Risk factors for clinically important upper gastrointestinal bleeding in patients requiring mechanical ventilation

OBJECTIVE: To evaluate the incidence and risk factors for clinically important upper gastrointestinal bleeding in critically ill patients requiring mechanical ventilation. DESIGN: In duplicate, blinded adjudicators determined the presence of clinically important gastrointestinal bleeding using a priori criteria, evaluating relevant clinical, laboratory, and diagnostic data. Cox proportional hazards regression analyses were used to examine baseline and time-dependent risk factors for bleeding. SETTING: Sixteen university-affiliated intensive care units (ICUs) in Canada. PATIENTS: A total of 1,077 critically ill ICU patients ventilated for at least 48 hrs. INTERVENTIONS: Patients were randomized to stress ulcer prophylaxis with intravenous ranitidine or nasogastric sucralfate; otherwise, management was at the discretion of the ICU team. MEASUREMENTS AND MAIN RESULTS: Demographic data included patient characteristics, Acute Physiology and Chronic Health Evaluation II score, and multiple organ dysfunction (MOD) score. Each day in the ICU, physiologic measurements including MOD score, feeding, and other treatment variables were recorded. The significant risk factors for upper gastrointestinal bleeding in the univariable analyses were low platelet count, maximum serum creatinine, maximum MOD score, maximum pulmonary component of the MOD score, maximum hepatic component of the MOD score, maximum renal component of the MOD score, enteral nutrition, and stress ulcer prophylaxis with ranitidine. The only independent predictors of bleeding in the multivariable analysis were maximum serum creatinine (relative risk = 1.16 [95% confidence interval = 1.02-1.32]), enteral nutrition (relative risk = 0.30 [95% confidence interval = 0.13-0.67]), and ranitidine administration (relative risk = 0.39 [95% confidence interval = 0.17-0.83]). CONCLUSIONS: In critically ill ventilated patients, renal failure was independently associated with an increased risk of clinically important gastrointestinal bleeding, whereas enteral nutrition and stress ulcer prophylaxis with ranitidine conferred significantly lower bleeding rates.