593 Impact of Computed Tomography Angiography-First Approach on Length of Stay and Time to Definitive Therapy in Patients Presenting for Lower GI Bleed

Abstract

INTRODUCTION: The initial study of choice in patients who present to the emergency department (ED) for lower gastrointestinal bleeding (LGIB) is a topic of debate. It is unclear how patient outcomes are affected if computed tomography angiography (CTA) of the abdomen and pelvis is a preliminary study. A major teaching hospital implemented a CTA-first protocol in patients presenting for LGIB in December of 2014. Our goal was to determine how this policy impacted outcomes like hospital length of stay (LOS), time to definitive intervention, number of colonoscopies or embolizations performed, and yield or impression of colonoscopy. METHODS: Electronic medical records were queried for patients presenting for acute LGIB from January 2011 to December 2016. We gathered healthcare and demographic information in patients older than 18. SAS© software was used to perform student's t-tests, chi-square and multivariable linear regression analysis of the data. RESULTS: Five-hundred seventy-four encounters were analyzed, 297 before CTA-first policy (B-CTA) and 277 after CTA-first (A-CTA) policy implementation. The groups were without significant differences in sex, race, age, admission day of the week, hemoglobin or creatinine on presentation (Table 1). There were significant differences in heart rate and systolic blood pressure on presentation (Table 1). LOS in the B-CTA group was 5.8 days compared to 5.1 days in the A-CTA group (P > 0.05; Table 2). In those hospitalized for more than two weeks, interestingly, LOS did seem shorter overall A-CTA (P < 0.01; Figure 2). More CTAs were performed after policy implementation (Table 2). Multivariate analyses showed a 77% reduction in the number of colonoscopies (P < 0.0001) and a 47% decrease in early (those within 24h of admission) colonoscopies (P = 0.0082) (Table 3). Time to colonoscopy was not significantly reduced B-CTA versus A-CTA (Table 2). Colonoscopy impressions detailing the source of bleeding pre- and post-policy implementation were similar (Figure 1). There was no significant change in the number of or time to embolizations (Table 2). CONCLUSION: The CTA-first protocol was intended to facilitate effective triage of patients, targeted definitive therapy, and decreased overall LOS. Our findings suggest that for patients presenting for LGIB, this approach did not significantly improve LOS unless admitted for more than 2 weeks. Moreover, it is questionable that it facilitated more targeted intervention with embolization or colonoscopy.