Endoscopic diagnosis,emergency therapy and outcome in 397 patients with acute gastrointestinal haemorrhage -a prospective study

Summary

The mortality of acute gastrointestinal haemorrhage in the literature varies significantly, depending on the study design and quality of the data. We therefore conducted a prospective study over 2.5 years of 397 patients with suspected acute Gl haemorrhage at the University Hospitals of Erlangen for internal quality control. In 99% of cases diagnostic endoscopy yielded one or more bleeding sources. Forty-six per cent of the patients bled from duodenal or gastric ulcers, 21% from oesophageal or gastric varices and 33% from other sources. Two hundred and twenty-eight of the 397 patients (57%) were initially treated endoscopically, 76 patients (19%) experienced a recurrence of a bleed. Patients older than 60 years of age with a haemoglobin below 8 g/dl had a significantly higher rate of recurrence of a bleed. The rate of complications during the hospital stay was 22% (n = 87), in-hospital mortality 17% (n=68). The rate of recurrence of the bleed (28 vs 20%), of complications (24 vs 18%) and mortality (20 vs 14%) was significantly higher for bleeding varices than for peptic ulcers. Patients with other causes of acute Gl-haemorrhage had a recurrence of the bleed in 13%, complications in 26% and a mortality of 19%. Bleeding varices as well as bleeding peptic ulcers and other causes of acute Gl haemorrhage still have a high mortality and require intensive medical surveillance besides diagnostic and therapeutic endoscopy.