ĐẶT VẤN ĐỀ Xuất huyết tiêu hoá là một cấp cứu thường gặp trên lâm sàng, không chỉ ở Việt Nam mà còn phổ biến ở các nước đã phát triển. Tại Mỹ và Anh, tỷ lệ nhập viện hàng năm do chảy máu đường tiêu hoá ước tính khoảng 150/100.000 dân và tỷ lệ tử vong khoảng 4 – 10%[1] . Biểu hiện xuất huyết tiêu hoá rất đa dạng, bao gồm những trường hợp chảy máu tiềm ẩn ở mức vi thể hoặc chạy máu đại thể mức độ nhẹ diễn biến từ từ hoặc từng đợt cho đến những trường hợp xuất huyết ồ ạt, liên tục gây thiếu máu nặng và đe doạ tính mạng người bệnh. Chảy máu ruột non là nhóm bệnh lý ít gặp, chiếm tỷ lệ khoảng 5% các trường hợp xuất huyết tiêu hoá nói chung[2]. Trước năm 2000, nhóm bệnh lý này vẫn là một thách thức chẩn đoán trên lâm sàng do ruột non được coi là “vùng tối” của các bác sỹ nội soi tiêu hoá vì hầu như không thể tiếp cận trực tiếp được toàn bộ bề mặt niêm mạc ruột ngoại trừ phẫu thuật mở. Giai đoạn 10 năm sau năm 2000 là thời điểm bùng nổ của các phương pháp nội soi ruột non mới có thể giúp bác sỹ tiêu hoá quan sát và tiếp cận được gần như toàn bộ niêm mạc ruột non. Nội soi ruột non được chia ra làm 2 nhóm chính, đó là nội soi viên nang có thể chụp ảnh, phát hiện tổn thương ở niêm mạc ruột và nội soi có dụng cụ hỗ trợ (như nội soi ruột non bóng kép, bóng đơn…) ngoài khả năng chẩn đoán nguyên nhân bệnh còn có thể tiến hành các can thiệp thủ thuật như sinh thiết chẩn đoán mô học hoặc điều trị cầm máu... Nhiều nghiên cứu trên thế giới đều cho thấy hiệu quả của các phương pháp nội soi này trong chẩn đoán và xử trí xuất huyết tiêu hoá tại ruột non. Nội soi viên nang có khả năng chẩn đoán nguyên nhân chảy máu từ 35% đến 83%, tuỳ từng nghiên cứu [3], [4]. Khả năng chẩn đoán của nội soi ruột non bóng kép trong chảy máu ruột non là từ 60% đến 86% và khả năng can thiệp cầm máu là từ 40% đến 73% [5-8]. Đối với nội soi ruột non bóng đơn, khả năng chẩn đoán tổn thương từ 41% đến 88% và khả năng can thiệp qua nội soi từ 7% đến 50% [9-11]. Khả năng chẩn đoán và can thiệp khác nhau tuỳ thuộc mức độ chảy máu là đại thể hay vi thể, thời điểm bệnh nhân được nội soi là cấp cứu hay không cấp cứu… Mỗi một phương pháp nội soi nói trên đều có ưu nhược điểm riêng, được áp dụng tuỳ thuộc tính chất xuất huyết của người bệnh, trang bị của các cơ sở y tế cũng như hướng dẫn điều trị của mỗi quốc gia và khu vực. Tuy nhiên, nội soi ruột non bóng kép có ưu thế hơn so với các phương pháp nội soi khác ở chỗ tỷ lệ soi hết được toàn bộ ruột non cao hơn [12]. Nghiên cứu của May và cộng sự thấy tỷ lệ soi hết ruột non của nội soi bóng kép là 66% còn nội soi bóng đơn là 22% [13]. Tương tự, Messer và cộng sự thấy tỷ lệ soi hết ruột non của nội soi xoắn ốc là 8%, trong khi tỷ lệ tương tự của nội soi bóng kép là 92% [14]. Tại Bệnh viện Bạch Mai, nội soi ruột non bóng kép được bắt đầu áp dụng trong thực hành lâm sàng từ năm 2014. Việc ứng dụng kĩ thuật nội soi mới này đã giúp bác sỹ tiêu hoá có thêm một phương pháp để chẩn đoán nguyên nhân gây chảy máu ở các bệnh nhân xuất huyết tiêu hoá tại ruột non và giúp can thiệp điều trị cầm máu cho các tổn thương phù hợp. Chính vì vậy, chúng tôi đã tiến hành đề tài “Nghiên cứu ứng dụng kĩ thuật nội soi ruột non bóng kép trong chẩn đoán và điều trị xuất huyết tiêu hoá tại ruột non” nhằm hai mục tiêu sau: 1. Mô tả đặc điểm lâm sàng, cận lâm sàng, hình ảnh nội soi ruột non bóng kép và nguyên nhân ở bệnh nhân xuất huyết tiêu hoá đại thể tại ruột non 2. Đánh giá kết quả của một số phương pháp điều trị xuất huyết tiêu hoá đại thể tại ruột non
MINISTRY OF EDUCATION TRAINING MINISTRY OF HEALTH HANOI MEDICAL UNIVERSITY ====== NGUYEN HOAI NAM RESEARCH ON THE APPLICATION OF DOUBLE BALLOON ENDOSCOPY IN DIAGNOSIS AND TREATMENT OF SMALL INTESTINAL BLEEDING Specialism: Internal Gastroenterology Code : 9720107 ABSTRACT OF THESIS HA NOI - 2022 THE LIST OF WORKS HAS PUBLISHED AND RELATED TO THE THESIS Nguyen Hoai Nam, Vu Truong Khanh, Dao Van Long (2020) Double Balloon Endoscopy in the diagnosis of causes of small intestinal bleeding Journal of Medical Research, 132 (8) 206-214 (Vietnamese) Nguyen Hoai Nam, Dao Van Long, Nguyen Cong Long, Vu Truong Khanh (2021) Small intestinal bleeding due to diverticulum detected by double balloon endoscopy Vietnam Journal of Gastroenterology, IX (62) 3849-3857 (Vietnamese) Nguyen Hoai Nam, Dao Van Long, Nguyen Cong Long, Vu Truong Khanh (2021) Clinical and subclinical characteristics of patients with overt small intestinal bleeding Vietnam Medical Journal, 507(1) October 94-99 (Vietnamese) Nguyen Hoai Nam, Dao Van Long (2022) Treatment of small intestinal bleeding by double balloon endoscopy” Vietnam Medical Journal, 510(1) January 207-212 (Vietnamese) INTRODUCTION Small intestinal bleeding is a rare disease, accounting for 5% of total gastrointestinal bleeding Its clinical manifestations are diverse, including occult bleeding or overt bleeding that may be mild, intermittent or massive and life-threatening haemorrhage This disease was previously challenging to diagnosis because endoscopy could not access the entire length of the small intestine Since 2000, several enteroscopies have been introduced that have helped to explore almost the entire mucosal surface of the small intestine Among them, Double balloon endoscopy (DBE) has the diagnostic yields from 60 to 86% and hemostasis therapeutic yields from 40 to 73% Furthermore, in complete enteroscopy rate, DBE was superior to single balloon endoscopy (66% vs 22%) and to spiral endoscopy (92% vs 8%) At Bach Mai hospital, DBE has been applied since 2014, so that gastroenterologists have had an additional method to diagnose causes of small intestinal bleeding and to perform hemostasis procedures for some suitable lesions Therefore, we conducted this thesis with the following objects: Describe the clinical and subclinical characteristics, DBE images and etiologies in patients with overt small intestinal bleeding Evaluate the results of treatment methods for overt small intestinal bleeding NEW SCIENTIFIC CONTRIBUTIONS OF THE THESIS This is the first study in Vietnam using DBE to diagnose for patients with overt small intestinal bleeding (OSIB) With strict patient selection criteria, the study has accurately described the clinical, subclinical characteristics of patients and causes of OSIB This is also the first study in Vietnam to evaluate the results of endoscopic hemostasis, surgery and medical treatment for patient with OSIB The results of the thesis have contributed to the literature on clinical, endoscopic and pathological features that help diagnose and treat OSIB STRUCTURE OF THE THESIS The thesis has 121 pages with main chapters, including Introduction (2 pages), Literature review (33 pages), Subject and Methodology (20 pages), Results (30 pages), Discussion (33 pages), Conclusion (2 pages) and Recommendation (1 pages) The thesis has 40 tables, charts, 21 figures, 176 references (including 168 documents in English and documents in Vietnamese; 15 documents published after 2020, 64 documents from 2015 – 2019 and 48 documents from 2010 – 2014) and appendix of lesions detected by DBE CHAPTER LITERATURE REVIEW According to American Gastroenterological Association (2015), if the source of bleeding is between the ampulla and the terminal ileum, it is designated as small intestinal bleeding Regarding the degree of bleeding, the literature divides it into types: overt bleeding when there is melena or bloody stool clinically and occult bleeding which manifests as iron deficiency anemia and positive fecal blood test To diagnose small intestinal bleeding, there are currently two approaches using imaging technique and/ or enteroscopy For small bowel imaging, contrast-enhanced computed tomography is the most commonly used However, Wang et al found its ability to diagnose the cause of bleeding was lower than that of DBE (38% and 78%) Enteroscopy is divided into groups, the first is capsule endoscopy in which only the mucosal surface can be visualized, and the second is device-assisted enteroscopy, beside of observing the mucosa it can allow us to perform endoscopic procedures Among the device-assisted enteroscopy, DBE has the highest rate of total enteroscopy and is the most commonly used The diagnostic yield of DBE ranges from 60 to 86% in patients with suspected small bowel bleeding Causes are divided into main groups: vascular abnormalities, neoplastic lesions (including tumors and polyps), diverticula, ulcer and some other rare causes Studies in different races and geographical regions have shown a different distribution of hemorrhage causes Xin et al found that in eastern countries, small bowel ulcers were the most common (37.6%), followed by vascular abnormalities (26.7%) and tumor lesions (26.4%) In contrast, in western population, vascular lesions account for a very high rate (65,9%) and inflammatory lesions and tumors account for a small proportion (15.7% and 14.4%) With the advantage of having an instrument channel, DBE allows hemostasis interventions for 16 to 57% of patients with OSIB According to the Japan Endoscopic Society, endoscopic hemostasis procedures is indicated for bleeding lesions and temporary stop-bleeding lesions, including vascular abnormalities < 1cm, diverticula or ulcers with non-bleeding visible vessel or having adherent clot and bleeding polyps with suitable size for endoscopic resection The success rate of endoscopic hemostasis has been reported range from 43 to 84%, thereby reducing the number of patients requiring surgical treatment, the more intensive method with more risks and complications than endoscopy Currently, surgery is indicated for cases of severe acute OSIB that is life-threatening, insufficient time to prepare for DBE or bleeding causes cannot be completely treated by enteroscopy such as tumor, Meckel’s diverticulum or when endoscopic hemostasis is unsuccessful In some patients whose general condition is not suitable for surgery, embolization can be performed, however the clinical success rate is 71,4% and complication of intestinal infarction is 4.3% For certainty stopped bleeding lesions such as ulcers, what the patients are treated depends on the cause, for example treatment of intestinal tuberculosis or Crohn’s disease, in combination with sucrafate and rebamipide In Vietnam, research on the clinical characteristics of patients with OSIB is very limited with a few case reports and two studies by Nguyen Cong Long et al in 2009 and 2020 with 17 and 54 patients undergoing surgical treatment About DBE, Pham Huu Tung et al initially described the application of DBE to patients with OSIB at Cho Ray hospital, Ho Chi Minh City At Bach Mai hospital, in 2015, our reasearch team published the first paper about the application of DBE in 38 suspected small bowel disease patients with the small lesion detection rate of 68,4% Using DBE to diagnose and treat patients with OSIB, our study is the first research conducted in Vietnam with a large enough number of patients CHAPTER SUBJECTS AND METHODOLOGY 2.1 SUBJECT Patients were diagnosed with suspected overt small intestinal bleeding according to the guideline of the American Gastroenterological Association in 2015 with the following characteristics: (1) Clinical presence of melena and/or bloody stools (2) No cause of bleeding was found on upper gastrointestinal endoscopy and total colonoscopy, or there was evidence of bleeding in the small intestine (capsule endoscopy found the bleeding site in small intestine or gastroscopy showed blood reflux from the jejunum) The patients were performed DBE to find out lesions If they meet the diagnostic criteria for small intestinal bleeding below, the patients will be included in the study Criteria for diagnosis of small intestinal bleeding: when DBE detects a small bowel lesion located below the papilla to the end of the ileum, meeting of following criteria: (1) Lesions are bleeding during DBE, or the lesions have stopped bleeding during DBE but having extravasation of contrast into the bowel lumen on computed tomography scan or bleeding on surgery (2) Lesions with signs of recent bleeding such as non-bleeding visible vessel or adherent blot or having blood in the small intestine nearby (3) Lesions that have stopped bleeding This kind of lesions has to meet the diagnostic criteria for identified causes of small intestinal bleeding defined by Shinozaki, Tanaka et al Table 2-1: Diagnostic criteria for lesions that stopped bleeding as the causes of small intestinal bleeding according to Shinozaki, Tanaka et al Lesion Identified causes of bleeding Vascular lesion type 1B, 2, and according to the classification of Yano Tumors/ polyps with ulcer, vascularization or > cm in diameter Diverticulum with ulcer Ulcer > cm in diameter Exclusion criteria: - DBE does not detect lesions, or detects the cause of bleeding in the upper or lower gastrointestinal tract - DBE detects lesions that belong to uncertain causes of bleeding according to the classification of Shinozaki, Tanaka et al - Pregnant patients or patients did not agree to participate in the study 2.2 LOCATION AND TIME OF STUDY: at the Gastroenterology Department of Bach Mai hospital, from May 2015 to May 2020 2.3 METHODOLOGY 2.3.1 Study design: a descriptive cross-sectional study with longitudinal followup to evaluate treatment outcomes in an uncontrolled clinical trial model 2.3.2 Sample size: because small intestinal bleeding is a rare disease, we conducted a convenient sampling method 2.3.3 Research equipment: - DBE system of Fujinon – Japan, including central processor VP-3500H, light source XL-4450, balloon pump and pressure control system and endoscopes EN-450T5 and EN-580T with suitable overtube - ERBE 200 electrosurgical unit and C-ARM x-ray machine - Interventional instruments: biopsy, hemostatic needle, clip, APC catheter, coagrasper, snare 2.4 RESEARCH PROCESS 2.4.1 Selection, assessment and management of patients prior to DBE - Patients diagnosed with suspected small intestinal bleeding were taken the history, clinical examination and performed tests - The clinical and laboratory severity of acute anemia is applied according to the following table: Characteristics Clinical Pulse (beats/min) Blood pressure Sensorium Laboratory - Hemoglobin (g/l) Mild Moderate Severe < 100 100 - 120 > 120 Normal Borderline Hypotensive Conscious Anxious Drowsy > 90 70 - 90 ≤ 70 The patients underwent an abdominal contrast-enhanced computed tomography If gastroscopy or colonoscopy showed red blood pouring out from the small intestine, the patients would be performed DBE immediately for hemostasis in time, no need to take computed tomography to ensure the benefit of the patients - The patients were resuscitated and transfused red blood cells to maintain hemoglobin level at 70 – 80 g/l If the patient was more than 60 years old or had concomitant cardiovascular disease, it was necessary to maintain the hemoglobin level about 90 – 100 g/l Patients with coagulopathy would receive fresh plasma, platelets transfusion 2.4.2 Double balloon endoscopy - Choice the insertion route: usually, transoral approach would be done first, unless the patient had fresh bloody stool and computed tomography suggested lesion in the terminal ileum, transanal approach would be performed first Transoral DBE required fasting at least hours only, whereas transanal DBE needed bowel preparation with Fortran - The patients were anesthetized with propofol alone or in combination with midazolam and fentanyl, with or without intubation They were monitored by an anesthesiologist before, during DBE and until recovery - If the first approach DBE detected the cause of bleeding, the second approach could not be necessary If no cause was found, or if there was a possible cause of multiple lesions, the last bowel segment would be marked with a clip When performing remaining approach DBE (if necessary), seeing the marked clip meant that the entire small intestine has been observed 2.4.3 Assessment of lesions detected by DBE - Lesions were bleeding or with signs of recent bleeding (non-bleeding visible vessel, adherent blot or having blood in the small intestine nearby) or stoppedbleeding lesions that met the diagnostic criteria of Shinozaki et al would be considered as the cause of small intestinal bleeding - Tumors or polyps were diagnosed histologically by specimens obtained after surgery, endoscopic resection or biopsy Specimens were evaluated and classified according to the WHO classification in 2019 at the Pathology Center, Bach Mai hospital - Vascular lesions were classified according to Yano et al Table 2-2:Yano endoscopic classification of vascular Type 1A Punctuate erythema (≤ mm) with or without oozing Angioectasia lesions Type 1B Patchy erythema (a few mm) with or without oozing Type 2A Punctuate lesions (≤ mm) with pulsatile bleeding Dieulafoy lesions Type 2B Pulsatile red protrusion without surrounding venous dilatation Type Pulsatile red protrusion with surrounding venous dilation AVM Type Other lesions not classified into above categories Other - Small bowel diverticulum was classified as Meckel’s and non-Meckel’s diverticulum + Meckel’s diverticulum located in the ileum section 80 – 100 cm from the Bauhin valve, at the anti-mesenteric border of the intestine seen at surgery, could be connected to the umbilicus by fibrous ligament and had ectopic tissues on histopathology + Non-Meckel’s diverticulum usually located in the duodenum or jejunum or in many different locations - Small bowel ulcers were classified according to the cause of the diseases - Lesion’s location: diviced into groups based on endoscopic images and routes + Proximal part: mucosa was thicker with larger folds, larger submucosal vessels and lesions were detected by peroral route Part III– IV of duodenum was determined when lesion located between papilla and Treitz angle + Distal part: mucosa was thinner with low and spare transverse mucosal folds, small and thin submucosal vessels and lesions was detected by peranal route 2.4.4 Endoscopic hemostasis - Indications for endoscopic hemostasis were applied according the Japanese Gastroenterology Association, including: + Bleeding lesions + Vascular abnormalities type 1, according to Yano classification and lesions type 3, if there diameter < cm + Ulcers or diverticula with non-bleeding visible vessel or adherent blot + Polyps that caused bleeding could be removed endoscopically - Endoscopic hemostasis was divided into types: + Endoscopic procedure was enough to stop bleeding; patients were monitored in the endoscopic therapy group + Endoscopic procedure achieved temporary hemostasis and patient needed to be transferred to surgery, applied to radically treated lesions such as tumors, Meckel’s diverticulum Patients were monitored in the surgery group - Indication of hemostasis techniques: + Adrenalin injection: indicated for bleeding lesions to reduce the flow rate to help determine the exact bleeding point for clipping or electrocoagulation afterwards Adrenalin injection was not used alone + Electrocoagulation: indicated for angiodysplasia (Yano classification type 1) + Clipping: indicated for invisible vessel, Dieulafoy lesions (Yano classification type 2) or other lesions ≤ cm + Polypectomy for pedunculated polyps, with looping before cutting - Clinically successful endoscopic intervention was defined as absence of rebleeding within 12 hours after hemostasis procedure Endoscopic intervention failure was defined as continued bleeding after intervention or having evidence of rebleeding within 12 hours of the procedure Diagnose rebleeding when: + Re-appearance of black or bloody stools after the patient had normal stools + Recurrence of hemodynamic instability after successful resuscitation in patients with black stools + Decrease in hemoglobin level > 20 g/l in 24 hours and/ or need red blood cells transfusion - If endoscopic hemostasis was unsuccessful, patients had persistently active bleeding in clinical; they would be treated with surgery or vascular intervention 2.4.5 Surgery - Indications for surgery: + Causes of bleeding couldn’t be completely treated by DBE, including tumors, larger polyps, vascular abnormalities > cm, ulcer suspected malignancy or causing bowel obstruction and perforation, and Meckel’s diverticulum + Bleeding lesions that could not be treated by endoscopic hemostasis or that the endoscopic intervention is unsuccessful - Surgery was not applied to end-stage cancer or diffuse small intestine lymphoma that had stopped bleeding by itself 2.4.6 Medical treatment, not endoscopic or surgical intervention - Including patients with: + Stopped-bleeding lesions such as ulcers and non-Meckel’s diverticulum, or diffuse intestinal lymphoma and end-stage cancer + Lesions having indications for operation but patients and their families did not agree to surgery - Medical treatment included specific treatment depending the bleeding causes or adjuvant therapy with sucrafate and rebamipide for ulcers 2.4.7 Follow-up patients after discharging from hospital - Patients were monitored periodically or immediately when there were signs of re-bleeding or other abnormalities by follow-up examination or telephone interview - Evaluation criteria: the rate of re-bleeding in all 3-treatment groups - The follow-up period was calculated from the first time when the patients had surgery, endoscopic hemostasis or endoscopic date (if patients were on medical treatment) and the last time when the patient had re-bleeding, died or the end of study, whichever occurred first 2.5 ANALYSIS METHOD OF RESEARCH DATA Data were processed by SPSS21 software, in which χ2 test, T-test, One-way INOVA test and Kaplan-Meier survival analysis and Log rank test were used The difference is statistically significant if p 90 g/l 68/84 (81,0%) patients required red blood cells transfusion with an average amount of 6.8 units (250 ml), medial of 5.5 units, IQR rang of [2.8 - 8.4], the minimum was unit and maximum was 32 units 3.1.3 Small bowel abnormalities detected on computed tomography (CT) Table 3-3: small bowel abnormalities detected on CT scan Small bowel Abnormalities n % Tumor 15 18.8 contrast extravasation into bowel lumen 7.5 Vascular lesions in small bowel wall 5.0 Small bowel diverticulitis 3.8 Thickening of small bowel wall 2.5 No abnormalities 50 62.5 Total 80 100.0 Comment: 37.5% of patients detected small bowel abnormalities on CT scan 10 3.2 ENDOSCOPIC CHARACTERISTIC OF BLEEDING CAUSES 3.2.1 Causes of small intestinal bleeding Table 3-4: causes of bleeding detected by DBE and their location Causes of small bowel Location n % Proximal part Distal part bleeding (P.III–IV duodenum) Neoplastic lesions Tumors Polyps Diverticulum Meckel (*) non-Meckel Vascular lesions Ulcer Total 30 35,7 24 (5)* 29,8 (2) 19 23,8 10,7 100 16 (1) 52 (61.9%) 32 (38.1%) 28 25 19 20 84 (*): number of lesions located in the third & fourth part of duodenum Comment: common causes of small intestinal bleeding were neoplastic lesions, diverticula and vascular abnormalities Table 3-5: histological types and endoscopic morphology of neoplastic lesions Endoscopic morphology n % Submucosal tumor Polypoid Ulcerative, Penduntumor infiltrated culated tumor polyp 16 57.2 15 Vascular tumor 21.4 Malignant lymphoma 7.1 1 Metastatic tumor 7.1 1 Adenocarcinoma 3.6 Carcinoid tumor 3.6 28 100 23 Bruner’s gland polyp 50.0 Inflammatory polyp 50.0 100 Histological types Tumors GIST Total 1 Polyps Total 11 Sessile polyp Comment: the most common tumors are GIST (57.2%) and vascular tumor (21.4%) Type 4 20,0% Type 1A 15.0% Type 3 5.0% Type 2B 5.0% Type 1B 45.0% Type 2A 10.0% Chart 3-1: types of vascular lesions according to Yano classification (n=20) Ulcer of unknown cause 55.6% Anasto -mosis ulcer 22.2% Ulcer due to Crohn NSAID 11.1% induced ulcer 11.1% Chart 3-2: types of ulcerative lesions (n=9) 3.2.2 Stool characteristic and lesion location, DBE insertion route Table 3-6: Stool color and lesion location Stool Lesion’s location in small intestine Total Proximal part Distal part characteristic Melena 41 14 55 Bloody 11 18 29 Total 52 32 84 p OR = 4.792 [95%CI 1.826 – 12.575], p 65 Total Neoplastic lesions 14 16 30 14 30 Diverticulum 21 25 16 25 Vascular lesions 13 20 20 Ulcer 9 Total 55 29 84 33 32 19 84 p p 0.05 p > 0.05 Comment: there was no difference in Hgb level and transfused RBC unit according to bleeding causes Table 3-9: Hgb level and transfused RBC unit according to lesion’s location Hgb level (g/l) RBC unit (250 ml/ unit) Lesion’s location n Medium ± SD n Medial (IQR) Proximal part 52 72.0 ± 17.7 45 5.6 (2.8 – 10.0) Distal part 32 77.7 ± 16.6 23 4.2 (2.8 – 7.6) p p > 0.05 p > 0.05 Comment: there was no different in in Hgb level and transfused RBC unit according to lesion’s location 13 3.3 TREATMENT OF SMALL INTESTINAL BLEEDING 3.3.1 The treatment methods applied in the study Small intestinal bleeding (n=84) Endoscopic hemostasis (n=29) n=6 Endoscopic intervention (n=23) n=44 Surgery (n=50) Medical (n=11) Imaging 3-1: Treatment methods applied for OSIB patients in the study Total (procedures) Loop + polypectomy (lesion) Electrocoagulation n Clipping Bleeding causes Adrenalin injection 3.3.2 Interventional hemostasis through DBE 29/84 patients (34.5%) received endoscopic hemostasis, of which 25/29 patients (86.2%) used one technique and the remaining patients (13.8%) required a combination of multiple hemostasis techniques Table 3-10: bleeding causes and endoscopic hemostasis techniques Vascular lesions 16 12 19 Diverticulum 7 Neoplastic lesions Tumors Polyps Ulcer 1 Total 29 17 13 33 Comment: Endoscopic hemostasis was mainly applied for vascular lesions (16/29, 55.2%) and diverticulum (7/29, 24.1% The two most commonly applied hemostatic techniques were clipping (17/33, 51.5%) and electrocoagulation (13/33, 39.4%) 3.3.3 Results of endoscopic hemostasis and rebleeding after discharge Among 29 patients undergoing endoscopic hemostasis, 100% of patients were clinically successful 6/29 patients underwent radical surgery later because of tumors and Meckel’s diverticulum 23/29 patients were discharged and followed for a mean of 122.2±80.4 weeks [0.4–312.1 weeks] 14 Table 3-11: Rebleeding in the endoscopic intervention group (n=23) Period after Patients had rebleeding (%) The remain non-rebleeding hemostasis patient ≤ 72 hours 23 72 hours - days 2* 21 – 30 days 21 31– 90 days 1* 20 > 90 days ** 19 Total (17.4%) (*) vascular lesions, (**) ulcers Comment: The rebleeding rate of endoscopic intervention group was 17.4% patients rebleeding within days were performed 2nd DBE, patient needed 2nd hemostatic procedure Follow-up until the end of the study, no further re-bleeding was recorded patients rebleeding after 30 days were operated because of recurrent anastomosis ulcer and vascular lesion (1 cm, Yano type 4) that did not completely cured after the first endoscopic intervention These surgical patients were followed up to the end of the study without rebleeding again 3.3.4 Surgery treatment Table 3-12: Bleeding causes of patients underwent surgical intervention Bleeding causes n % Tumors 26 52.0 Diverticulum 19 38.0 Meckel 18 non-Meckel (> cm with diffuse angiodysplasia) Vascular lesion(Yano type 4, > 1.5 cm) 6.0 Ulcer 4.0 Crohn’ disease ulcer, with obstruction Ulcer, bleeding after biopsy Total 50 100 Comment: surgery was applied mainly for tumors (52.0%) and diverticulum (38.0%) 100% of patients had resection of small intestine segment containing lesions, 5/50 (10%) patients were combined with adjuvant chemotherapy Follow-up of 50 patients for a mean period of 185.3 ± 75.8 weeks [54.3 – 332.0 weeks], no patient had rebleeding 15 3.3.5 Medical treatment Table 3-13: Bleeding causes of patients underwent medical treatment Bleeding causes n % Ulcers 54.5 Tumors 18.2 End-stage pancreatic tumors with small metastasis Diffuse intestinal lymphoma Diverticulum 18.2 Meckel * non- Meckel Vascular lesions Yano type 3/ end-stage kidney failure * 9.1 Total 11 100 (*) patients and their families did not consent to surgical intervention Comment: 54.5% of medical treatment patients were bleeding due to ulcers Following 11 patients for a period of 128.9±110.2 weeks [1.7–320.7 weeks]ư, there were 3/11 (27.3%) patients had rebleeding with lesions including small bowel tumor metastasis from pancreas tumor, intestinal lymphoma and ulcer 3.3.6 Comparison of rebleeding rates in treatment groups Chart 3-3: comparison of rebleeding rates in treatment groups Comment: By Kaplan-Meier and Log rank test, the study found statistically significant differences in the probability of rebleeding between surgical and endoscopic treatment (p