A Retrospective Analysis of a Lower GI Bleeding Protocol Implemented in the Emergency Department at a Tertiary Care Hospital

Abstract

Introduction: Acute lower gastrointestinal (LGI) bleeding is a common cause of visit to the emergency department with subsequent hospitalization, morbidity and mortality in the United States. The management of acute LGI is not standardized. This retrospective quality improvement study assesses the impact of a standardized acute lower GI protocol for clinically significant bleeding implemented in the ED at a tertiary care hospital in the US, and interim results are presented. Methods: A multidisciplinary, institutional LGI bleed protocol was implemented, which addressed severity of bleeding and provided diagnostic management guidance, including CT angiography (CTA) and colonoscopy. A retrospective analysis of two groups, pre and post-protocol, was performed, with each having a defined period of 1.5 years pre and post protocol initiation (March 2014). All adult patients with clinically significant bleeding (ED, HR>100, or Hemoglobin drop > 1 g/dl, or 1 unit PRBU transfusion) were selected. ICD 9,10 codes for LGI bleeding were used and manual review was performed subsequently. Sixteen patients were randomly selected in each group for the interim analysis using Chisquare and Wilcoxon rank sum test. Results: There was equal representation of males and females (50%) in each group. The mean age of the pre-group was 62 (SD +-21) with the post-group slightly older 67 (SD +-10, P=0.4). The BMI was similar 28 (+-7) and 26 (+-5) respectively. The mean baseline admission characteristics were similar including Heart Rate (92 vs 90), Hemoglobin (10.6 g/dl vs 10.3 g/dl), platelets (222 vs 231), however not statistically significant (p>0.05). Prior NSAID (11 vs 12) and anti-coagulant (2 vs 2) use were similar. The hospital length of stay was statistically significantly shorter in the post-protocol group (9 days for pre-group vs 3 days, P=0.03), while the number of CTA (2 vs 5, p=0.2), colonoscopy (6 vs 8, p=0.4) and PRBU (0.8 vs 1.3, p=0.4) were higher, though non-significant.Figure 1Conclusion: This interim analysis of the first retrospective study of a standardized LGI bleeding protocol in the ED demonstrates a significantly shorter length of hospital stay post implementation. Given it's potential to reduce health care costs, further investigation with an increased sample size is needed.