OBJECTIVEThe objective of this chart review was to determine the frequency of transfusion and prevalence of anemia (hemoglobin result < 100 g/L) in patients receiving chemotherapy.DESIGNThis study was a retrospective review of medical charts.SETTINGPatients receiving chemotherapy were included from 12 tertiary care comprehensive cancer centres across Canada.MAIN OUTCOME MEASUREThe primary study outcome measure was red blood cell transfusion rate, controlling for patient variables.RESULTSThe 616 patients included had started chemotherapy in January-June 1992. For each subject, data collection finished 4 weeks after the end of the first regimen or after a maximum follow-up period of 26 weeks. Seventy-two patients (12%; 95% confidence interval 9.5% to 14.5%) were transfused for anemia (reasons other than blood loss), and 28% (95% confidence interval 24.5% to 31.5%) of the subjects were anemic during treatment. The univariate analyses of transfusion for anemia yielded significant associations with prognostic factors. In the multivariate analyses, platinum (odds ratio [OR] = 6.69) and anthracycline (OR = 3.56) chemotherapy, baseline hemoglobin (OR = 0.96) and disease stage (OR = 1.72) were statistically significant contributors.CONCLUSIONIn this patient cohort, red blood cell transfusion was infrequent (12%). However, patient groups at high risk of transfusion could be identified, with platinum-based chemotherapy being the most significant contributing factor. The information obtained from this multicentre study may prove helpful in developing supportive care guidelines for the management of chemotherapy-related anemia requiring transfusion.

The objective of this historical prospective review of 12 medical centres across Canada was to determine the prevalence of anaemia and the frequency of transfusion in patients (pts) receiving chemotherapy for the treatment of selected cancers (lung, breast, ovary, sarcoma Hodgkin's/non-Hodgkin's lymphoma, melanoma, bladder, colorectal). The 616 pts recruited had started chemotherapy in January–June 1992, continued to have treatment at regular intervals for at least 6 weeks (mean duration 24 weeks), and had received no chemotherapy within the previous 12 months. The pts were not treated with any investigational drug. Data collection finished 4 weeks after the end of the first chemotherapy regimen, to a maximum follow-up period of 26 weeks. A full multivariate analysis of transfusion and anaemia will be presented. Preliminary data indicate that 87 pts (14%) were transfused—72 (12%) for anaemia. In the different diagnostic groups (n > 30), 4% of pts with colorectal cancer were transfused for anaemia (13% anaemic), compared with 5% of breast cancer pts (17% anaemic), 24% non-Hodgkin's lymphoma pts (53% anaemic), 25% ovarian cancer pts (51% anaemic), and 28% of lung cancer pts (52% anaemic). For all subjects, the mean baseline Hb level of pts not transfused was significantly higher than that of pts transfused for anaemia overall (12.8 vs 11.8 g/dl, P < 0.001), in the lung cancer sub-group (13.6 vs 12.7 g/dl, P < 0.05), and the ovarian cancer sub-group (12.1 vs 10.6 g/dl, P < 0.05). Overall, the mean nadir Hb level was also significantly higher in pts not transfused vs pts transfused for anaemia, and in the 5 sub-groups with n > 30 (P < 0.001). The results of the full analysis, including stage of disease and chemotherapy regimen, will identify factors associated with anaemia and transfusion. The objective of this historical prospective review of 12 medical centres across Canada was to determine the prevalence of anaemia and the frequency of transfusion in patients (pts) receiving chemotherapy for the treatment of selected cancers (lung, breast, ovary, sarcoma Hodgkin's/non-Hodgkin's lymphoma, melanoma, bladder, colorectal). The 616 pts recruited had started chemotherapy in January–June 1992, continued to have treatment at regular intervals for at least 6 weeks (mean duration 24 weeks), and had received no chemotherapy within the previous 12 months. The pts were not treated with any investigational drug. Data collection finished 4 weeks after the end of the first chemotherapy regimen, to a maximum follow-up period of 26 weeks. A full multivariate analysis of transfusion and anaemia will be presented. Preliminary data indicate that 87 pts (14%) were transfused—72 (12%) for anaemia. In the different diagnostic groups (n > 30), 4% of pts with colorectal cancer were transfused for anaemia (13% anaemic), compared with 5% of breast cancer pts (17% anaemic), 24% non-Hodgkin's lymphoma pts (53% anaemic), 25% ovarian cancer pts (51% anaemic), and 28% of lung cancer pts (52% anaemic). For all subjects, the mean baseline Hb level of pts not transfused was significantly higher than that of pts transfused for anaemia overall (12.8 vs 11.8 g/dl, P < 0.001), in the lung cancer sub-group (13.6 vs 12.7 g/dl, P < 0.05), and the ovarian cancer sub-group (12.1 vs 10.6 g/dl, P < 0.05). Overall, the mean nadir Hb level was also significantly higher in pts not transfused vs pts transfused for anaemia, and in the 5 sub-groups with n > 30 (P < 0.001). The results of the full analysis, including stage of disease and chemotherapy regimen, will identify factors associated with anaemia and transfusion.