Frequency of upper gastrointestinal bleeding in a pediatric intensive care unit.
OBJECTIVE: To determine the frequency of upper gastrointestinal (GI) bleeding in pediatric ICUs. DESIGN: Prospective, descriptive study. SETTING: Pediatric ICU in a university hospital. PATIENTS: All children admitted to a pediatric ICU over a 55-wk period. INTERVENTION: None. MEASUREMENTS AND MAIN RESULTS: Upper GI bleeding was considered to be present if there was an episode of hematemesis or if any amount of blood was seen in drainage from a nasogastric tube. Sixty-three (6.4%) upper GI bleeds were detected among 984 patients: 5.2% in 698 patients who did not receive upper GI bleeding prophylaxis, and 9.4% in 286 patients who did receive some prophylaxis. Density was defined as the number of events/1000 days.patient. The mean density was 10.8 GI bleeding episodes/1000 days.patient in a pediatric ICU. A multivariate analysis detected four independent risk factors or risk markers for upper GI bleeding: high Pediatric Risk of Mortality score, coagulopathy, pneumonia, and multitrauma. Age, sex, hepatic and respiratory failures were identified as confounding variables. An upper GI bleeding episode was defined as being clinically important if hypotension, death, or transfusion occurred within 24 hrs after the bleeding. There were four clinically important GI bleeding episodes. All were caused, at least in part, by a coagulopathy. The GI bleeding was associated with a need for transfusion in four children, and with hypotension in two. CONCLUSIONS: The frequency of upper GI bleeding is substantial, but the rate of occurrence of clinically important upper GI bleeding is low, even in a pediatric ICU where most patients do not receive any prophylaxis.