Factors associated with failure of propranolol for the prevention of first bleeding in cirrhotic patients. The Study Group of Primary Prophylaxis
Calès, P.
Gastroenterologie Clinique et Biologique 16(6-7): 504-510
1992
ISSN/ISBN: 0399-8320 PMID: 1526413 Document Number: 395951
Beta blockers are effective in the prevention of first bleeding of esophageal varices in cirrhotic patients; however, the bleeding incidence is reduced by only 20 percent. This study was aimed at investigating the factors associated with the occurrence of first bleeding and death rate in patients receiving propranolol for primary prevention. One hundred and nine patients were included in the study, 92 percent of whom were alcoholic. According to Child-Pugh's classification, the distribution was grade A: 16 percent, grade B: 36 percent, grade C: 48 percent. During the study, 18 patients bled and 21 died. In survivors without bleeding, the median follow-up was 14 months. The predictive value of 34 biochemical or clinical variables, recorded at entry or during the course of the study, was first tested using univariate analysis. A higher initial Child-Pugh score and a smaller decrease in Child-Pugh score during the study were significantly associated with an increased bleeding risk; the above factors along with a low initial value and a higher decrease in plasma sodium concentration during the study were significantly associated with an increased death rate. Next, we tested these factors as well as others with potential explicative value, using multivariate analysis according to a multiple regression model.