Upper gastrointestinal bleeding (UGIB) is one of the serious complications after major orthopedic operation, which is mainly caused by stress-related ulcer (SU) . While showing no symptom in mild cases, UGIB can affect the rehabilitation and life of patients in critical cases. The symptoms of UGIB are usually hard to be detected during the early postoperative stage, which brings some difficulties to the diagnosis and treatment. Usually, more cases of UGIB are reported after major orthopedic surgery. Shoulder and elbow surgery is in rapid development recently, and most of them are small and medium surgeries including internal fixation of peripheral shoulder and elbow fractures, ligament injuries of peripheral shoulder and elbow, repair and reconstruction of rotator cuff injury, shoulder and elbow arthroplasties, etc. Thus, the postoperative UGIB is seldomly reported in these cases. Clinically, the application of prophylactic drug on every patient with shoulder and elbow surgery in avoiding digestive tract bleeding will cause unnecessary medical waste. This study retrospectively analyzed the clinical data of shoulder and elbow surgery in our hospital for nearly 8 years and collected the high-risk factors of postoperative UGIB to better grasp the clinical data of UGIB after shoulder and elbow surgery, prevent postoperative UGIB, and optimize the guideline of clinical individualized treatment.

(1) Research object. From January 1, 2007 to December 12, 2015, the information of patients with shoulder and elbow surgery under the department of orthopedics and traumatology was collected, and the patients with UGIB were included in this study. Inclusive and exclusive criteria: within one week after operation, the patients had situations including hematemesis, melena and positive occult test. The bleeding of mouth, nasopharynx and biliary tract as well as preoperative active peptic ulcer and gastrointestinal bleeding should be excluded. (2) Research methods. The general and clinical data of patients were statistically analyzed. The general data includes demographic characteristics and lifestyle habits such as age, gender, smoking, drinking, etc. The clinical data includes causes of shoulder and elbow surgery, operation time, operation method, anesthesia, gastrointestinal clinical manifestations, UGIB diagnosis, clinical outcomes, laboratory results, history of peptic ulcer and hemorrhage, anticoagulant or antiplatelet drug history, etc. The risk factors for UGIB were analyzed by single-factor analysis and multiple-factor analysis. (3) Statistical analysis. SPSS 19.0 software was used for statistical analysis. The enumeration data was analyzed by Chi-square test, and the multiple-factor analysis was performed by Logistic regression analysis. A P value < 0.05 was regarded as statistically significant.

(1) General information and clinical data. A total of 33 patients developed UGIB after shoulder and elbow surgery, including 17 males (51.56%) and 16 females (48.44%) . The incidence rate was 3.36% (33/983) , and the average age was (67.609±13.948) years old. The operations that involved UGIB included shoulder and elbow arthroplasty, shoulder and elbow internal fixation, ligament repair and reconstruction, and open reduction and internal fixation of humeral shaft fracture. The average operation time was (131.00±93.36) mins. Gastrointestinal symptoms include hematemesis or vomiting coffee ground vomitus, melena or dark red colored bloody stool, abdominal pain or discomfort, nausea, etc. All the patients included in this study had positive fecal occult blood. A total of 27 cases (81.82%) underwent gastroscopy: 25 cases of gastric mucosal erosion and ulcer were diagnosed as SU (8 cases of which had active bleeding) ; 1 case had esophageal and gastric varices bleeding (the patient suffered from hepatitis B virus-related cirrhosis) ; 1 case had anastomotic ulcer after subtotal gastrectomy. Among the patients who received gastroscopy, 8 out of 23 patients who underwent rapid urease test for detection of Helicobacter pylori were positive. Based on clinical manifestations and the endoscopic diagnosis of upper gastrointestinal bleeding, there were 31 cases of SU, 1 case of esophageal and gastric varices bleeding, and 1 case of anastomotic ulcer after subtotal gastrectomy. All patients with UGIB were treated with proton pump inhibitors (PPI) , and 11 cases were treated with blood transfusion. 31 patients were discharged, and 2 patients died. One case died of UGIB, and the other died of respiratory failure caused by severe pulmonary infection. (2) Single-factor analysis of UGIB after shoulder and elbow surgery. Single-factor analysis was conducted on gender, age, international normalized ratio (INR) , platelet count, operation time, operation method, anesthesia, drinking history, smoking history, peptic ulcer or gastrointestinal bleeding history, anticoagulant or antiplatelet drug application, etc. The results showed that the risk factors of UGIB after shoulder and elbow surgery included male, advanced age, long time operation, smoking history, drinking history, peptic ulcer or gastrointestinal bleeding history, anticoagulant or antiplatelet drug usage, etc. (P <0.05) . (3) Non-conditional multiple-factor Logistic analysis of UGIB after shoulder and elbow surgery. The factors were taken as independent variable for non-conditional multiple-factor Logistic analysis. The results showed that male, advanced age, long operation time, previous history of peptic ulcer or gastrointestinal bleeding and application of anticoagulant or antiplatelet drug were the independent risk factors of UGIB after shoulder and elbow surgery (P <0.05) .

UGIB in patients with shoulder and elbow surgery is relatively uncommon. Advanced age, male, long operation time, history of peptic ulcer or gastrointestinal bleeding, usage of anticoagulant or antiplatelet drug are the independent risk factors of UGIB after shoulder and elbow surgery.