25 January 2007 (online)
Ramsoekh D, Leerdam van ME, Rauws EA, et al. Outcome of peptic ulcer bleeding, nonsteroidal anti-inflammatory drug use, and Helicobacter pylori infection. Clin Gastroenterol Hepatol 2005; 3: 859 - 864
This epidemiologic survey focused on two university hospitals and 12 regional hospitals around Amsterdam, in which 14 % of patients admitted with hematemesis, melena, or hematochezia, or who developed peptic ulcer bleeding as inpatients, died. Mortality rate was nearly double the mortality rate of 5 - 8 % reported by specialized units that emphasize prompt resuscitation, early endoscopic diagnosis, and treatment followed by dedicated multidisciplinary postoperative care  . Also, the 19 % recurrent bleeding rate and 7 % need for surgery rate are higher than rates reported elsewhere, and no doubt contributed to the overall mortality.
Some aspects of this interesting paper deserve to be underlined. First, the surprisingly high mortality rate is in line with the outcome of acute upper gastrointestinal bleeding in other European epidemiologic surveys . This raises the issue of selection bias, which is often present in clinical trials, where patients with severe or life-threatening comorbidity, or who are older, or have coagulation disorders for example, are usually excluded. This results in a selected population that cannot be compared with the population of epidemiologic surveys that include all patients. Second, patients with severe acute upper gastrointestinal bleeding should be offered optimal management in specialized units. Management might have been suboptimal in regional hospitals, where early assessment and expert endoscopic intervention might not always be available. Most patients in this study who were given endoscopic treatment received epinephrine injection therapy alone, rather than combining this with a second hemostatic method, either thermal or mechanical, as recommended for patients with high-risk stigmata . Last but not least, mortality was significantly higher in patients already admitted for other illness, 28 % vs. 5.3 % for newly admitted patients (P < 0.01), respectively. Severe or life-threatening comorbidity was present in 76 % of the patients who died. Therefore, the high mortality rate reported in this survey was mainly attributed to the high mortality rate among inpatients with severe comorbidity. The majority of deaths in these high-risk patients could not have been prevented, but they resulted from the comorbidity, independently of any effective endoscopic intervention.