[A multivariate prognosis model for Hodgkin disease]

The authors describe the application of the linear discriminatory analysis as a pattern of prognosis for a group of 83 irradiated patients with Hodgkin disease. They investigated which combination of primary characteristics allows the most accurate discrimination between a group of patients with favorable prognosis and a group with unfavorable prognosis. Stage, histology, age, and clinical symptomatology proved to be valuable discriminatory factors. The accuracy of the discriminatory criterium is 75%. The results are compatible with the temporally comparable observations about prognostic factors. The method is suitable for gaining further quantitative knowledge.