Upper gastrointestinal stress ulceration was once viewed as a potentially life-threatening complication of critical illness, for which effective prophylactic measures were endorsed. However, the incidence of gastrointestinal bleeding in critically ill patients varies depending on the population studied, the definitions employed, and the methods used to diagnose the bleeding.
Microscopic gastrointestinal bleeding is common in critically ill patients and may be due to nasogastric trauma more often than stress ulceration. This type of bleeding rarely progresses to bleeding that is clinically relevant. Overt bleeding is defined as hematemesis, bloody gastric aspirate, melena, or hematochezia; this is less common, occurring in approximately 5 per cent of a heterogeneous group of intensive care unit (ICU) patients. Clinically important bleeding can be defined as overt bleeding and one of the following: drop in blood pressure of 20 mmHg within 24 h of bleeding, blood pressure drop of 10 mmHg and heart rate increase of 20
beats/min on orthostatic change, or hemoglobin decrease of 20 g/l and transfusion of 2 units of blood within 24 h ( Co.o.k,®La.l 1994). Recent reports have suggested that the incidence of clinically important gastrointestinal bleeding is extremely low, and question the widespread use of stress ulcer prophylaxis for all critically ill patients (Cook. etai 1994; Z.an.d.stra,a.Qd SíOyl®Qb®®k,19.9.4.). This decline in clinically important bleeding rates may be due to improvements in critical care over the last decade (e.g. more aggressive hemodynamic support, earlier attention to enteral nutrition, avoidance of corticosteroids in sepsis, etc).
One of these studies, demonstrating a decreased incidence, is a prospective multicenter natural history study evaluating the incidence of, and risk factors for, clinically important bleeding (Cookela/ 1994). The study population consisted of consecutive patients more than 16 years old who were admitted to four mixed medical-surgical ICUs. Patients were followed throughout their stay in the ICU for the development of clinically important gastrointestinal bleeding. Of the 2252 patients followed in this study, 33 developed clinically important gastrointestinal bleeding (1.5 per cent; 95 per cent confidence interval, 1.0-2.1 per cent). Gastrointestinal bleeding occurred a mean of 14 ± 1. days after admission to the ICU.
Was this article helpful?