Introduction
Massive gastrointestinal bleeding is a clinical condition that is poorly studied due to its abrupt presentation, involvement of various clinical services, and the high and rapid mortality of patients. We present the preliminary results of a multicenter registry that includes 21 hospitals from Spain and Hispanic America, analyzing factors related to mortality and need for intervention.
Aims & Methods
Since January 2023, 403 patients over 18 years old who presented to the emergency departments of participating hospitals with massive gastrointestinal bleeding, defined as bleeding accompanied by signs of shock and requiring transfusion of at least 2 units of blood within the first 24 hours of admission, were included. Demographic variables, comorbidities, treatments including NSAIDs, antiplatelets, and anticoagulants, clinical presentation variables, medical, endoscopic, and surgical variables, and a 6-month follow-up were collected.
The aim of this study was to analyze factors related to mortality and the need for intervention in patients with massive gastrointestinal bleeding.
Results
A total of 403 patients with a mean age of 64.5 ± 15.7 years were included (male 68.7%). Main comorbidities were cirrhosis (31%), diabetes (28%) and a previous UGIB (25%). 8.4% of patients had already been admitted to the hospital, and 29% were on antithrombotic (16.6% antiplatelets and 16.4% anticoagulants). On admission we observed SPB: 91±22 mmHg, heart rate 101 ± 21; and an altered mental status in 31%. Mean hemoglobin levels 7.6±2.5 g/dl, urea: 98±67, albumin 2.16±1.3, INR: 3.9±7. The most common causes were esophageal varices (29%), gastric ulcer (25%) and duodenal ulcer (25%). Endoscopic intervention was needed in 60%, with mortality rate of15.4%. Mean admission was 8.6±10 days, and patients needed 4±3 red blood cells units. When analyzing in-hospital mortality, we observed that the age of patients who died was higher than that of survivors (68.2±15 vs. 63.6±16; p=0.033), with the former requiring more packed red blood cells (5±5.5 vs. 3.5±2; p=0.043). Patients who died had a lower frequency of gastric ulcer (14% vs. 27%; p=0.034), a higher frequency of altered mental status (42.6% vs. 29%; p=0.036), and a lower rate of endoscopic therapy applied (43.3% vs. 63.6%; p<0.0001). In logistic regression, gastric ulcer etiology and endoscopic therapy were protective factors against mortality (Table 2).
Regarding the need for intervention, more patients with liver cirrhosis required it (73% vs. 60%, p=0.018), patients with hematemesis (73% vs. 52%; p<0.0001), patients without melena (60% vs. 76%; p=0.002), with syncope (76.3% vs. 61%; p=0.06), etiologies other than gastric ulcer (78% vs. 56%; p=0.028), and the presence of esophageal varices (83% vs. 57%; p<0.0001). In binary logistic regression, liver cirrhosis, hematemesis, and syncope were risk factors for the need for intervention.
Conclusion
Age, altered mental status, and lower frequency of endoscopic therapy are risk factors for mortality due to massive gastrointestinal bleeding. Gastric ulcer as an etiology and receiving endoscopic therapy acted as protective factors against mortality. Indeed, in both situations, it might be related to an opportunity window to receive appropriate treatment, especially when compared with more cataclysmic types of bleeding, such as variceal or posterior duodenal bulb wall, which sometimes can be challenging and utterly abrupt. Regarding the need for intervention, liver cirrhosis, hematemesis, and syncope were independent risk factors for the need for endoscopy, surgery, or interventional radiology.