This represented a 0·1% excess of hospital admissions for cardiovascular or respiratory problems compared with age-matched and gender-matched controls who had not undergone an upper endoscopy.