RÉSUMÉ
This study was to evaluate the epidemiological characteristics, etiology and therapeutic outcome of active upper gastrointestinal bleeding in patients who underwent emergency gastroscopy outside official hours at the Gastroenterology Unit, King Chulalongkorn Memorial Hospital. From January to December 2002, 103 emergency gastroscopies were performed in 99 patients. There were 66 men and 33 women (mean age 55.4 years, range 22-98 years). Causes of bleeding were esophageal varices (29/99; 29.3%), gastric ulcer (25/99; 25.3%), duodenal ulcer (9/99; 9.1%), gastric varices (9/99; 9.1%) and miscellaneous (12/99; 12.1%). Etiology of bleeding was uncertain in 10.1 per cent of the cases. Therapeutic modalities for variceal bleeding were banding (78.6%), sclerotherapy (10.7%) and glue injection (10.7%). Endoscopic therapies for patients with non variceal bleeding were: epinephrine injection with bipolar coaptation (48.1%), epinephrine injection only (11.1%), bipolar coaptation alone (7.4%), heater probe (7.4%), epinephrine injection combined with heater probe (11.1%), epinephrine injection with bipolar coaptation and hemoclipping (7.4%), hemoclipping (3.7%), epinephrine injection with hemoclipping (3.7%). Initial hemostasis was achieved in 91.2 per cent of the patients (91/99). Recurrent bleeding within 72 hours developed in 9.1 per cent of patients (9/99). Of these, eight patients (88.9%) underwent re-endoscopy and bleeding was stopped in 62.5 per cent (5/8). And 2.0 per cent of patients (2/99) had to go for emergency surgery after failed therapeutic endoscopy. Overall mortality was 15.2 per cent (15/99). In conclusion, emergency gastroscopy can offer not only diagnostic but also therapeutic modality for patients with acute upper gastrointestinal bleeding. Endoscopic therapy is effective for both initial hemostasis and recurrent bleeding.