I read with interest the article on measurement of hepatic venous pressure gradient in patients with active variceal bleeding.1 I would like to mention one therapeutic aspect in this article that I suggest is not appropriate. The authors randomized the patients with cirrhosis and active bleeding into two arms—injection sclerotherapy and band ligation—and measured the hepatic venous pressure gradient before and after the procedure until the 5th day of admission. They concluded that injection sclerotherapy had caused a sustained increase in hepatic venous pressure gradient, which is followed by a higher rebleeding rate. The authors did not administer any pharmacological treatment to the patients who exhibited signs and symptoms of acute variceal bleeding within 12 hours of admission.
I believe it is inappropriate to withhold vasoactive drugs from patients who experience active variceal bleeding. The authors claim that at the time they designed the study, endoscopic treatment was the treatment of choice for acute variceal bleeding.2 However, in an American Association for the Study of Liver Diseases single-topic symposium in 1998, pharmacological treatment was established as an effective treatment for variceal bleeding.3 Furthermore, D'Amico et al.4 published an elegant meta-analysis in 1999 in which they concluded that pharmacological treatment should be started immediately even if variceal bleeding is suspected before endoscopic confirmation.
The present study was performed between 1998 and 2001, the period in which the pharmacological treatment consensus had already been accepted. I believe that researchers should either stop or modify a study when a consensus on treatment modalities is altered completely during the study period.