M. Shimada et al., ESTIMATION OF DEGREE OF LIVER-CIRRHOSIS USING A FIBROSIS SCORE - A MULTIVARIATE-ANALYSIS OF CLINICAL-PARAMETERS AND RESECTED SPECIMENS, Hepato-gastroenterology, 41(2), 1994, pp. 177-180
We designed this study to estimate the degree of liver cirrhosis using
clinical parameters and surgically resected specimens. One hundred an
d twenty-nine cases who underwent hepatic resection in the Department
of Surgery II, Kyushu University Hospital during the period between Ap
ril 1985 and July 1991 for whom data for all 29 variables evaluated we
re available, were admitted to this study. On the basis of the histolo
gical findings of fibrosis of the liver, the nonneoplastic part of the
resected specimens were classified into 3 groups; Z0, no cirrhosis (n
= 63), Z1, mild cirrhosis (n = 38), and Z2, severe cirrhosis (n = 28)
. A univariate analysis revealed 14 significant variables. After multi
ple logistic regression analysis of these, five independent variables
(low platelet count, female, low value of hepaplastin test, a high Pug
h's score and a high value of ICG) were identified. We obtained a fibr
osis score using the coefficient of each above-mentioned variable. Thi
s score increased the discriminative power. The fibrosis score is ther
efore considered useful for estimating the severity of liver cirrhosis
.