BLOOD-TRANSFUSION FOR RADICAL HYSTERECTOMY BEFORE AND AFTER THE DISCOVERY OF TRANSFUSION-RELATED HUMAN-IMMUNODEFICIENCY-VIRUS INFECTION

Objective: To compare the patterns of perioperative blood transfusion in patients undergoing radical hysterectomy in time periods before and after the discovery of transfusion-related human immunodeficiency vir us (HIV) infection. Methods: We reviewed the medical records of all pa tients who underwent radical hysterectomy and pelvic lymphadenectomy a t Memorial Sloan-Kettering Cancer Center during two time periods, an e arly period (January 1, 1980 through December 31, 1993) and a late per iod (July 1, 1991 through June 30, 1993), The early period preceded an d the late period coincided with the era of increased awareness of tra nsfusion-related HIV infections. Results: One hundred twenty-eight pat ients underwent radical hysterectomy in the early period and 71 in the late period. In the late period, markedly fewer units of blood were t ransfused per patient (0.62 versus 3.5). Most patients in the early pe riod received at least one unit (117 of 128, 91%), compared with less than half (31 of 71, 44%) in the late period. For patients who receive d transfusions, notably fewer received homologous blood during the lat e (6 of 31, 19%) versus the early period (117 of 117, 100%). Using an unpaired t test and the 95% confidence interval (CI), we found that th e patients in the late period had a shorter mean postoperative length of stay (II versus 14 days, P < .0001, 95% CI +/- 1.3) and were discha rged with a significantly lower mean hemoglobin level (9.7 versus 11.4 g/dL, P < .0001, 95% CI +/- 0.35). The mean estimated blood loss was lower in the late period (756 versus 1598 mL). We defined the transfus ion index as the number of units transfused per 500 mL of estimated bl ood loss. The mean transfusion index was significantly lower during th e late period (0.38 versus 1.1, P < .001, 95% CI +/- 0.19). Conclusion s: During the perioperative period for radical hysterectomy and pelvic lymphadenectomy, the incidence of blood transfusion at our institutio n has markedly decreased over the past decade without immediate advers e effects on postoperative recovery. The reasons for this are probably multifactorial. However, the contribution of increased concern about transfusion-related HIV infections must be considered. Preoperative au tologous donation can notably decrease the need for homologous transfu sion.