Time to endoscopy and its effect on outcomes in acute nonvariceal and variceal upper gastrointestinal bleeding.

The role of early endoscopy in patients with nonvariceal upper gastrointestinal bleeding (UGIB) is controversial; in addition, the timing of endoscopy in patients with variceal hemorrhage has been poorly studied. We aimed to determine the effect of time to endoscopy on clinical outcomes in nonvariceal and variceal UGIB. We identified patients with acute UGIB admitted from the emergency room in an academic medical center. Patients were stratified by risk (low and high) and source of bleeding (variceal or nonvariceal). We examined time to endoscopy (within 3, >3-6, >6-12, and >12-24 h) and assessed the following outcomes: stigmata of bleeding, transfusions, rebleeding, surgery, mortality, ICU admission, readmission (within 42 days), weekday versus weekend admission, and length of stay (LOS). We identified 987 patients with nonvariceal UGIB and 288 patients with variceal UGIB. Clinical outcomes, including transfusions, rebleeding, surgery, mortality, ICU admission or readmission, did not vary as a function of the time to endoscopy. However, a shorter time to endoscopy was an independent predictor of identifying stigmata of bleeding (p < 0.001) and increases in time to endoscopy reduced the likelihood of identifying stigmata of bleeding to 66%, 35%, and 23% at 2, 6, and 12 h, respectively (p < 0.001). Endoscopy performed within 3 h was also associated with a significantly reduced LOS in low-risk patients with nonvariceal UGIB (p < 0.001). Our data indicate that although earlier endoscopy does not improve clinical outcomes, the earlier it is performed, the more likely stigmata are identified, and in certain types of patients (especially low-risk), the earlier discharge.