520 Pre-Endoscopic Risk Stratification Scores Associated With Intubation for Emergent Endoscopy for Severe Gastrointestinal Hemorrhage

Abstract

INTRODUCTION: Approximately 300,000 patients are admitted annually for gastrointestinal bleeding (GIB) in the US and many require emergent endoscopy. Many patients are prophylactically intubated to protect the airway in hopes of avoiding respiratory complications intra-operatively. Currently there are no guidelines to determine which patients should be intubated prior to endoscopy. STUDY AIM: To identify pre-procedural factors associated with intubation compared to those who undergo endoscopy without intubation and compare post-procedural outcomes. METHODS: Retrospective chart review from Sept 2017 to Dec 2018 of all emergent endoscopies for acute GIB. Data are presented as means with 95% ± t-test confidence intervals, or frequency (%). Analysis of variance (ANOVA), or univariate logistic regression was used for continuous variables, the Pearsons χ2 test was used for categorical variables. RESULTS: There were 210 procedures performed during this time frame. 142 patients were included in the final analysis after excluding endoscopies done for non-GIB indications. There was no statistically significant difference between the intubated and non-intubated groups for the following: gender, race, BMI, number of transfusions needed, or anticoagulation use [Table 1]. Mean Glasgow-Blatchford scores (GBS) and APACHE scores were higher in the intubated group than in the non-intubated group (12 vs. 10, P = .03 and 20 vs. 12, P < .01, respectively). In the intubated group, 28% of patients developed pneumonia (PNA) [Table 1]. The difference in cardiovascular outcomes trended towards significance (P = 0.07). The intubated group had higher cardiac arrest (5.7% vs. 0%) and mortality rates (24% vs. 11%) than the non-intubated group. CONCLUSION: We found that higher GBS and APACHE scores were associated with intubation. We also found that patients with cirrhosis or who were on pressors were more likely to be intubated. In our study, 28% of intubated patients developed pneumonia, whereas prior studies report a 14% incidence of pneumonia after intubation. The decision to intubate a patient is complex but gastroenterologists and intensivists must weigh the risks associated with intubation, such as PNA, ICU delirium, and increased length of stay in the ICU, and avoid reflexively intubating all patients. Our study shows significant differences in pre-procedure GBS and APACHE scores that could be useful for future studies as a possible tool to aid in this decision-making process.