Study	Reason for exclusion
Gameel 2005	Randomised clinical trial evaluating band ligation (n = 16), sclerotherapy (n = 17), or no intervention (n = 17) for primary prevention of variceal bleeding in participants with portal hypertension secondary to schistosomiasis and 'high risk' varices; 84% of the participants also had chronic viral hepatitis and cirrhosis. One participant in the band ligation group died compared with none in the no intervention group while no bleeding episodes occurred in the band ligation group in the first six months compared to three bleeding episodes in the no intervention group. This trial, which is published in abstract form only, was excluded because the portal hypertension was secondary to schistosomal portal hypertension.
Lim 2009	Retrospective observational study undertaken in 258 adults with cirrhosis awaiting liver transplantation who underwent upper gastrointestinal endoscopy. Of these, 101 had varices deemed to be at high risk of bleeding and underwent banding until variceal eradication was achieved or they were transplanted. Failed prophylaxis occurred in 2 participants (2%), and there were 3 episodes (1.2%) of acute haematemesis from band‐induced ulceration. One patient (1%) had mild oesophageal stricturing postbanding without dysphagia. None died. This study was excluded as it was observational with no control group.
Omar 2000	Randomised clinical trial evaluating band ligation (n = 36) versus no intervention (n = 38) for primary prophylaxis of variceal bleeding in participants with non‐alcoholic cirrhosis and/or schistosomal hepatic fibrosis and large oesophageal varices. No deaths were reported. Variceal bleeding occurred in one participant in the banding group compared to four in the no intervention group (P > 0.05). This trial, which was published in abstract form only, was excluded as no separate analyses were provided of the participants with and without cirrhosis.