Methods: Introduction of the acute upper gastrointestinal bleeding proforma during continued use of "consult forms" gave a proforma group (P) and a "consult form" group (C). The quality of referral information was prospectively assessed. Five critical fields made up the primary
outcome measure: Hb, INR, platelets, heart rate and blood pressure. Sample size calculation estimated that 73 referrals would be required to detect a 33% difference between groups P and C for the presence of all 5 of these outcome measures with p= 0.05 and beta = 0.8.
Results: A total of 87 referrals were examined: 29 group C and 58 group P. No group C and 69% of the group P (p < 0.001)referrals fulfilled the primary outcome measure. In addition no group C and 97% of group P (p < 0.001) referrals included some form of risk assessment.
Conclusions: An endoscopy referral proforma significantly improved the quality of information provided for endoscopy triage and significantly improved the communication of pre-endoscopic risk. This is likely to result in improved prioritization of endoscopy and might influence quality of care in acute upper gastrointestinal bleeding . The effect of this simple approach on acute upper gastrointestinal bleeding outcomes merits further investigation.