Purpose: The outcome of patients with upper gastrointestinal hemorrhage is greatly influenced by recurrence of bleeding, but it may be possible to identify patients who have a low risk for rebleeding, and can be discharged after a short hospitalization. To examine the effect of an early discharge protocol (length of hospital stay < or =3 days), we conducted a 2-year prospective study in patients with upper gastrointestinal bleeding at low risk for rebleeding, as selected by clinical and endoscopic criteria.
Methods: During the first year of the study, patients were managed according to the standard criteria by any of six surgical teams (control period). During the second year, patients were managed by only one surgical team under the early discharge protocol guidelines (study period).
Results: Overall, 488 of 942 (52%) patients were considered as low risk. Early discharge was achieved in 26 of 230 (11%) patients in the control period and in 191 of 258 (74%) in the study period (P <0.001). Age and number of compensated comorbidities did not affect the rate of early discharge. Length of hospital stay was reduced from (mean +/- SD) 6 +/- 2.7 days (control period) to 3 +/- 2.3 days (study period, P <0.001). No differences were observed in rates of rebleeding, need for surgery, readmission or mortality. By contrast, no differences in lengths of stay were observed during that time period among patients admitted with coronary artery disease, colorectal cancer, or acute pancreatitis.
Conclusion: Most patients with upper gastrointestinal bleeding who are at low risk for rebleeding can be discharged early, leading to important cost savings.