Summary
Human and canine liver transplants representing two sample populations of living and dead, were studied by means of multivariate discriminant analysis using ten coagulation assays as input parameters. This procedure produced correct sample classifications at the .975 significance level in all cases. It is suggested that such a procedure, or similar procedures, may be used, when integrated within the proper information system, as an early adjunct to prognosis. It is further concluded that coagulation parameters represents, at least in part, the basis set which spans a multivariate discriminant partition space for liver transplantation.
