Clinical factors predicting good outcome in upper gastrointestinal bleeding
The level of care provided to patients with upper gastrointestinal bleeding (UGIB) prior to endoscopy is usually based on clinical, non-endoscopic factors. The goal of this study is to identify features of the history, physical exam, and laboratory data which accurately predict an uncomplicated hospital course and which could help clinicians triage patients to an appropriate level of care prior to endoscopy. METHODS: We conducted a retrospective chart review of patients admitted through our Emergency Department with a diagnosis of UGIB between 7/93 and 7/95. Seventy-six patients were analyzed. Twelve predictor variables were recorded: age, h/o cirrhosis, co-morbid illness, initial systolic BP and HR, orthostatic vital signs, NG lavage color(red, coffee grounds, clear), presence of ascites, stool color(red, maroon, black, brown) , initial Hct, PT ratio, and BUN. Endoscopic diagnoses were recorded. An uncomplicated hospital course was defined by the absence of: rebleeding; need for surgery, interventional radiologic procedure, or endoscopic hemostasis; ICU admission; and death. RESULTS: Seventy-six patients were analyzed. Categorical variables were analyzed using ChiSquare or Fisher's Exact Test. Continuous variables were analyzed using logistic regression. Of the 12 variables, NG lavage color (p=0.02) and stool color (p=0.006) significantly differentiated between patients who had a complicated versus an uncomplicated hospital course. CONCLUSION: When evaluating patients with UGIB in the Emergency Department, NG lavage color and stool color are useful predictors of the need for a more aggressive level of care and potential need for intervention.
Strobel, J; Eloubeidi, M; Onken, J
Volume / Issue
Start / End Page
International Standard Serial Number (ISSN)
Digital Object Identifier (DOI)