Management practices for gastrointestinal hemorrhage related to portal hypertension in cirrhotic patients: evaluation of the impact of the Paris consensus workshop.

OBJECTIVES: The purpose of this before-after observational survey was to evaluate compliance with good clinical practice guidelines for gastrointestinal hemorrhage related to portal hypertension and the impact of the French Consensus Workshop held in Paris in 2003. METHODS: Data were recorded concerning episodes of gastrointestinal hemorrhage occurring in cirrhotic patients using a survey questionnaire in 2003 before the workshop and again in 2004. RESULTS: Seventy-six index episodes were included in 2003 and 84 in 2004 in patients attending French hospitals. Before hospital admission, primary prophylaxis was similar in 2003 and 2004, but beta blockers were used alone more often in 2004 for secondary prophylaxis (42% vs 19%, P=0.018). The time from onset of bleeding to hospital admission was greater than 12 hours for 43 and 42% of patients and was not shorter in the event of recurrent hemorrhage. At admission, vasoactive drugs were given earlier in 2004 (<2h: 68% vs 35%, P<0.001). Use of antibiotic prophylaxis was similar in 2003 and 2004 (70% vs 61%, P=0.098), and was more common for Child-Pugh B or C patients (P=0.044). CONCLUSION: The Paris Consensus Workshop enabled improved clinical practices. Weak points were insufficient screening for cirrhosis, long delay before admission, insufficient use of antibiotic prophylaxis which should be systematic.