[Blood transfusion and survival of patients with neoplasms of the colon and rectum].
The fate of 248 patients diagnosed of colorectal malignancies between 1973 and 1982 and subjected to radical surgery with curative purposes was analysed. The effect of peroperative whole blood and packed red cells transfusions, along with other co-variants, on the survival of the patients was studied. It was found, by means of the Cox analysis, that whole blood and the Duke's staging system had the most important correlation (p = 0.003 and p = 0.02, respectively) on the patients' prognosis and survival, no such relationship being found with packed red cells transfusion (p = 0.25). Dose-related association between the relapse index and the number of whole blood units transfused was shown by the Mantel's global trend test (p = 0.015). The different effect of the transfusion products would confirm the hypothesis that the relationship between transfusion and poorer prognosis, thought to be due to patient's immunosuppression and hence higher relapse index, could be determined by any ill-defined plasmatic factor.