Prognosis of human liver transplantation.

The results of publications on liver transplantation were diverse since several years, without model prognosis. The impossibility was due to the international system of measurement. We resorted to vector functions for calculating the ratios of biological values. We studied 2 samples with the same total number (35 patients) in the same conditions. We proposed 2 vector functions of transplants: (alpha)v1 weight/age donor and recipient in proportion to obtain a medium coefficient; (gamma)v2 ratio of biliary volume/700 mL (minimum secretion); beta was the coefficient of ratio ALT/AST (transaminases). After evaluation of 560 observations and mathematical control about 3000 numbers, we compared the samples with 10 parameters without significant difference between variances, means, other values; with consented errors alpha= beta = 0.05; gamma < 10(-7); means of relative errors = +/- 0.03 negligible. The results were verified by diverse tests (standard deviation of differences, chi2-test, relative risk, odds ratio, comparisons of distributions, parent population, equations of normality, partial correlations, partial regression coefficients, multiple regression, coefficient beta. Final results : quantitative prognosis by grading ; right responders to immunosuppressive treatment without complications, RR1 fast response (scores 3.5 ; 4) ; RR2 slow response (scores 2 ; 2.5 ; 3). Partial responders: very slow response (score 2; 2.5; 3) with transitory complications. Those patients were in recovery (81.5c/o). Wrong responders (score 2), 4 deaths (5.55c/o) by ARS; score 2.5, 1 death (1.5 5c/o) by ARS. We subtracted beta from these scores to differentiate them. Non-responders (score 1.5), 2 deaths (35c/o) by ARS.