A rectal bleeding algorithm can successfully reduce emergency admissions

Abstract

Aim: Acute lower gastrointestinal bleeding (LGIB) is a common cause of emergency admissions yet rarely requires blood transfusion or radiological/surgical intervention. We aimed to develop a risk assessment tool to identify patients with acute LGIB who can be safely managed in primary care.

Method: We retrospectively applied an existing nomogram to 20 admissions to obtain criteria that could predict the need for transfusion. We simplified the algorithm to three criteria and developed an associated care pathway. If haemoglobin was > 13 g/dl, systolic blood pressure > 115 mmHg and the patient was not anticoagulated, admission could be avoided. These criteria were then applied to 57 prospective patients attending during a 16-week period. This was implemented with education of primary and secondary care staff, access to an emergency clinic and provision of patient information.

Results: We applied our algorithm and care pathway to 57 patients with uncomplicated rectal bleeding. Thirty-five per cent (20/57) of potential admissions were avoided. Instead, patients received written information and underwent flexible sigmoidoscopy as outpatients within 6 weeks. One discharged patient was readmitted from endoscopy with severe colitis. There were no other readmissions or complications. Of the 36 patients for whom the algorithm predicted admission was needed, 33% (12/36) were anticoagulated, 94% (34/36) had haemoglobin < 13 g/dl and 42% (15/36) had a systolic blood pressure < 115 mmHg. Only one admission (1.8%) did not fulfil the admission criteria and could have potentially been avoided. Avoidable admissions reduced from 50 to 1.8%.

Conclusion: The application of a simple rectal bleeding algorithm can safely prevent unnecessary admissions.

Keywords: Rectal bleeding; outpatient management; triage.