RESUMO
The authors presented the application of the Cox proportional hazards stepwise regression procedure in prognostic factor analysis of Hodgkin's disease. The data of 233 patients in our hospital from 1966 to 1989 were collected. There were 19 Variates used for univariate and multivariate analysis. The result showed that seven variates were important prognostic factors of Hodgkin's disease, namely; short-term effect, doses of treatment, pathological type, short breath, elevated ESR, clinical stage and lymph node enlargement. Among them, the probability of achieving complete remission (CR) after initial treatment is a favorable prognostic factor. Moreover, the rate of CR is consistently associated with clinical stage doses etc. It is important to make great efforts to avoid misdiagnosis. Early diagnosis and sufficient doses of medication are the crucial points for prolonged Survival.