Current issues of ACP Journal Club are published in Annals of Internal Medicine


Endoscopic injection reduced rebleeding in bleeding peptic ulcers

ACP J Club. 1992 Sept-Oct;117:38. doi:10.7326/ACPJC-1992-117-2-038

Source Citation

Oxner RB, Simmonds NJ, Gertner DJ, Nightingale JM, Burnham WR. Controlled trial of endoscopic injection treatment for bleeding from peptic ulcers with visible vessels. Lancet. 1992 Apr 18;339:966-8.



To determine whether endoscopic injection of adrenaline and ethanolamine oleate is an effective treatment for adults with bleeding or nonbleeding peptic ulcers with visible vessels.


Randomized controlled trial. Analysis was by intention-to-treat.


A British district general hospital.


All adult patients (n = 555) hospitalized for suspected gastrointestinal hemorrhage had endoscopy within 24 hours of admission. 98 patients had bleeding or nonbleeding visible vessels within a peptic ulcer, and 93 (53 men; mean age, 69 y, range 18 to 96 y) were randomized.


The adherent clots were removed from the peptic ulcers during endoscopy, and all patients received histamine antagonists or omeprazole plus standard medical care from physicians blinded to injection status. 48 patients received injection therapy. Using a flexible variceal needle injector, endoscopic injections of 1 to 2 mL of 1 in 10 000 (0.1 mg/mL) adrenaline were made around (4 to 6 injections) and into the vessel followed by a 1- to 2-mL injection of 5% ethanolamine oleate into the vessel. The mean volume of adrenaline used was 8 mL (range 5 to 17.5 mL) and of ethanolamine was 0.8 mL (range 0 to 2 mL). All patients with serious rebleeding received additional injection treatment.

Main outcome measures

Repeat endoscopies were done 5 days after initial endoscopy, before discharge, and 6 weeks later. Rebleeding was defined as fresh hematemesis or melena associated with either shock (pulse > 100 beats/min, systolic blood pressure < 100 mm Hg, or a fall in hemoglobin > 20 g/L in 24 h) or fresh blood found during endoscopy. Chart audit provided data on blood transfusions, length of stay, surgery, and death.

Main results

Rebleeding occurred in 8 patients (17%) in the injection group and 21 (47%) in the control-treatment group (P = 0.01). {This absolute risk reduction of 30% means that 4 patients would need to receive endoscopic injection (rather than usual care) to prevent 1 additional patient from rebleeding, 95% CI 2 to 9; the relative risk reduction was 64%, CI 30% to 82%.}* Rebleeding was treated with injection therapy in 5 patients in the injection group and 11 patients in the control group. The groups did not differ for number of blood transfusions; length of hospital stay; surgery (4 in the injection group vs 8 control group); or death (4 in injection group vs 9 in the control group).


Injection therapy using adrenaline and ethanolamine oleate reduced rebleeding for patients with bleeding or nonbleeding peptic ulcers and visible vessels.

Source of funding: Not stated.

Address for article reprint: Dr. W.R. Burnham, Department of Gastroenterology, Oldchurch Hospital, Romford, Essex RM7 0BE, United Kingdom.

*Numbers calculated from data in article.


Endoscopic injection reduced further bleeding and need for surgery in patients with bleeding duodenal ulcers

The mortality rate from bleeding ulcers has remained stable over the last 30 years and ranges from 6% to 10% (1). Because no medical therapy is effective for the acute treatment of bleeding ulcers, nonsurgical means of controlling bleeding have been sought. Various endoscopic forms of hemostatic therapy that use heat to stop bleeding have been assessed. On the basis of several randomized trials, a 1989 National Institutes of Health Consensus Conference concluded that bipolar electrocoagulation and heater probe were the treatments of choice in patients with bleeding ulcers (1).

Recently, the use of injection therapy has been evaluated. A major attraction of injection therapy is its simplicity: An injection catheter is the only equipment required. A number of injection solutions (absolute alcohol, polidocanol, epinephrine [1:10 000], saline) have been used with success. At least 7 previous trials have shown some benefit of injection therapy when compared with standard management of patients with bleeding ulcers. The trials by Oxner and colleagues and Moretó and colleagues further support the use of injection therapy.

The clinical and endoscopic entry criteria must be examined carefully when assessing therapeutic trials for bleeding ulcers. Clinical evidence of major bleeding (e.g., hemodynamic instability, transfusions) increases the likelihood of further bleeding, surgery, and death (1). Endoscopic features provide even better prognostic information. Patients with active bleeding or nonbleeding visible vessels have much higher rates of further bleeding and need urgent intervention more often than those with clots, flat pigmented spots, or clean-based ulcers (2). Patients without high-risk features generally do well with standard management alone and should not be included in therapeutic trials. It should be noted that patients with high-risk features are in the minority: fewer than 20% of those suspected of gastrointestinal hemorrhage in the Oxner study.

The study by Oxner and colleagues had no specific clinical entry criteria other than suspected upper gastrointestinal hemorrhage, and only half the patients had hemodynamic instability. All patients had visible vessels, and 5% had spurting vessels (the highest risk endoscopic feature). Rebleeding was significantly decreased with injection therapy. Unfortunately, the authors treated rebleeding episodes in the control group with injection therapy so that the relative effect of treatment on the more important and less subjective parameters such as surgery, hospital days, and transfusions cannot be properly assessed.

As in most trials of endoscopic therapy, the 2 trials under discussion failed to show a significant decrease in mortality with treatment. This finding may relate to the relatively small size of the individual studies. Recent meta-analyses have shown a significant decrease in mortality, evident only in patients with active bleeding or non-bleeding visible vessels, providing further justification for the use of endoscopic therapy (3).

Most articles now suggest that endoscopic hemostatic therapy should be used in patients with bleeding ulcers who have clinical and endoscopic features predicting a high risk for further bleeding or death (1). At present, bipolar electrocoagulation, heater probe, and injection therapy can be considered comparable in safety and efficacy for the treatment of bleeding ulcers.

Loren Laine, MD
University of Southern California Medical CenterLos Angeles, California, USA


1. N.I.H. Consensus Conference. Therapeutic endoscopy and bleeding ulcers. JAMA. 1989;262:1369-72.

2. Laine L, Cohen H, Brodhead J, et al. Prospective evaluation of immediate versus delayed refeeding and prognostic value of endoscopy in patients with upper gastrointestinal hemorrhage. Gastroenterology. 1992;102:314-6.

3. Cook DJ, Guyatt GH, Salena BJ, Laine LA. Endoscopic therapy for acute nonvariceal upper gastrointestinal hemorrhage: a meta-analysis. Gastroenterology. 1992;102:139-48.