 for the degree of Doctor of Philosophy (Faculty of Medicine). The examination will be conducted in Swedish. Faculty examiner: Associate Professor Jacob Freedman ( Department of Clinical Sciences, Danderyd Hospital, Karolinska Institutet, Stockholm).

Abstract Vidarsson, B. 2020. Complications in bariatric surgery with focus on gastric bypass. Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine 1676. 56 pp. Uppsala: Acta Universitatis Upsaliensis. ISBN 978-91-513-0991-0.

Obesity is rising in pandemic proportions. At present, one third of the world’s population has become overweight or obese, and estimates predict 60% in 2030. Thus, the problem is gigantic. Obesity is associated with numerous diseases such as diabetes, high blood pressure, sleep apnea and cancer. Untreated obesity decreases life expectancy by about 10 years. Gastric bypass has been one of the cornerstones of surgical treatment. Since 1994 this is done by laparoscopic technique (LRYGB) In this thesis, we have primarily used data from our national quality register, the Scandinavian Obesity Surgical Registry (SOReg), on patients that have been operated with LRYGB. In the first paper, we evaluated the use of a novel suture for closing the gastrojejunostomy (upper anastomosis). Paper II and III focused on incidence, risk factors, treatment and outcome of anastomotic leaks. Paper IV compares the weight results, quality of life, use of medications and healthcare consumption in patients suffering from a serious complication within 30 days after LRYGB. In Paper I, the use of the barbed suture resulted in shorter operative time compared to a standard polyfilament, without increased risk for complications. Paper II showed that the incidence of anastomotic leaks at the gastrojejunostomy was 0.6%. Risk factors were male sex, higher age (≥49 years), diabetes, conversion to open surgery and prolonged operative time (≥ 90 minutes). Almost all patients were reoperated and 1% died. Paper III showed that the incidence of small bowel leaks was 0.3% and these leaks were associated with prolonged operative time, and surgery at a low-volume centre for leaks at the enteroaenteral anastomosis. Surgical re-intervention was common. Paper IV showed that severe complications within 30 days postoperatively after LRYGB occurred in 2.9% of cases. Two years later, the patients still reported inferior quality of life and had a higher use of antidepressants, proton pump inhibitors and opioids compared to uncomplicated cases. The need for additional in-hospital care was higher, even after the first 30 days. In conclusion, the novel barbed suture reduced operative time without increasing risks. Anastomotic leaks are rare, but serious complications in LRYGB do affect the patient in numerous ways and increase healthcare costs.

Keywords: Barbed suture; Operative time; Complications; LRYGB; Gastric bypass; RYGB; Leaks; Reoperation; Enteroenteral anastomosis; Small bowel; Reintervention; Healthcare consumption; Roux-en-Y gastric bypass; RAND-36; Quality of life

Bjarni Vidarsson, Department of Surgical Sciences, Upper Abdominal Surgery, Akademiska sjukhuset ing 70 1 tr, Uppsala University, SE-751 85 Uppsala, Sweden.

ISSN 1651-6206 ISBN 978-91-513-0991-0 urn:nbn:se:uu:diva-417549 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-417549)

To my beloved family Harpa, Viðar Elí, Daníel and Óskar

List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I Bjarni Vidarsson, Magnus Sundbom and David Edholm. (2017) Shorter overall operative time when barbed suture is used in primary laparoscopic gastric bypass: A cohort study of 25,006 cases. Surgery for Obesity and Related Diseases. 13(9):1484- 1488.

II Bjarni Vidarsson, Magnus Sundbom and David Edholm. (2019) Incidence and treatment of leak at the gastrojejunostomy in Roux-en-Y gastric bypass: A cohort study of 40,844 patients. Surgery for Obesity and Related Diseases. 15(7):1075-1079.

III Bjarni Vidarsson, Magnus Sundbom and David Edholm. (2020) Incidence and treatment of small bowel leak after Roux-en-Y gastric bypass: A cohort study from the Scandinavian Obesity Surgery registry. Surgery for Obesity and Related Diseases.16 (8):1005-1010.

IV Bjarni Vidarsson, Martin Löfling Skogar and Magnus Sund- bom. Impact of a severe complication two years after laparoscopic Roux-en-Y gastric bypass: A cohort study from the Scandinavian Obesity Surgery registry. Manuscript

Reprints were made with permission from the respective publishers.

Contents

Introduction ...... 11 Background ...... 14 Definitions of overweight and obesity ...... 14 Risk factors of obesity ...... 14 Gastric bypass ...... 14 Early complications in Roux-en-Y Gastric Bypass (RYGB) ...... 15 Anastomotic leak ...... 16 Suture types ...... 17 Other complications in RYGB ...... 17 Grading complications by the Clavien Dindo system ...... 18 Operative time ...... 19 In-hospital stay ...... 19 Weight loss after RYGB ...... 19 Effect on quality of life ...... 19 Quality of life after RYGB ...... 20 Aims ...... 21 Material and methods ...... 22 Paper I ...... 22 Statistics ...... 22 Paper II ...... 22 Statistics ...... 23 Paper III ...... 23 Statistics ...... 23 Paper IV ...... 24 Statistics ...... 24 Ethics ...... 24 Results ...... 25 Paper I ...... 25 Paper II ...... 26 Paper III ...... 30 Paper IV ...... 31

Discussion ...... 36 Paper I ...... 36 Paper II ...... 37 Paper III ...... 38 Paper IV ...... 39 Strength and limitations ...... 41 Conclusions ...... 42 Paper I ...... 42 Paper II ...... 42 Paper III ...... 42 Paper IV ...... 42 Future projects ...... 43 Populärvetenskaplig sammanfattning på svenska ...... 45 Arbete I ...... 45 Arbete II och III ...... 46 Arbete IV ...... 46 Acknowledgements ...... 48 References ...... 50

Abbreviations

ANOVA Analysis of variance ATC Anatomic Therapeutic Chemical Classification system ASMBS American Society for Metabolic and Bariatric Surgery BMI Body Mass Index BS Barbed suture BV Bjarni Vidarsson CD Clavien Dindo CI Confidence interval CPAP Continuous positive airway pressure DDD Defined daily dose EA Enteroenteral anstomosis EWL Excess weight loss GSRS Gastrointestinal Symptom Rating Scale GJ Gastrojejunostomy HADS Hospital Anxiety and Depression Scale ICU Intensive Care Unit ICD International Statistical Classification of Diseases and Health Related Problems LABS Longitudinal Assessment of Bariatric Surgery LRYGB Laparoscopic Roux-en-Y gastric bypass LOS Length of stay MCS Mental component summary MME Morphine milligram equivalent NPR National Patient Register OME Oral morphine equivalent OR Odds Ratio OSA Obstructive sleep apnea syndrome PCS Physical component summary PIN Personal identification number PPI Proton pump inhibitors PS Polyfilament suture RYGB Roux-en-Y gastric bypass SC Serious complication SOReg Scandinavian Obesity Surgery Registry SOS Swedish Obese Subjects TWL Total weight loss QoL Quality of life

Introduction

Obesity has become a worldwide disease and undergoing bariatric surgery is the best possibility for the severely obese to achieve, and maintain, substantial weight loss (1). Bariatric surgery produces a significant and more durable weight loss compared to any other non-surgical treatment modality (2, 3). Roux-en-Y gastric bypass (RYGB) has been a common procedure for the treatment of severe obesity in Sweden for many years. At present, sleeve gastrectomy comprises about 48% of all bariatric surgery in Sweden followed by laparoscopic gastric bypass (47%), according to the Scandinavian Surgery Obesity Registry (SOReg) (4). SOReg is a nationwide quality registry containing more than 75,000 bariatric procedures done at the 38 bariatric centres in Sweden. SOReg has high coverage; indeed, 99.1% of all of bariatric surgeries in Sweden are registered (5), and the registry has been supported by The National Board of Health and Welfare and the Swedish Association of Local Authorities and Regions since the start in May 2007 (5). In SOReg, defined variables are registered preoperatively, at surgery, and at follow up after 6 weeks and 1, 2 and 5 years. Va- lidity has been found to be high with a positive predictive value of 99.7% for registration of performed procedures (6). All registrations are based on the unique personal identification number (PIN), routinely used in the Swedish health care system, to allow cross-matching to other official registers, such as the National Inpatient Register, containing all given in-patient care, and the Prescribed Drug Register. These registries are run by The National Board of Health and Welfare and considered to have 100% coverage (7). Today, almost all the surgeries (99%) are done with laparoscopy, compared to about 75% in 2007 (4). Laparoscopy facilitates the postoperative course, particularly for the length of in-hospital stay and postoperative con- valescence (8). A low incidence of postoperative complications, both in the short and long term, is of considerable importance when operating benign high-volume surgery, in addition to a good postoperative weight result. SOReg data show that the RYGB technique is improving with shorter operative time, shorter length of stay and reduced complication frequencies. The learning curve of the procedure has been eased thanks to the standardization of the procedure in Sweden, using the double omega loop method to create the pouch and the two anastomoses (9).

Figure 1. Schematic description of a Roux-en-Y Gastric Bypass with a gastrojejunostomy and an enteroenteral anastomosis.

It is important to continuously lower the complications frequency to minimize the individual patient suffering as well as overall cost for bariatric surgery in Sweden. Therefore, we have studied the risk of complications when constructing the two anastomoses in RYGB on a nationwide basis. In Paper I, ‘Shorter overall operative time when barbed suture is used in primary laparoscopic gastric bypass‘, we studied a cohort of primary laparoscopic RYGB patients in which two different types of sutures were used to close the defect after constructing the linear stapled gastrojejunostomy (upper anastomosis). We studied differences between the groups for overall operative time, early complications and hospital stay. The novel barbed suture shortened the operative time without increasing the risk of complications compared to conventional polyfilament suture. In Paper II, ‘Incidence and treatment of leak at the gastrojejunostomy in Roux-en-Y gastric bypass‘, we retrospectively investigated all patients having a leak at the gastrojejunostomy (GJ) within 30 days after primary RYGB. GJ leaks occurred in 0.6% of patients and risk factors for leak were male sex, higher age, diabetes, prolonged operative time and conversion to open surgery. In Paper III, ‘Incidence and treatment of small bowel leak after Roux-en- Y gastric bypass‘, we focused on non-GJ leaks as most surgeons seldom experience this complication and there are no randomized studies to guide us in the selection of potential treatment options. Incidence of small bowel leaks was 0.3%. Risk factors were prolonged operative time and surgery at a

12 low volume centre for enteroenteral anastomosis (EA, lower anastomosis) leak. In Paper IV, ‘Impact of a severe complication two years after laparoscopic Roux-en-Y gastric bypass‘, patients with a severe complication after laparoscopic RYGB were compared to controls with the help of SOReg data and three national registers. The results showed that a serious complication leads to inferior improvement in quality of life (QoL) and higher use of antidepressants, proton pump inhibitors (PPI) and opioids. Additional in-hospital care was increased during the first two years.

13 Background

Definitions of overweight and obesity Body Mass Index (BMI, kg/m2) can be used to measure variation from ideal weight. Normal BMI lies between 18.5 and 24.9, overweight between 25 and 29.9 and obesity above 30. Obesity is usually further subgrouped into class 1 obesity with BMI between 30 and 34.9, class 2 with BMI between 35 and 39.9 and class 3 with BMI above 40. Class II is usually referred to as severe obesity and class III as morbid obesity or super obesity). According to Swe- dish guidelines, patients with BMI above 35 are eligible for bariatric surgery (10). Other methods for measuring overweight is waist circumference and waist-hip ratio. Increased central obesity has been linked to increased mortality in patients with coronary artery disease (11).

Risk factors of obesity Overweight affects over one-third of the world’s population today (1.9 billion adults and children) and this proportion has tripled since 1975 (12). If the trend continues about 58% of the world population could be overweight or obese (3.3 billion people) by the year 2030 (13). Overweight and obese individuals run a higher risk of premature death due to diseases such as stroke, coronary heart disease, congestive heart failure and cardiomyopathy (14). Mortality is also associated to obesity per se (15-17). The risk of hyperlipidaemia, hypertension, type 2 diabetes and sleep apnea is increased, in turn leading to shorter life expectancy compared to non-obese individuals (18-20).

Gastric bypass Gastric bypass for the treatment of obesity was first described by Edward E Mason in 1967 (21). The pouch was then created by dividing the fundus hori- zontally and a loop gastrojejunostomy (GJ) was added (22). The procedure has evolved through the years and now a small pouch (15-30 mL) is created on the lesser curvature and a Roux limb from the proximal jejunum is anas- tomosed to the pouch (23) with a linear stapler (Figure 2). This alteration decreases the bile reflux into the pouch(24).

Figure 2. Construction of the gastrojejunostomy with a linear stapler during a laparoscopic RYGB.

The enteroenteral anastomosis (EA, lower anastomosis between the distal part of the Roux limb and the biliopancreatic limb) is the second anastomoses that created in a gastric bypass operation. In the laparoscopic approach, this anastomosis is usually done with a linear staple line between two jejunal bowel segments. The remaining opening, after the stapler has been removed from the gut, is closed manually with a suture of the surgeon’s preference (Figure 3). Bile, pancreatic and gastric juices merge with indigested nutrients at the enteroenteral anastomosis often placed 70-100 cm distal to the GJ. This construction makes it possible for the patient to lose substantial amounts of weight without risking nutritional deficiencies.

Figure 3. Construction of the enteroenteral anastomosis with a linear stapler during a laparoscopic RYGB.

Early complications in Roux-en-Y Gastric Bypass (RYGB) Early complications, such as bowel obstruction, bleeding, leakage and deep infections are rare in RYGB. In the SOReg material, leaks and deep infections/abscesses occur in less than 1% (4) and general complications such as deep vein thrombosis, pulmonary embolus and other pulmonary and cardiovascular complications occur in <0.5%. However, all of these complications can have serious outcome for patients.

15 Anastomotic leak The gastrojejunostomy (upper anastomosis between the small gastric pouch and proximal end of the Roux limb) is the most common location for anastomotic leak in RYGB (25, 26). Other leak sites include the enteroenteral anastomosis and iatrogenic perforations in the small- and large bowel, excluded stomach and esophagus. The incidence of leak from the GJ in RYGB is about 1-2% (27, 28) and it has not been found to differ between open and laparoscopic operative technique in a randomized trial (8). According to data from SOReg, anastomotic leak occurs in 0.6% of patients operated with laparoscopic RYGB in Sweden (4). An early leak can be a consequence of a technical error in the construction of the anastomosis; however, a leak occurring after a few days is more likely associated to patient-related factors and tissue healing. The management of anastomotic leaks has been suggested to include prompt surgery and drainage (26). Many emphasize the importance of observing postoperative patients for early signs of leakage such as tachycardia and accentuating upper abdominal pain postoperatively. The use of endo- scopically applied stents has been discussed and might have its place in certain patient categories (29, 30).

Figure 4. Patient with a suspected leak in the gastrojejunostomy at postoperative day 1. Injection of methylene blue dye into the gastric pouch confirms the leakage site. After suture of the defect with a barbed suture, no leakage of methylene blue was observed.

Leak at the EA is rare (0.2% in our results) and thus more unusual than the feared GJ leak, despite the fact that they are probably as dangerous (31). The introduction of laparoscopic approach to RYGB in the 90s facilitated the perioperative phase, allowing more operations to be done per day. The new minimal invasive operative technique also eased the patient’s postoperative period. The introduction of staplers to divide tissue and construct anastomoses is probably one of the factors contributing to the explosion in gastric bypass operations done worldwide: 685,000 procedures in the year 2016 (32). Interestingly, the stapling of the GJ has not been shown to decrease operative time or complications in a meta-analysis from 2016 (33).

16 Suture types Polyfilament suture is one of many common bowel sutures. It is a synthetic, often braided, suture that loses 75% of its strength in four weeks. It is easy to handle and can be grabbed with a laparoscopic instrument (grasper or needle driver) without the risk of fragmenting the suture. The new barbed suture has, as the name implies, numerous barbs on the side that make the suture unidirectional (Figure 5). The barbs anchor to the tissue at different points making the need for a final knot unnecessary. Stud- ies have shown that its use has not increased risk of complications and that it shortens the operative time (34, 35), which is in line with our results in Paper I.

Figure 5. Barbed suture, the barbs make the suture unidirectional and eliminates the need for knot tying.

Other complications in RYGB Anastomotic stricture is a well-known complication in RYGB. The most common location is at the GJ and it often presents 4-6 weeks after surgery. The incidence in Sweden is low, 0.3% at one year (36) but can be underre- ported because other studies have reported higher numbers (37-39). The treatment is endoscopic dilatation and prescription of proton pump inhibitors. In extreme refractory cases a resection of the pouch with a direct esophagojeju- nal anastomosis is required. Stricture at the EA is uncommon and usually caused by a technical error or a kinking in the making of the anastomosis.

Internal hernias are a common complication to RYGB and have been seen in up to 6% of patients in a study with a 50-month follow up(40). There are two potential internal hernia sites that are created in the making of an antecolic laparoscopic RYGB (LRYGB). One is the space between the mesentery of the biliopancreatic and the Roux limb. The other is the so-called “Petersen

17 space” between the mesenteries of the Roux limb and the colon. The most common symptoms are colicky abdominal pain caused by the entrapment of the bowel within the internal opening. This can, in extreme cases, cause bowel infarction and is almost always an indication for explorative laparoscopy or open exploration. A negative CT scan should not rule out the diagnosis of an internal hernia if the patient shows symptoms of pain out of proportion. The needed treatment is reposition of the bowel and closure of the hernia defect, commonly with a non-absorbable suture.

Grading complications by the Clavien Dindo system The Clavien Dindo system is a complication-grading instrument that was presented in 1992, initially grading the complications after cholecystectomy (41). The system has since then been improved to more accurately be able to grade complications in other surgeries (42). The system grades complications from 1 to 5 where 1 stands for “any deviation from the normal postoperative course without the need for intervention” and 5 stands for “death of a patient”. This system is frequently used when measuring and comparing postoperative complications.

Table 1. Clavien Dindo complication grading system. Grade Intervention 1 Any deviation from the normal postoperative course without the need for pharmacological treatment or surgical, endoscopic and radiologic interventions. 2 Requiring pharmacological treatment with drugs other than such allowed for grade I complications. Blood transfusions, antibiotics and total par- enteral nutrition are also included. 3a Requiring surgical, endoscopic or radiological intervention under regional or local anaesthesia. 3b Requiring surgical, endoscopic or radiological intervention under general anaesthesia. 4a Life-threatening complication requiring intensive care unit management, single organ failure. 4b Life-threatening complication requiring intensive care unit management, multi-organ failure. 5 Death of patient

18 Operative time The operative time for primary laparoscopic bariatric surgery in SOReg is now 57 minutes. When the register was started in 2007, operative time was twice as long, about 120 minutes (4). The reason for the successful shortening of operative time is probably more-experienced surgeons, better equipment, standardized operative technique, and that we passed the early learning curve.

In-hospital stay A laparoscopic operative technique leads to a shorter in-hospital stay compared to an open operation for the same procedure (43). Length of stay was short, 1.5 days on average, for all bariatric surgeries in SOReg in 2017 (4). Length of stay depends on the experience of the whole surgical team as well as staff at the surgical ward. It is also affected by the postoperative dietary regime and the rate of postoperative complications.

Weight loss after RYGB RYGB has a restrictive effect and increases early satiety by alteration in hormonal levels. These two mechanisms in combination cause the massive weight loss after surgery (44). Studies on RYGB show that patients can expect between 50% and 80% excess weight loss i.e., reduction of former weight above BMI 25 (45-48). Thus, the average patient operated in Sweden with BMI 42 will reach a BMI of about 28-33. In clinical practice, we are cautious to not “guarantee” a specific amount of weight loss.

Effect on quality of life Quality of life (QoL) is difficult to measure and can be demanding to compare between individuals. The available methods include specific questionnaires for general and disease specific QoL. One of the most used questionnaires for general QoL, the RAND-36, comprises 36 short questions with categorical answers (49). The responses are classified into 8 domains and two summary scores for physical and mental components. RAND-36 is validated for the use in patients with obesity (50, 51). It is frequently used in evaluating the effects of bariatric surgery (49). The Swedish version of RAND-36 is used in SOReg and validated for the Swedish population (52). Disease-specific questionnaires like the Obesity-related Problems scale (OP), developed to measure the impacts of obesity on psychosocial functioning (53) have, however, not been used in this thesis.

19 Quality of life after RYGB Bariatric surgery and long-lasting weight reduction improves patient scored QoL (53). This is probably achieved through less body pain, improved self- esteem and improved confidence. It is also known that patients who suffer from a complication after colorectal surgery have decreased QoL compared to those without complications (54). In Paper IV, we studied general QoL in patients that had suffered from a serious complication after primary laparoscopic RYGB.

20 Aims

The aim of the thesis is to assess the impact of two early complications after gastric bypass surgery and to evaluate whether the use of a novel type of suture influences the complication risk.

The specific aims are: I To compare the use of a novel barbed suture to standard techniques in closing the remaining opening in a linear stapled gastrojejunostomy for operative time, length of stay and surgical complications such as leak, small bowel obstruction and anastomotic stricture

II To assess incidence, risk factors, treatment and outcome of leaks at the gastrojejunostomy after RYGB

III To assess incidence, risk factors, treatment and outcome of leaks at the enteroenteral anastomosis and iatrogenic small bowel perforations after RYGB

IV To compare the weight result, quality of life, use of antidepressants, proton pump inhibitors, opioids, and healthcare consumption in laparoscopic RYGB patients suffering a severe complication to a matched cohort.

21 Material and methods

Paper I We identified 25,006 patients in SOReg that had undergone primary LRYGB with a linear stapled gastrojejunostomy (GJ). The type of suture that was used to close the defect after stapling the anastomosis decided whether the patient was put in the barbed suture group (BS) (2211 patients) or the polyfilament group (PS) (22,795 patients). Centres with 10 or fewer patients were excluded to minimize the effects of the learning curve. SOReg definitions were used for the classification of complications.

Statistics Student’s t test was used for parametric data and chi2 test for proportions. A p< 0.05 was considered statistically significant.

Paper II Primary RYGB procedures in SOReg between 2007 and 2014 were included, a total of 40,844 patients. Almost all (97%) were operated with laparoscopy. When we had identified all patients with leak, the peri- and postoperative charts were collected and reviewed. We were able to locate 94% of the charts. After the revision, 262 leaks at the GJ were verified and used for further analysis (Figure 6).

Figure 6. Consort diagram that shows how the 262 study patients were selected.

Statistics Logistic regression was done to estimate odds ratios (OR). Student’s t-test and analysis of variance (ANOVA) were used for parametric data and chi2 test was used to compare proportions. A p< 0.05 was considered statistically significant. Age and operative time were divided into quartiles.

Paper III In this retrospective cohort study, 41,342 primary RYGB procedures (97% laparoscopic) from 2007-2014 were included. As in Paper II, the peri- and postoperative charts of all patients with a registered postoperative leak were collected and reviewed to verify the leak and determine its exact location. Patients were divided into two groups based on the location of the leak, at the enteroenteral anastomosis or in the small bowel.

Statistics Logistic regression, based on important clinical risk factors from the univariate analyses, was done to estimate odds ratios (OR) for significant risk factors for leak. Student’s t-test and ANOVA were used to compare parametric data and Chi2 test was used to compare proportions. Age and operative time were divided into quartiles. A p<0.05 was considered statistically significant.

23 Paper IV In this retrospective cohort study, we included 48,201 primary LRYGB operated from 2007 to June 2016. Information about postoperative complications within 30 days was retrieved from SOReg and the most severe complication determined the Clavien-Dindo (CD) score(55). A Clavien-Dindo score ≥3b was considered a severe complication (SC). A 1:1 case-control matching without replacement was done. Data on weight loss and QoL were gathered from SOReg. Medication use was gathered from the Prescribed Drug Register. We studied prescribed antidepressants, proton pump inhibitors and opioid analgesics. Opioids were furthermore converted into average daily oral morphine equivalents (OMEs) using the morphine equivalent conversion factor per milligram for each specific opioid(56). Comparisons between the groups were done at six time points. Data on healthcare consumption were extracted from the National Patient Register (NPR), containing data from 1st January 1997 to 31st of De- cember 2016. All types of in-hospital care were studied up until two years after the primary procedure. Principal diagnoses were used to classify the care events according to the main chapters of the International Statistical Classification of Diseases and Related Health- problems (ICD-10).

Statistics Independent-samples t-test and ANOVA were used to compare parametric data between groups and Chi2 test was used to compare proportions. Paired- samples t-test was used for comparisons within groups and McNemar’s test was used for comparison of proportions within groups. A p<0.05 was considered statistically significant.

Ethics Paper I: The study was approved by the regional ethical review board in Stockholm (Dnr 2013/535-31/5).

Paper II and III: The studies were approved by the regional ethical review board in Uppsala (Dnr 2015/481).

Paper IV: The study was approved by the regional ethical review board in Stockholm, Sweden (Dnr: 2013/535-31/5 and 2017/857-32).

24 Results

Paper I The barbed suture group and polyfilament group were similar at baseline except for BMI and diabetes, as can be seen in Table 2.

Table 2. Data of the 25,006 included primary laparoscopic gastric bypass patients grouped according to suture type used for the gastrojejunostomy.

Barbed Polyfilament p-value suture (BS) (PS) n=2211 n=22,795 Baseline Gender (male) 23.4% 24.8% 0.14 Age (years) 41.0 ±11.6 40.9 ±11.2 0.76 Body Mass Index (kg/m2) 41.3 ±5.1 42.3 ±5.2 <0.001 Diabetics 13.7% 15.4% 0.03 Abbreviations: BS - barbed suture, PS - polyfilament suture

Total operative time in the BS group was 58 minutes compared to 69 minutes in the PS group (p<0.001), which is a difference of 16%. Not all patients were, however, operated with same type of suture in both anastomoses and the operative time for the four different combinations are seen in Table 3.

Table 3. Details on the type of suture used for the gastrojejunostomy and enteroenteral anastomosis. Gastro- Entero- N Operative time jejunostomy enterostomy (min) BS BS 1932 57 BS PS 279 67 PS BS 360 81 PS PS 22,435 69 Abbreviations: BS – barbed suture, PS - polyfilament suture

25 No difference in complication rate was observed between the groups. Table 4.

Table 4. Operative data and complications grouped according to suture type used for the gastrojejunostomy.

Barbed suture Polyfilament p-value (BS) (PS) n=2211 n=22,795 Operative data Mesenteric openings closed 94.3% 70.6% <0.001 Leak test performed 99.3% 99.3% 0.76 Total operative time (min) 58 ± 22 69 ± 33 <0001 Length of stay (days) 1.5 ± 2.8 1.9 ± 2.6 <0.001 Early complications Leak or abscess (n) 1.8% (39) 1.4% (319) 0.17 Complications at one year Stricture at one year (n) 0.13% (2) 0.17% (32) 0.73 Small bowel obstruction 1.8% (27) 1.6% (311) 0.69 (n) Abbreviations: BS – barbed suture, PS - polyfilament suture

Paper II Leak at the GJ was confirmed in 262 patients (0.6%). Patients with leaks were more often male (36% vs 24%), older (44.5 years vs 40.9 years) and had diabetes more often (26% vs 14%) (p<0.001 for all) than patients who did not develop a GJ leak. Hypertension and hyperlipidaemia were also more prevalent in the leak group, 33% vs 25% (p<0.01), and 15% vs. 10% (p<0.01). Waist circumference was somewhat higher in the leak group, 130.4 cm vs 126.5 cm (p<0.001).

Operative risk factors Mean operative time for patients with subsequent leaks was 97.7 ± 54.7 min compared to 74.1 ± 36.8 min (p<0.001) for patients that did not develop a leak. In addition, operative time was longer in patients that developed an early leak (within three days) compared to patients developing a leak later, 109 min vs 89 min (p=0.004). Patients with early leaks were somewhat older than those suffering late leaks (46 vs 43 years, p=0.04). The leak rate occur-

26 ring at high volume centres did not differ from low volume centres (0.5% vs 0.8%, p=0.24). A leak occurred in 0.6% (240/39,455) and 0.7% (8/1085) of the laparoscopic and open procedures, respectively. In the converted group, 4.6% (14/304) suffered from a leak (p<0.001).

Time of diagnoses Almost half (44%) of the leaks were identified within 3 days of surgery (early leaks) with most leaks diagnosed on day 2. Of all patients with leaks, 56% had been discharged and were readmitted due to their leak.

Figure 7. Percentage of leaks diagnosed by a) postoperative day and b) mean annual number of procedures at each bariatric centre where the line separates high and low volume centres (<125 cases/year).

In the multivariate analysis, risk factors for developing a leak at the GJ after RYGB were male gender, older age, diabetes, conversion to open surgery and prolonged operative time Table 5.

28 Table 5. Multivariate analysis of risk factors predisposing for developing a leak at the gastrojejunostomy within 30 days after Roux-en-Y gastric bypass.

Variable Total n (%) Number of Odds 95% Conf. leaks (%) Ratio Interval

Gender

Female 30,915 (76%) 167 (64%) Ref Male 9929 (24%) 95 (36%) 1.47 1.13 - 1.90

Age (years)

≤32 9827 (24%) 41 (16%) Ref 33 to 40 9739 (24%) 48 (18%) 1.11 0.73 - 1.69 41 to 48 10,601 (26%) 67 (26%) 1.31 0.88 - 1.94 ≥49 10,677 (26%) 106 (40%) 1.83 1.25 - 2.67

Comorbidity No diabetes 34,943 (86%) 195 (74%) Ref Diabetes 5901 (14%) 67 (26%) 1.43 1.06 - 1.93

Surgical access Laparoscopic 39,455 (96%) 240 (92%) Ref Open 1085 (3%) 8 (3%) 0.73 0.35 – 1.49 Converted 304 (1%) 14 (5%) 3.88 2.19 – 6.88

Operative time <50 minutes 10,004 (24%) 39 (15%) Ref 50-65 minutes 10,502 (26%) 40 (15%) 0.96 0.62 – 1.50 66–89 minutes 9828 (24%) 58 (22%) 1.40 0.96 – 2.16 ≥90 minutes 10,508 (26%) 125 (48%) 2.60 1.79 – 3.76

Treatment and outcome

Surgical reintervention was done in 85% of patients, and simultaneously, the defect was sutured in 45%. A feeding gastrostomy was placed in 24% of patients. Stents were used in 31% of GJ leaks. Overall, drainage was placed in 89% of patients under either general or local anaesthesia. Six percent of all leaks were treated conservatively. Almost all leaks (n=235) lead to a severe complication (Clavien Dindo ≥3b), i.e. intervention under general anaesthesia (64%), intensive care (25%) or death (1%). The two deaths that occurred were caused by cardiac arrest shortly after the leak was diagnosed (0-2 days). Mean duration of stay after a leak was 22 days compared to 2 days for those that did not have leak (p<0.001). The construction of a new anastomosis was associated with increased duration of stay, 58 days (p<0.001). Pa- tients treated with stents also required a prolonged duration of stay, 30 days vs 18 days for non-stented patients (p=0.001).

29 Paper III Leaks at the EA were confirmed in 75 patients (0.2%) and small bowel perforations in 54 patients (0.1%). In the univariate analyses, patients with EA leak and small bowel perforations had higher waist circumference and longer operative time than patients who did not develop a leak. EA leaks were more frequent in patients operated in low-volume centres, 0.3%, compared to 0.1% in high-volume centres (p<0.001). In the multivariate analysis, risk factors for developing an EA leak were surgery at a low-volume centre (OR 2.1 [1.0 – 4.1]) and prolonged operative time (> 90 min, OR 3.5 [1.1–11.0]). Risk factor for a small bowel leak was prolonged operative time (> 90 min, OR 3.4 [1.2–9.4]) Table 6.

Table 6. Multivariate analysis of risk factors predisposing for developing a leak at the enteroenteral anastomosis (EA, n=75) or being treated for a small bowel perforation (n=54) within 30 days after Roux-en-Y gastric bypass.

EA leak Small bowel perforation

Variable Total (%) Number of OR 95% CI Number of OR 95% CI leaks (%) leaks (%) Gender Female 30 748 (76%) 50 (67%) Ref 36 (67%) Ref Male 9834 (24%) 25 (33%) 1.0 0.5 – 2.2 18 (33%) 0.9 0.4 – 1.9 Age (years) ≤32 9786 (24%) 17 (22%) Ref 10 (18%) Ref 33 to 40 9691 (24%) 14 (19%) 0.5 0.2 - 1.5 9 (17%) 0.9 0.3 - 2.4 41 to 48 10,534 (26%) 18 (24%) 0.4 0.2 - 1.3 16 (30%) 1.4 0.6 - 3.4 ≥49 10,571 (26%) 26 (35%) 1.4 0.6 - 3.1 19 (35%) 1.71 0.7 – 4.0 Waist cir- 33,116 53 1.0 0.9 – 1.1 43 1.0 0.9 – 1.1 cumf. (cm) Op. time <50 min 9964 (24%) 9 (12%) Ref 6 (11%) Ref 50-65 min 10,461 (26%) 17 (23%) 3.1 0.9 - 9.8 16 (30%) 2.2 0.8 – 6.5 66–89 min 9769 (24%) 15 (20%) 1.70 0.5 – 5.0 10 (18%) 1.9 0.6 – 5.6 >90 min 10,388 (26%) 34 (45%) 3.5 1.1 - 22 (41%) 3.4 1.2 – 9.4 11.0 Op. volume High 26,030 (64%) 33 (44%) Ref 34 (63%) Ref Low (<125 14,552 (36%) 42 (56%) 2.1 1.0 - 4.1 20 (37%) 1.0 0.5 – 1.9 cases /year) Smoking 4204 2 (3%) 0.4 0.9 – 1.5 - - -

30 Treatment and outcome Almost half (45%) of the EA leaks and 85% of the small bowel perforations were identified within 3 days of surgery (early leaks), with most leaks (33%) diagnosed on postoperative day 1. Surgical reintervention was done in more than 97% of the patients, evenly distributed between laparoscopic, open and converted operations. When specified in the charts, EA leaks were more common in the suture line (31%) than the staple line (20%). Likewise, the small bowel perforations were mainly believed to be caused by grasping forceps (41%). During the reintervention, the defect was sutured in 75% of EA leaks and in 93% of small bowel perforations. Drainage was placed in 86% and 72% of the two groups at the end of the procedure, while 43% and 44%, respectively, also received a gastrostomy. Almost all leaks (97-98%) lead to a severe complication (Clavien-Dindo ≥3b) and intensive care was needed in 35% of patients with EA leaks and in 24% of patients with small bowel perforations. In the group with EA leaks, one patient died on the third postoperative day due to septic shock. LOS was 18 days for patients with EA leaks and 13 days for patients with small bowel perforations, compared to two days for the remaining patients (p<0.001). One year after RYGB, the weight result in patients with EA leaks or small bowel perforations did not differ from the remaining patients.

Paper IV We identified 1411 (2.9%) patients having had a severe complication. Six- teen patients (0.03%) had a fatal complication (CD 5) and were not studied further. In total, 1403 SC patients were successfully matched to 1403 controls (42 years, 75% females, 15.7% diabetics and BMI 41.9 kg/m2). The eight remaining patients, which did not meet matching criteria, were excluded.

Complications and treatment The most common complications within 30 days postoperatively were bowel obstruction 32.4% (n=454), leak 29.6% (n=415) and bleeding 26.5% (n=372). The overall 30-day reoperation rate was 94.2% (n=1321), while intensive care treatment (CD ≥4a) was needed in 6.8% (n=96) and 1.2% (n=17) suffered multiorgan failure (CD 4b).

Weight loss Two years after LRYGB, the weight result in the SC group did not differ from the controls.

31 Quality of life The physical component summary (PCS) was improved at two years after surgery in both groups; however, the SC group had lower scores than controls at all time points (p<0.01). When analysing the eight domains in the RAND-36, both groups showed an improvement in all domains (p<0.01) except for role emotional in SC group. Furthermore, the SC group had lower scores at two years compared to controls in all domains (p<0.05) but mental health (p=0.34) (Figure 8).

Baseline (Severe complication) 2 years (Severe complication) Baseline (Controls) 2 years (Controls) Physical function 100 Mental health 80 Role physical 60 40 20 Role emotional 0 Bodily pain

Social function General health

Vitality

Figure 8. Patient scored quality of life at baseline and two years after surgery.

Medications Antidepressants No differences between groups were observed preoperatively nor during the first year after surgery. After an increase in both groups, the proportion of patients taking antidepressants at 2 years was somewhat higher in the SC group (31% vs 26% for controls, p=0.01) (Figure 9).

Figure 9. The proportion of patients using antidepressants before and after surgery in SC and control group (±95% CI).

Proton pump inhibitors At all time points, including the preoperative period, the proportion of patients taking PPI was higher in the SC group (p<0.01 for all). In both groups, a transient increase was observed during the first 90 days after operation. At two years, a higher proportion of patients were on PPI compared to before surgery (SC 35% vs 15% and controls 20% vs 10%, both p<0.01) (Figure 10).

Figure 10. The proportion of patients using PPI before and after surgery in SC and control group (±95% CI).

33 Opioids Prior to surgery, the use of opioids did not differ between the groups. After surgery, and throughout the 2-year follow up, the proportion of patients using opioids was higher in the SC group (p<0.05 for all time points). Moreo- ver, an increase in mean OMEs was observed in the SC group, from 7.3 mg/day prior to surgery to 17.0 mg/day at two years (Figure 11).

Figure 11. Opioid use as mean oral morphine equivalents per day (mean ±% CI)

34 Healthcare consumption We had access to data from the National Patient Register on 1364 SC patients and 1308 controls. Length of stay, within the first 30 days after surgery, was 5.6 and 1.8 days for SC and controls, respectively (p<0.01). Dur- ing the remaining part of the first year, the SC group had more additional hospital admissions (0.6 vs 0.2) as well as in-hospital days (3.8 vs 0.9 days) compared to controls. This was also true for the second year. The total number of additional hospital admissions from >30 days to 2 years were 1331 for SC and 516 for controls (p<0.01). The SC group had more in-hospital care, based on principal diagnoses, in 9 of the 18 ICD chapters (Figure 12).

Figure 12. The total number of additional hospital admissions from >30 days to 2 years for severe complications and controls. The main diagnoses are classified according to the International Statistical Classification of Diseases and related health Problem (ICD). Significant difference between groups is marked with ** (p<0.01).

35 Discussion

Bariatric surgery is the only method for the obese patient to lose weight in the long term. It is also known that by losing weight diseases such as diabetes and high blood pressure can go into regress, which increases overall survival in the operated patient. Therefore, it is of considerable value to be able to help those patients with an operation that does not decrease their quality of life due to a complication. We have shown in short that barbed suture used in gastric bypass surgery does not increase the complication rate and that gastric bypass surgery in Sweden can be considered safe. However, patients that suffer from a complication have lower quality of life compared to controls.

Paper I The total operative time was 11 minutes shorter when the novel barbed suture was used to close the defect after a stapled anastomosis in LRYGB. This is probably because the barbed suture eliminates the need for manual knot tying, a technically demanding task in LRYGB. A similar decrease in operative time (74 to 63 minutes and 125 to 119 minutes) has been found in other single centre trials (57-59). The small difference in BMI (41.3 vs 42.3) and diabetes (15.4% vs 13.7%) between the two groups is unlikely to affect the operative time or incidence of complications. On the other hand, closing of the mesenteric openings was more common in the BS group, which should increase operative time considering it takes about 2-3 minutes for each opening(60). No difference was observed in leak rate between the groups, which is in concordance with the earlier published studies (57, 59, 61). We believe that a meticulous surgical technique and allowing the surgeon to use the operative technique with which he or she is most familiar is most important. Although the stricture rate is lower in a linear stapled GJ compared to circular anastomosis(62, 63), the switch to a new type of suture could be problematic. Howev- er, the present stricture rate was not increased when BS was used. When comparing our results to others, it is important to remember that only symptomatic strictures, leading to an upper endoscopy, were reported and registered.

36 Small bowel obstruction after the use of BS has been described(64, 65) but our data from the groups did not differ. To minimize this risk, the loose end of the BS should be cut close to the bowel. In our opinion, small bowel obstruction is probably most commonly due to internal hernias or kinking of the enteroenteral anastomosis, rather than a long end of the BS becoming entangled in the bowel in turn causing obstruction.

Paper II A leak incidence of 0.6% at the GJ was observed, which is low compared with other studies (3-4%) (29, 66). Notably those studies have included all leaks after RYGB, not only leaks at the GJ. We were able to define the following risk factor for leak: male gender, older age, diabetes, conversion to open surgery and long operative time. Male gender and age as a risk factor is supported in previous papers. This may be because surgery in men with increased abdominal obesity is technically demanding and older patients have impaired tissue healing as well as increased number of comorbidities (29, 67, 68). Similarly, patients with diabetes are reported to be more prone to suffer from leak due to increased risk for infection and decreased tissue healing(68). Conversion to open surgery increases the risk for leak as does long operative time (>66 minutes). This is probably because the perioperative difficulties requiring the conversion continued during open surgery and leads to a suboptimal anastomosis. Long operative time per se is a proxy for technical difficulties as shown by Reames et al (69). Most patients suffering from leak required surgical intervention and 89% of patients had drainage. Moreover, about one-third of the present leaks were treated with stents, a feasible nonoperative approach (29, 70), keeping in mind that stents lead to their own complication panorama, such as migration and potential erosion. We believe that the mainstay of patients will require surgical management to achieve adequate drainage, a belief in line with the ASMBS guidelines(71). One-fourth of our patients needed treatment in the intensive care unit. This proportion differs to that found in other studies, which probably depends on local traditions and the distribution of high risk patients between private and public centres (26, 30). Mortality was 1% (2 patients), which we consider low. In the literature, there is a wide spread from 0% to 17% in cohorts from 1980 to 2000(25, 29, 68). Excess weight loss at one year was the same for the two groups. This was expected because both groups had the same postoperative anatomy, giving them the same possibility of achieving the expected weight loss after gastric bypass.

37 Paper III EA and small bowel leak after RYGB is a severe threat to the bariatric patient (68). The incidence of leak was low, 0.2% and 0.1% respectively, and in line with the literature (0.1-0.2%) (72, 73).

Risk factors In the multivariate analysis, the risk factors associated with EA leak is surgery at a low-volume centre and prolonged operative time, which agrees with other studies(74, 75). Prolonged operative time also tripled the risk for undiagnosed iatrogenic small bowel perforations. It has been shown that long operative time predisposes for leak(69) likely because it is a proxy for technical difficulties. The increased technical demands and poor visibility, due to the large amount of intraabdominal fat in bariatric surgery, can easily increase the risk of traumatic injuries to the small bowel by grasping forceps. The pressure of a grasping forceps on the bowel can be as high as 600 kilopascal (kPa), depending on the angle, which is similar to the pressure generated by a high heeled shoe when all weight is put on one heel (76). Moreover, it is important to avoid unnecessary traction on the small bowel. Interestingly, none of the present leaks occurred in patients having had a converted operation, i.e., surgeons converting had used a wise safety-first management. In the literature, damage to the small bowel is often reported to be caused by the blind insertion of a Veress needle or trocar, or by an energy device(77-79). In line with previous studies (67, 68) BMI was not found to be a risk factor for leak. Possibly patients with high BMI get preferential surgical treatment by a more senior surgeon than patients with a BMI just above 35 kg/m2. Fernandez et al found obstructive sleep apnea to be a risk factor for gastrointestinal leaks after open gastric bypass (68); however, we were unable to con- firm this in our cohort possibly due to the tenfold difference in leak rate (3.2 vs 0.3% in the present study). We were also unable to verify that smoking is a risk factor for leaks in this large nationwide study (80); despite this, Swedish patients are diligently counseled to stop smoking preoperatively. In elective colorectal surgery, preoperative antibiotics, used in all our patients, have been shown to reduce anastomotic leak (81); however, the anastomotic leak rate is much lower in small bowel anastomoses.

Choice of treatment and outcome The earlier diagnosis of iatrogenic small bowel perforations (3.6 vs 6.5 days) in the EA leaks is probably due to a larger defect in the bowel, thus leading to a larger spill of bowel contents. Earlier studies have shown that two-thirds of traumatic perforations are diagnosed within 48 hours (77) which is in line with our results (74% diagnosed within 48 hours). Most patients (97%) with a leak had surgical reintervention and were thus treated in line with ASMBS

38 guidelines (71). Closing the defect and adequate drainage is essential in man- aging a leaking EA or small bowel perforation. In our experience, an adequate way to manage a suspected small bowel leak is early diagnosis, prefer- ably by laparoscopy, followed by an intervention that minimizes the spill of bowel contents into the abdominal cavity. Unfortunately, our data are unsuf- ficient to help us decide whether to close the defect or to create a totally new anastomosis. Treatment in the intensive care unit (ICU) was required by 35% of patients with an EA leak and by 24% of those with small bowel perforation. This agrees with other studies, reporting the need for ICU treatment ranging from 20-67% (82, 83). In our cohort, one patient (1%) died after an EA leak in the third postoperative day, while the overall 30-day mortality in SOReg is 0.03%. In the literature, mortality from postoperative leaks range from below 1% (25, 29) to 17%, in a study from 2004 (68). The most plausible explanation for decreasing mortality is improved patient care, with earlier diagnosis and treatment, as a result of increased experience. The present LOS for patients suffering from leaks and perforations (18 and 13 days, respectively) was shorter than that Marshall et al reported (33 days) in 2003 (29, 84). The shorter LOS for patients with small bowel perforations diagnosed within three days indicates that early detection of signs and symptoms, e.g. tachycardia, fever and abdominal pain, of leak (83) can reduce patient morbidity and hospital expenses. It is important to remember that patients with leaks will eventually have the same postoperative anatomy as complication-free patients, giving them the same opportunity to achieve the expected postoperative weight loss. This outcome may provide some consolation to patients suffering from a leak, while they endure the extended hospital stay and various leak-related procedures. However, the leak will probably lead to increased risk of future small bowel obstruction due to intra-abdominal adhesions.

Paper IV We found the incidence of serious complications to be 2.9% and in line with the literature (85) as were the most common complications: bowel obstruction, bleeding and leakage (86). Treatment in the ICU was required by 7% which is lower than in other studies. The difference may partly be due to local rou- tines and the level of care provided in ordinary wards.

Quality of life Quality of life was considerably improved in both groups according to the frequently used RAND-36 scale (49). However, the PCS increased more in the control group and almost reached the normal score of 50 for the general population (87). Weight loss was strongly correlated with improved physical

39 health as early as in the late 1990s by the Swedish Obese Subjects (SOS) study (53). Patients who suffer from a surgical complication have lower QoL compared to controls (88, 89) a result our data verified.

Medications Antidepressants reflect an increased anxiety and mental affection, PPI acid- related conditions, e.g. ulcers and reflux, due to a suboptimal anatomical result and opioids reflect chronic postoperative pain. These medications were therefore used as proxies for those symptoms. The use of antidepressants increased in the end of the follow-up period in both groups, resulting in a higher use in the SC group compared to controls. These results are in contrast to the Longitudinal Assessment of Bariatric Surgery (LABS) study, in which a decrease was observed, from 35.3% at baseline to 27.5% at 3-year follow-up (90). However, although we show an increase in the medication for depression, our 2-year incidence (29%) is similar to LABS and Cunningham et al. (91). Proportion of patients using PPI doubled after surgery in both groups and a higher consumption was observed in the SC group at all time points. We have earlier shown that an increased use of PPI preoperatively is associated with an increased risk for complications (92). The reason for this is unknown but speculated about in more complicated surgery to be due to hiatal hernia or a gastrointestinal vulnerability at large. The transient increase in PPI use until 3 months postoperatively can partly be due to local guidelines recom- mending PPI as prophylaxis for marginal ulcers although this effect has been questioned in Sweden (93). The greater prescription of opioid analgesics in the SC group probably re- flects more reoperations, thus increasing the risk for both acute and chronic abdominal pain. An increased opioid consumption after RYGB is observed in other studies, even in patients that have not suffered from a complication (94-96). Furthermore, the opioid abuse factor must also be kept in mind as the risk for abuse increases after surgery, e.g., alcohol disorders increase after bariatric surgery (97, 98).

Healthcare consumption The SC group had both more hospital admissions and in-hospital days than the control group. The difference at 30 days was expected, and in line with other studies (10 in-hospital days after LRYGB with a complication) (99). The increased need for in-hospital care due to conditions in the gastrointestinal system, such as small bowel obstruction, hernias and psychiatric diseases was expected.

40 Strength and limitations An overall strength of all studies lies in the use of SOReg data, resulting in large cohorts with nationwide coverage as well as high validity (98%). The follow up at 30 days is high (99%) and in Paper II and III we have also been able to review the charts of 94% of the patients suffering from a leak, a task done according to a predefined protocol and by a single examiner (BV). Moreover, the nationwide use of an almost identical operative technique, the laparoscopic double omega-loop technique (100), produces homogenous perioperative and complication data. In Paper IV, the matching of three national registries with high coverage reduces the risk of missing patients treated for a complication at another centre. The limitations lie in the nature of retrospective register studies where centres might underreport data on patients suffering from complications such as leaks. Lack of detailed information on differences in case mix and numbers of procedures for individual surgeons can vary between centres. Furthermore, the low incidence of certain complications makes it difficult to show both associations and statistical differences between groups. Moreover, as a diagnostic laparoscopy is preferred to a CT scan in some centres, the presented number of early reoperations could be unnecessarily high. Our determination of quality of life in Paper IV lacks disease-specific questionnaires, such as the Gastrointestinal Symptom Rating Scale (GSRS)(101) and the Hospital Anxiety and Depression Scale (HADS)(102).

41 Conclusions

Paper I The use of barbed suture shortened the operative time, without increased risk of complications. Barbed suture is still uncommon in LRYGB, but its im- plementation could significantly affect operating room time, which is bene- ficial considering the high volume of bariatric surgeries.

Paper II Leaks at the gastrojejunostomy occurred in 0.6% of patients. Risk factors were older age (≥49 years), male gender, diabetes, conversion to open surgery and prolonged operative time (>66 minutes). Most patients had surgical interventions and one-quarter needed intensive care. Length of stay was increased by 20 days. Mortality was low, 1%.

Paper III Leak at the EA occurred in 0.2% of cases, while 0.1% suffered from an iatrogenic small bowel perforation, undiagnosed at the primary procedure. Risk factors were prolonged operative time and operation at a low-volume centre for EA leaks, tripling the risk. Most patients had surgical re-interventions, including closure of the defect and drainage of the abdominal cavity. None- theless, 3 in 10 patients needed intensive care, but mortality was below 1%. Length of stay was increased by more than two weeks.

Paper IV The incidence of a severe complication was 2.9%. A severe complication resulted in an inferior improvement in quality of life, higher use of antidepressants, PPI and opioids as well as higher need of in-hospital care, even after the first 30 days. Risk of affected patients needing treatment at the intensive care unit was 7% and the mortality was low, 0.03%.

42 Future projects

As demonstrated in this thesis, complications after gastric bypass surgery are rare and can therefore be challenging to study. To determine whether treatment methods differ, we would, in a perfect world, do a randomized study with well-defined end points. It will be challenging to include and start the randomizing process, when confronted with an acute complication and the patient is in no condition to be enrolled in such a project. Because of the low incidence of the specific complications, descriptive studies such as those carried out will probably be used to better optimize treatment on a case by case basis. Therefore, making our registers more robust and more accurate seems to me the best way to collect reliable data to study the infrequent, but dangerous complications. However, as shown in Paper IV, it is important to extend such studies to several years after the complication.

The following questions have become arisen from my work:

1. Is there enough evidence to be more alert when operating older male patients with diabetes? Our results state that there is an increased risk of leaks in this group of patients. Yet, does this mean that these patients should be operated by a more senior surgeon or should we care for these patients differently in the perioperative period. This needs to be studied further in national multicentre trials.

2. Can leaks be diagnosed earlier? New ideas have been launched that have focused on early detection of leaks after gastrointestinal surgery with measurement of hydrogen in the abdominal cavity (103). This might help to prevent delay in the treatment and diagnosis of leaks but further studies are needed, particularly because meticulous routine manual abdominal examination is difficult in the bariatric patient group.

3. Can we further reduce the complication rates? Preventive measures to avoid serious complications, such as obligatory sim- ulator training, centralization of difficult cases and continuous education and research, are probably the best alternatives for the individual patient and the

43 healthcare system. These factors could, however, be difficult to study, particularly in the short term.

4. Should patients that suffered from a serious complication receive special- ized postoperative surveillance? As shown in Paper IV, no major improvement was observed in the complication-related variables at two years. At present, these patients are in routine follow up. We could increase the number of annual outpatient visits or increase the psychological support in order to increase the overall quality of life after their bypass surgery. The effect of this could be studied in a multiprofessional study.

5. Are other procedures better? Other bariatric procedures such as sleeve gastrectomy and minigastric bypass are becoming more common. Sleeve gastrectomy is now the most common bariatric surgical method in Sweden, but it is unclear whether it is a preferable method. The ongoing Bypass Equipoise Sleeve Trial (BEST) compares gastric bypass to sleeve gastrectomy in a randomized, multicentre study with substantial complications and total body weight loss as primary endpoints (104). If sleeve gastrectomy is less prone to cause complications perhaps gastric bypass will lose its role as the gold standard of bariatric surgery, but in order to determine this we need good, robust data.

6. Are we as surgeons choosing the correct fight? Lastly, one can start to wonder whether the fight against obesity and related- diseases can be won with surgery when estimates tell us that 60% of the world’s population will be overweight or obese in less than 10 years. This calls for multiple studies on new types of conservative treatments ranging from socioeconomic interventions to targeted drug therapies. The future success in overcoming the obesity epidemic lies in multiprofessional and political cooperation.

44 Populärvetenskaplig sammanfattning på svenska

Övervikt och fetma blir alltmer vanlig i värden. Nu är mer än halva världens befolkning överviktig med ett kroppsmasseindex (Body Mass Index, BMI) på 25 eller högre. Övervikt leder till flertal olika sjukdomar till exempel diabetes, högt blodtryck och cancer, som in sin tur leder till förkortad livs- längd. I Sverige har andelen överviktiga och feta också stigit i takt med ut- vecklingen i övriga världen. Idag är nästan 60 % av svenska män och 45 % av svenska kvinnor överviktiga. Den rekommenderade behandlingen vid övervikt är kostråd och ökad fy- sisk aktivitet med målet att minska kaloriintaget och öka förbrukningen. Om dessa råd fungerar inte finns det olika medicinska behandlingar som bland annat kan minska upptag av fett i magtarmkanalen och påverka hungerkäns- lan. Vid kraftig fetma (BMI≥35) kan patienter i Sverige erbjudas kirurgi. De vanligaste ingreppen som utförs är gastric bypass och sleeve gastrektomi. Dessa operationer har blivit alltmer vanliga och det utförs omkring 5000 ingrepp om året på olika center runtom i Sverige. Fetmaoperation ger de flesta hjälp med en bestående viktnedgång genom att minska mängden föda som man kan äta och skapa hormonförändringar i kroppen som bidrar till viktminskningen I denna avhandling har användningen av en ny kirurgisk tråd och komplikationer efter fetmakirurgi granskats.

Arbete I I arbetet jämför vi två grupper av gastric bypasspatienter som opereras med titthålsteknik (laparoskopi). En grupp opererades med en ny hullingförsedd tråd som inte behöver knytas, medan den andra gruppen opererades med den tråd som normalt används. Information om grupperna samlade vi från svenska kvalitetsregistret för fetmakirurgi (SOReg). Över 25,000 patienter inkluderades och operationstid, komplikationer och vårdtid registrerades. Våra resultat visar att operationstiden är kortare i gruppen som opererades med hullingförsedd tråd och att deras vårdtid är något kortare. Komplikat- ionsfrekvensen är lika i båda grupperna.

45 Vår slutsats är att hullingförsedd sutur förkortar operationstiden utan att öka komplikationsrisken.

Arbete II och III Läckage efter kirurgi kan vara en livsfarlig komplikation hos nyopererade patienter. I arbete II och III har vi tittat på de vanligaste platserna för läckage efter gastric bypassoperation. Arbete II granskar läckage vid övre kopplingen, den som kallas gastrojejunostomi. Arbete III fokuserar på läckage från nedre anastomosen eller tunntarmskopplingen och andra läckage från tunn- tarmen, till exempel efter skada på tarmen som orsakas av operatören. Vi har registrerat alla läckage inom 30 dagar som har dokumenteras i SOReg. Granskning av alla patientjournaler har gjorts och på det sättet har den ex- akta lokalisationen av läckaget fastställts. Patienter med läckage har jämförts med patienter som inte har drabbats av läckage med målet att ta fram riskfaktorer. Läckage vid övre kopplingen förekommer i 0,6 % av alla gastric bypassoperationer i Sverige. Riskfaktorer för läckage är manligt kön, högre ålder (över 49 år), diabetes, lång operationstid och om man behövde byta till öp- pen kirurgi (operation med vanligt snitt). Läckage från kopplingen resulterade oftast i en omoperation då man vanligtvis sydde över hålet i kopplingen och lade dränage. En fjärdedel av patienterna behövde vistas på intensiv- vårdsavdelning i samband med sitt läckage. Vårdtiden förlängdes med 20 dagar och 1 % av patienterna avled. Arbete III visade att den nedre kopplingen läckte i 0,2 % av alla gastric bypassoperationer och att 0,1 % fick läckage från en tunntarmsskada. Risk- faktorer för detta är lång operationstid och om man opereras på ett center som gör färre än 125 fetmaoperationer per år. Liksom i arbete II resulterade läckagen oftast i en omoperation då man vanligtvis sydde över hålet i kopplingen och lade drän. En tredjedel av patienterna behövde vistas på intensiv- vårdsavdelning i samband med sitt läckage och 1 % av patienterna avled. Vår slutsats är att läckage är sällsynt men att de som drabbas av läckage oftast behöver omopereras.

Arbete IV Konsekvenserna av en komplikation kan på kort sikt vara allvarliga, men hur de drabbade patienterna mår på lång sikt är inte väl känt. Vi valde därför att jämföra en grupp med patienter som hade drabbats av en allvarlig komplikation med grupp som inte hade någon komplikation eller endast en mindre allvarlig komplikation (kontrollgrupp). Allvarlighetsgraden bestämdes uti- från en välbeprövad skala (Clavien Dindoskalan). Mer än 1400 patienter

46 med en allvarlig komplikation kunde jämföras med kontrollgruppen angå- ende viktminskning, livskvalitet, användning av vissa utvalda mediciner och sjukvårdsbehov. Tre olika nationella register användes för att samla in data till studien. Samma viktminskning noterades i bägge grupperna, den fysiska hälsan hos komplikationsgruppen var sämre än hos de som inte drabbades av komplikation två år efter kirurgi. Ökad användning av smärtlindrande mediciner, läkemedel som minskar magsyran (protonspumpshämmare) och antidepres- siva läkemedel noterades. Slutligen såg vi att patienter som drabbas av en allvarlig komplikation har ett ökat vårdbehov, även efter de första 30 dagar- na. Vår slutsats är att följderna av en svår komplikation påverkar patienten även två år efter operationen. Sammanfattningsvis har vi sett att hullingförsedd tråd förkortar operationstiden vid en gastric bypassoperation utan att öka risken för komplikation. Läckage och andra allvarliga komplikationer efter gastric bypassoperation är trots allt ganska sällsynta i Sverige. Om en svår komplikation uppstår leder det oftast till en omoperation med förlängning av vårdtiden men patienterna uppnår samma viktminskning som de som inte har drabbats av läckage eller annan allvarlig komplikation. Två år efter en allvarlig komplikation kvarstår dock ett ökat behov av vissa läkemedel och vårdresurser jämfört med patienter som inte drabbas av en svår komplikation.

47 Acknowledgements

I want to express my sincere gratitude and appreciation to all of you who have helped me complete this work. A special thanks to:

My main supervisor, Magnus Sundbom, for all your help, patience and that giving up does not exist in your vocabulary. You can always find the time to discuss and improve manuscripts and your door is always open. I’m fortu- nate to both been able to do research with you and learning the art of upper GI surgery. You are constantly open to new ideas and new areas to explore. Without your energy, outstanding support and expertise this work would never have been accomplished.

My co-supervisor, David Edholm, for helping me with statistics and for your good comments in the process of writing manuscripts. You are always willing to go through the text without delay and constantly giving encourag- ing remarks. I would also like to thank you for all family dinners; we miss you all.

Peter Stålberg, Head of Department of Surgery and Per Hellman, head of Department of Surgical Sciences, for giving me the opportunity and means to accomplish these studies simultaneously with my clinical work.

My co-author and colleague, Martin Löfling Skogar, for your invaluable help with statistics, scientific writing both in a manuscript and thesis produc- tion.

Gustav Linder, a wonderful colleague, and friend, always positive and ready to take on the next surgical case. We want you back.

My fellow surgeon at the Division of Esophageal and Gastric Surgery, Jakob Hedberg, who always is willing to help, give good advice or just discuss sports.

My fellow surgeons at the Division of Esophageal and Gastric Surgery, Ed- uardo Sima, Khalid Elias and Filip Möller. You make the everyday work easy and there is always room for laughter.

48 The former Head of Departement of Surgery, Claes Juhlin and Kristina Ahlgren, for giving me the opportunity to work in Uppsala and conduct scientific research.

Joakim Folkesson, for the seminars where the manuscripts have been criti- cally reviewed and my former and present fellow PhD students Ann Langerth, Christopher Månsson, Eladio Cabrera, Emmanuel Ezra, Henrik Benoni, Josef Urdzik, Josefine Kjaer, Katharina Stevens, Lana Ghanipour, Kostas Tsimogiannis, Malin Enblad, Matilda Annebäck, Petter Fruhling, Peter Moberger, Sara Artursson, Åsa Collin and Kristín Jónsdóttir for the pertinent comments and interesting discussions during the seminars which improved my work.

Steve Scott Robson my outstanding proofreader who seems, like all surgeons, to be on call around the clock. Steve - I have now learned the difference between 25 006 and 25,006. I promise!

All my fellow colleagues at the Department of Surgery, Uppsala University Hospital for all help.

My parents, Viðar and Guðmunda, for your never ending encouragement, unconditional support and love throughout the years. For that I can never thank you enough. My siblings, Sindri, Margrét Lára and Elísa with fami- lies, for your support and all the good times that we have had together, may they be more. I am proud of you.

Finally, the love of my life, Harpa and our three boys Viðar Elí, Daníel and Óskar. This work could never have been done without your support. Every- day problems become small when we are together. I love you all.

49 References

1. Fisher BL, Schauer P. Medical and surgical options in the treatment of severe obesity. The American journal of surgery. 2002;184(6):S9-S16. 2. Gloy VL, Briel M, Bhatt DL, Kashyap SR, Schauer PR, Mingrone G, et al. Bariatric surgery versus non-surgical treatment for obesity: a systematic review and meta-analysis of randomised controlled trials. Bmj. 2013;347:f5934. 3. Maggard-Gibbons M, Maglione M, Livhits M, Ewing B, Maher AR, Hu J, et al. Bariatric surgery for weight loss and glycemic control in nonmorbidly obese adults with diabetes: a systematic review. Jama. 2013;309(21):2250-61. 4. Årsrapport SOReg 2019 [Internet]. 2017. Available from: http://www.ucr.uu.se/soreg/dokument. 5. Hedenbro J, Näslund E, Boman L, Lundegårdh G, Bylund A, Ekelund M, et al. Formation of the Scandinavian Obesity Surgery Registry, SOReg. Obesity surgery. 2015:1-8. 6. Tao W, Holmberg D, Näslund E, Näslund I, Mattsson F, Lagergren J, et al. Validation of Obesity Surgery Data in the Swedish National Patient Registry and Scandinavian Obesity Registry (SOReg). Obesity surgery. 2015:1-7. 7. Ludvigsson JF, Andersson E, Ekbom A, Feychting M, Kim JL, Reuterwall C, et al. External review and validation of the Swedish national inpatient register. BMC Public Health. 2011;11:450. 8. Nguyen NT, Goldman C, Rosenquist CJ, Arango A, Cole CJ, Lee SJ, et al. Laparoscopic versus open gastric bypass: a randomized study of outcomes, quality of life, and costs. Annals of surgery. 2001;234(3):279. 9. Suzanne Johansson M, Hans Lönroth M, Torsten Olbers M. The Gothenburg technique for laparoscopic Roux-en-Y gastric bypass. Journal of the Interna- tional Bariatric Club. 2012;22:12. 10. Nationella Indikationer för obesitas-kirurgi. 2009. [Internet] Available from: http://www.sfoak.se/wp-content/niok_2009.pdf 11. Coutinho T, Goel K, Correa de Sa D, Kragelund C, Kanaya AM, Zeller M, et al. Central obesity and survival in subjects with coronary artery disease: a systematic review of the literature and collaborative analysis with individual sub- ject data. J Am Coll Cardiol. 2011;57(19):1877-86. 12. Organization. WH. Obesity and overweight 2020 [Available from: https://www.who.int/news-room/fact-sheets/detail/obesity-and-overweight. 13. Kelly T, Yang W, Chen CS, Reynolds K, He J. Global burden of obesity in 2005 and projections to 2030. Int J Obes (Lond). 2008;32(9):1431-7. 14. Eckel RH, Barouch WW, Ershow AG. Report of the National Heart, Lung, and Blood Institute-National Institute of Diabetes and Digestive and Kidney Diseases Working Group on the pathophysiology of obesity-associated cardiovascular disease. Circulation. 2002;105(24):2923-8. 15. Adams KF, Schatzkin A, Harris TB, Kipnis V, Mouw T, Ballard-Barbash R, et al. Overweight, obesity, and mortality in a large prospective cohort of persons 50 to 71 years old. New England Journal of Medicine. 2006;355(8):763-78.

50 16. Katzmarzyk P, Janssen I, Ardern C. Physical inactivity, excess adiposity and premature mortality. Obesity reviews. 2003;4(4):257-90. 17. Ärnlöv J, Ingelsson E, Sundström J, Lind L. Impact of body mass index and the metabolic syndrome on the risk of cardiovascular disease and death in middle-aged men. Circulation. 2010;121(2):230-6. 18. Kirk EP, Klein S. Pathogenesis and pathophysiology of the cardiometabolic syndrome. The Journal of Clinical Hypertension. 2009;11(12):761-5. 19. Oguma Y, Sesso HD, Paffenbarger RS, Lee IM. Weight change and risk of developing type 2 diabetes. Obesity research. 2005;13(5):945-51. 20. Malhotra A, White DP. Obstructive sleep apnoea. The lancet. 2002;360(9328):237-45. 21. Mason EE, Ito C. Gastric bypass in obesity. Surg Clin North Am. 1967;47(6):1345-51. 22. Mason EE, Ito C. Gastric bypass. Annals of surgery. 1969;170(3):329. 23. Torres JC, Oca CF, Garrison RN. Gastric bypass: Roux-en-Y gastrojejunostomy from the lesser curvature. South Med J. 1983;76(10):1217-21. 24. McCarthy H, Rucker JR, Chan E, Rupp W, Snover D, Goodale R, et al. Gas- tritis after gastric bypass surgery. Surgery. 1985;98(1):68-71. 25. Ballesta C, Berindoague R, Cabrera M, Palau M, Gonzales M. Management of anastomotic leaks after laparoscopic Roux-en-Y gastric bypass. Obesity surgery. 2008;18(6):623-30. 26. Jacobsen H, Nergard B, Leifsson B, Frederiksen S, Agajahni E, Ekelund M, et al. Management of suspected anastomotic leak after bariatric laparoscopic Roux‐en‐y gastric bypass. British Journal of Surgery. 2014;101(4):417-23. 27. Stenberg E, Szabo E, Ågren G, Näslund E, Boman L, Bylund A, et al. Early complications after laparoscopic gastric bypass surgery: results from the Scandinavian Obesity Surgery Registry. Annals of surgery. 2014;260(6):1040-7. 28. Podnos YD, Jimenez JC, Wilson SE, Stevens CM, Nguyen NT. Complica- tions after laparoscopic gastric bypass: a review of 3464 cases. Arch Surg. 2003;138(9):957-61. 29. Freedman J, Jonas E, Näslund E, Nilsson H, Marsk R, Stockeld D. Treatment of leaking gastrojejunostomy after gastric bypass surgery with special empha- sis on stenting. Surgery for Obesity and Related Diseases. 2013;9(4):554-8. 30. Gonzalez R, Sarr MG, Smith CD, Baghai M, Kendrick M, Szomstein S, et al. Diagnosis and contemporary management of anastomotic leaks after gastric bypass for obesity. Journal of the American College of Surgeons. 2007;204(1):47-55. 31. Marshall JS, Srivastava A, Gupta SK, Rossi TR, DeBord JR. Roux-en-Y gastric bypass leak complications. Arch Surg. 2003;138(5):520-3; discussion 3-4. 32. Angrisani L, Santonicola A, Iovino P, Vitiello A, Higa K, Himpens J, et al. IFSO Worldwide Survey 2016: Primary, Endoluminal, and Revisional Proce- dures. Obes Surg. 2018;28(12):3783-94. 33. Jiang HP, Lin LL, Jiang X, Qiao HQ. Meta-analysis of hand-sewn versus mechanical gastrojejunal anastomosis during laparoscopic Roux-en-Y gastric bypass for morbid obesity. Int J Surg. 2016;32:150-7. 34. Gys B, Gys T, Lafullarde T. The Use of Unidirectional Knotless Barbed Su- ture for Enterotomy Closure in Roux-en-Y Gastric Bypass: a Randomized Comparative Study. Obes Surg. 2017;27(8):2159-63.

51 35. Bures C, Seika P, Denecke C, Pratschke J, Zorron R. Routine Use of V- Lock(R) Suture for Bariatric Anastomosis Is Safe: Comparative Results from Consecutive Case Series. Arq Bras Cir Dig. 2019;32(3):e1452. 36. Almby K, Edholm D. Anastomotic Strictures After Roux-en-Y Gastric By- pass: a Cohort Study from the Scandinavian Obesity Surgery Registry. Obes Surg. 2019;29(1):172-7. 37. Bell RL, Reinhardt KE, Flowers JL. Surgeon-performed endoscopic dilatation of symptomatic gastrojejunal anastomotic strictures following laparoscopic Roux-en-Y gastric bypass. Obes Surg. 2003;13(5):728-33. 38. Rossi TR, Dynda DI, Estes NC, Marshall JS. Stricture dilation after laparoscopic Roux-en-Y gastric bypass. Am J Surg. 2005;189(3):357-60. 39. Dillemans B, Sakran N, Van Cauwenberge S, Sablon T, Defoort B, Van Dessel E, et al. Standardization of the fully stapled laparoscopic Roux-en-Y gastric bypass for obesity reduces early immediate postoperative morbidity and mortality: a single center study on 2606 patients. Obes Surg. 2009;19(10):1355-64. 40. Al-Mansour MR, Mundy R, Canoy JM, Dulaimy K, Kuhn JN, Romanelli J. Internal Hernia After Laparoscopic Antecolic Roux-en-Y Gastric Bypass. Obes Surg. 2015;25(11):2106-11. 41. Clavien PA, Sanabria JR, Strasberg SM. Proposed classification of complications of surgery with examples of utility in cholecystectomy. Surgery. 1992;111(5):518-26. 42. Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg. 2004;240(2):205-13. 43. Tian H, Tian J, Yang K, Yi K, Li L. The effects of laparoscopic vs. open gastric bypass for morbid obesity: a systematic review and meta‐analysis of randomized controlled trials. obesity reviews. 2011;12(4):254-60. 44. Bueter M, Ashrafian H, le Roux CW. Mechanisms of weight loss after gastric bypass and gastric banding. Obesity facts. 2009;2(5):325-31. 45. Peterli R, Borbély Y, Kern B, Gass M, Peters T, Thurnheer M, et al. Early results of the Swiss Multicentre Bypass or Sleeve Study (SM-BOSS): a prospective randomized trial comparing laparoscopic sleeve gastrectomy and Roux-en-Y gastric bypass. Annals of surgery. 2013;258(5):690. 46. Nguyen NT, Slone JA, Nguyen X-MT, Hartman JS, Hoyt DB. A prospective randomized trial of laparoscopic gastric bypass versus laparoscopic adjustable gastric banding for the treatment of morbid obesity: outcomes, quality of life, and costs. Annals of surgery. 2009;250(4):631-41. 47. Angrisani L, Lorenzo M, Borrelli V. Laparoscopic adjustable gastric banding versus Roux-en-Y gastric bypass: 5-year results of a prospective randomized trial. Surgery for obesity and related diseases. 2007;3(2):127-32. 48. Angrisani L, Cutolo PP, Formisano G, Nosso G, Vitolo G. Laparoscopic adjustable gastric banding versus Roux-en-Y gastric bypass: 10-year results of a prospective, randomized trial. Surgery for obesity and related diseases. 2013;9(3):405-13. 49. Mazer LM, Azagury DE, Morton JM. Quality of Life After Bariatric Surgery. Curr Obes Rep. 2017;6(2):204-10. 50. Karlsen TI, Tveita EK, Natvig GK, Tonstad S, Hjelmesaeth J. Validity of the SF-36 in patients with morbid obesity. Obes Facts. 2011;4(5):346-51.

52 51. Corica F, Corsonello A, Apolone G, Lucchetti M, Melchionda N, Marchesini G, et al. Construct validity of the Short Form-36 Health Survey and its relationship with BMI in obese outpatients. Obesity (Silver Spring). 2006;14(8):1429-37. 52. Orwelius L, Nilsson M, Nilsson E, Wenemark M, Walfridsson U, Lundstrom M, et al. The Swedish RAND-36 Health Survey - reliability and responsive- ness assessed in patient populations using Svensson's method for paired ordi- nal data. J Patient Rep Outcomes. 2017;2(1):4. 53. Karlsson J, Taft C, Ryden A, Sjostrom L, Sullivan M. Ten-year trends in health-related quality of life after surgical and conventional treatment for severe obesity: the SOS intervention study. Int J Obes (Lond). 2007;31(8):1248- 61. 54. Dumitra TC, Trepanier M, Fiore JF, Jr., Kaneva P, Carli F, Fried GM, et al. The relationship of two postoperative complication grading schemas with postoperative quality of life after elective colorectal surgery. Surgery. 2019;166(4):663-9. 55. Clavien PA, Barkun J, de Oliveira ML, Vauthey JN, Dindo D, Schulick RD, et al. The Clavien-Dindo classification of surgical complications: five-year experience. Annals of surgery. 2009;250(2):187-96. 56. Nielsen S, Degenhardt L, Hoban B, Gisev N. A synthesis of oral morphine equivalents (OME) for opioid utilisation studies. Pharmacoepidemiol Drug Saf. 2016;25(6):733-7. 57. Costantino F, Dente M, Perrin P, Sarhan FA, Keller P. Barbed unidirectional V-Loc 180 suture in laparoscopic Roux-en-Y gastric bypass: a study comparing unidirectional barbed monofilament and multifilament absorbable suture. Surgical endoscopy. 2013;27(10):3846-51. 58. De Blasi V, Facy O, Goergen M, Poulain V, De Magistris L, Azagra JS. Barbed versus usual suture for closure of the gastrojejunal anastomosis in laparoscopic gastric bypass: a comparative trial. Obesity surgery. 2013;23(1):60- 3. 59. Facy O, De Blasi V, Goergen M, Arru L, De Magistris L, Azagra J-S. Laparo- scopic gastrointestinal anastomoses using knotless barbed sutures are safe and reproducible: a single-center experience with 201 patients. Surgical endoscopy. 2013;27(10):3841-5. 60. Aghajani E, Jacobsen HJ, Nergaard BJ, Hedenbro JL, Leifson BG, Gislason H. Internal hernia after gastric bypass: a new and simplified technique for laparoscopic primary closure of the mesenteric defects. Journal of gastrointestinal surgery. 2012;16(3):641-5. 61. Tyner RP, Clifton GT, Fenton SJ. Hand-sewn gastrojejunostomy using knotless unidirectional barbed absorbable suture during laparoscopic gastric bypass. Surgical endoscopy. 2013;27(4):1360-6. 62. Csendes A, Burgos AM, Altuve J, Bonacic S. Incidence of marginal ulcer 1 month and 1 to 2 years after gastric bypass: a prospective consecutive endoscopic evaluation of 442 patients with morbid obesity. Obesity surgery. 2009;19(2):135-8. 63. Edholm D, Sundbom M. Comparison between circular-and linear-stapled gastrojejunostomy in laparoscopic Roux-en-Y gastric bypass—a cohort from the Scandinavian Obesity Registry. Surgery for Obesity and Related Diseases. 2015;11(6):1233-6. 64. Oor J, Castro S, Wagensveld B. V‐loc™ capable of grasping surrounding tissue causes obstruction at the jejunojejunostomy after Roux‐en‐Y laparoscopic gastric bypass. Asian journal of endoscopic surgery. 2015;8(2):209-11.

53 65. Buchs NC, Ostermann S, Hauser J, Roche B, Iselin CE, MoreL P. Intestinal obstruction following use of laparoscopic barbed suture: A new complication with new material? Minimally Invasive Therapy & Allied Technologies. 2012;21(5):369-71. 66. Lee S, Carmody B, Wolfe L, DeMaria E, Kellum JM, Sugerman H, et al. Effect of location and speed of diagnosis on anastomotic leak outcomes in 3828 gastric bypass cases. Journal of Gastrointestinal Surgery. 2007;11(6):708-13. 67. Livingston EH, Huerta S, Arthur D, Lee S, De Shields S, Heber D. Male gender is a predictor of morbidity and age a predictor of mortality for patients undergoing gastric bypass surgery. Annals of surgery. 2002;236(5):576. 68. Fernandez Jr A, DeMaria E, Tichansky D, Kellum J, Wolfe L, Meador J, et al. Experience with over 3,000 open and laparoscopic bariatric procedures: multivariate analysis of factors related to leak and resultant mortality. Surgical Endoscopy And Other Interventional Techniques. 2004;18(2):193-7. 69. Reames BN, Bacal D, Krell RW, Birkmeyer JD, Birkmeyer NJ, Finks JF. Influence of median surgeon operative duration on adverse outcomes in bariatric surgery. Surg Obes Relat Dis. 2015;11(1):207-13. 70. Salinas A, Baptista A, Santiago E, Antor M, Salinas H. Self-expandable metal stents to treat gastric leaks. Surgery for Obesity and Related Diseases. 2006;2(5):570-2. 71. Kim J, Azagury D, Eisenberg D, DeMaria E, Campos GM. ASMBS position statement on prevention, detection, and treatment of gastrointestinal leak after gastric bypass and sleeve gastrectomy, including the roles of imaging, surgical exploration, and nonoperative management. Surg Obes Relat Dis. 2015;11(4):739-48. 72. Thodiyil PAM, Yenumula, Panduranga MD, Rogula, Tomasz MD et al. Se- lective Nonoperative Management of Leaks After Gastric Bypass: Lessons Learned From 2675 Consecutive Patients. Annals of surgery. 2008;248(5):782-92. 73. Jacobsen H, Nergard BJ, Leifsson BG, Frederiksen SG et al. Management of suspected anastomotic leak after bariatric laparoscopic Roux‐en‐y gastric bypass. British Journal of Surgery. 2014;101(4):417-23. 74. Nguyen NT, Paya M, Stevens CM, Mavandadi S, Zainabadi K, Wilson SE. The relationship between hospital volume and outcome in bariatric surgery at academic medical centers. Ann Surg. 2004;240(4):586-93; discussion 93-4. 75. Vidarsson B, Sundbom M, Edholm D. Incidence and treatment of leak at the gastrojejunostomy in Roux-en-Y gastric bypass: a cohort study of 40,844 patients. Surg Obes Relat Dis. 2019;15(7):1075-9. 76. Cartmill JA, Shakeshaft, A.J., Walsh, W.R. and Martin, C.J. High pressures are generated at the tip of laparoscopic graspers. Aus NZJSurg. 1999;69:127- 30. 77. van der Voort M HE, Gouma DJ. Bowel injury as a complication of laparoscopy. British Journal of Surgery. 2004;Oct; 91(10):1253-8. 78. Sundbom M, Ottosson J. Trocar Injuries in 17,446 Laparoscopic Gastric By- pass-a Nationwide Survey from the Scandinavian Obesity Surgery Registry. Obes Surg. 2016;26(9):2127-30. 79. Llarena NC SA, Milad MP. Bowel injury in gynecologic laparoscopy: a systematic review. Obstet Gynecol. 2015;Oct; 126(6):1407-17. 80. Haskins IN, Amdur R, Vaziri K. The effect of smoking on bariatric surgical outcomes. Surg Endosc. 2014;28(11):3074-80.

54 81. Scarborough JE, Mantyh CR, Sun Z, Migaly J. Combined Mechanical and Oral Antibiotic Bowel Preparation Reduces Incisional Surgical Site Infection and Anastomotic Leak Rates After Elective Colorectal Resection: An Analy- sis of Colectomy-Targeted ACS NSQIP. Ann Surg. 2015;262(2):331-7. 82. Jacobsen HJ, Nergard BJ, Leifsson BG, Frederiksen SG, Agajahni E, Ekelund M, et al. Management of suspected anastomotic leak after bariatric laparoscopic Roux-en-y gastric bypass. Br J Surg. 2014;101(4):417-23. 83. Gonzalez R, Sarr MG, Smith CD, Baghai M, Kendrick M, Szomstein S, et al. Diagnosis and contemporary management of anastomotic leaks after gastric bypass for obesity. J Am Coll Surg. 2007;204(1):47-55. 84. Marshall JS, Srivastava A, Gupta SK, Rossi TR, DeBord JR. Roux-en-Y gastric bypass leak complications. Archives of Surgery. 2003;138(5):520-4. 85. Schauer PR, Ikramuddin S, Gourash W, Ramanathan R, Luketich J. Outcomes after laparoscopic Roux-en-Y gastric bypass for morbid obesity. Ann Surg. 2000;232(4):515-29. 86. Ladak F, Dang JT, Switzer NJ, Mocanu V, Birch DW, Karmali S. Rates of reoperation and nonoperative intervention within 30 days of bariatric surgery. Surg Obes Relat Dis. 2019;15(3):431-40. 87. Hawthorne G O, RH, Taylor A, Sansoni J. The SF 36 Version 2: critical analyses of population weights, scoring algorithms and population norms. Quality of life research. 2007;16:661-73. 88. Pinto A, Faiz O, Davis R, Almoudaris A, Vincent C. Surgical complications and their impact on patients' psychosocial well-being: a systematic review and meta-analysis. BMJ Open. 2016;6(2):e007224. 89. Bouras G, Burns EM, Howell AM, Bagnall NM, Lee H, Athanasiou T, et al. Systematic review of the impact of surgical harm on quality of life after general and gastrointestinal surgery. Ann Surg. 2014;260(6):975-83. 90. Courcoulas AP, Christian NJ, Belle SH, Berk PD, Flum DR, Garcia L, et al. Weight change and health outcomes at 3 years after bariatric surgery among individuals with severe obesity. JAMA. 2013;310(22):2416-25. 91. Cunningham JL, Merrell CC, Sarr M, Somers KJ, McAlpine D, Reese M, et al. Investigation of antidepressant medication usage after bariatric surgery. Obes Surg. 2012;22(4):530-5. 92. Elias K, Hedberg J, Sundbom M. Prevalence and impact of acid-related symptoms and diarrhea in patients undergoing Roux-en-Y gastric bypass, sleeve gastrectomy, and biliopancreatic diversion with duodenal switch. Surg Obes Relat Dis. 2020;16(4):520-7. 93. Wennerlund J, Gunnarsson U, Strigard K, Sundbom M. Acid-related complications after laparoscopic Roux-en-Y gastric bypass: risk factors and impact of proton pump inhibitors. Surg Obes Relat Dis. 2020;16(5):620-5. 94. W.C. King JYC, S.H. Belle, et al. Use of prescribed opioids before and after bariatric surgery: prospective evidence from a U.S. multicenter cohort study. Surg Obes Relat Dis. 2017;13(8):1337-46. 95. S. Wallen ES, M. Palmetun-Ekback, I. Naslund. Use of opioid analgesics before and after gastric bypass surgery in Sweden: a population-based study. Obes Surg. 2018;28(11):3518-23. 96. Heinberg LJ, Pudalov L, Alameddin H, Steffen K. Opioids and bariatric surgery: A review and suggested recommendations for assessment and risk reduction. Surg Obes Relat Dis. 2019;15(2):314-21. 97. C. Lloret-Linares DH, C. Bardin, et al. Effect of a Roux-en-Y gastric bypass on the pharmacokinetics of oral morphine using a population approach. Clin Pharmacokinet. 2014;53(10):919-30.

55 98. W.C. King JYC, J.E. Mitchell, et al. Prevalence of alcohol use disorders before and after bariatric surgery. JAMA. 2012;307(23):2516-25. 99. Weiner RA, El-Sayes IA, Theodoridou S, Weiner SR, Scheffel O. Early postoperative complications: incidence, management, and impact on length of hospital stay. A retrospective comparison between laparoscopic gastric bypass and sleeve gastrectomy. Obes Surg. 2013;23(12):2004-12. 100. Olbers T, Lonroth H, Fagevik-Olsen M, Lundell L. Laparoscopic gastric bypass: development of technique, respiratory function, and long-term outcome. Obes Surg. 2003;13(3):364-70. 101. Svedlund J, Sjodin I, Dotevall G. GSRS--a clinical rating scale for gastrointestinal symptoms in patients with irritable bowel syndrome and peptic ulcer disease. Dig Dis Sci. 1988;33(2):129-34. 102. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand. 1983;67(6):361-70. 103. Wynne EK AD. Novel device to detect enterotomies in real time during laparoscopy: first in human trial during Roux-en-y gastric bypass. Surg Endosc. 2019;33(5):1687-92. 104. Hedberg S, Olbers T, Peltonen M, Osterberg J, Wiren M, Ottosson J, et al. BEST: Bypass equipoise sleeve trial; rationale and design of a randomized, registry-based, multicenter trial comparing Roux-en-Y gastric bypass with sleeve gastrectomy. Contemp Clin Trials. 2019;84:105809.

Acta Universitatis Upsaliensis Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine 1676 Editor: The Dean of the Faculty of Medicine

A doctoral dissertation from the Faculty of Medicine, Uppsala University, is usually a summary of a number of papers. A few copies of the complete dissertation are kept at major Swedish research libraries, while the summary alone is distributed internationally through the series Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine. (Prior to January, 2005, the series was published under the title “Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine”.)

ACTA UNIVERSITATIS UPSALIENSIS Distribution: publications.uu.se UPPSALA urn:nbn:se:uu:diva-417549 2020