geographical area. The California, Florida, and Massachusetts cord blood banks recruited a lower percentage of minorities than the corresponding bone marrow donor centers. In four of the five areas studied, cord blood banks recruited a lower percentage of minorities (in comparison with the census data for the corresponding collection hospitals).
A more recent study from the American Red Cross showed that racial diversity can be achieved in a national network of cord blood banks (Ballen et al., 2004). The population of that network is 64 percent Caucasian, 16 percent African-American, 12 percent Hispanic, 4 percent Asian, 1 percent Native American, and 3 percent other. Diversity was achieved by focusing collections in specific geographic areas; Detroit, for example, had the highest percentage of African-American donors and San Diego had the highest percentage of Hispanic donors.
Individual sites reported wide ranges in the distributions of their units by race and ethnicity. For example, the Karmanos Cancer Institute/JP McCarthy Cord Blood Bank in Detroit is a smaller, minority-focused bank that stores about 442 cord blood units; 81 percent of these units are from African-American donors, and less than 10 percent are from Caucasian donors. Duke University provided information on 3,870 banked units; 59 percent of these are from Caucasian donors, 19 percent are from African-American donors, and 9 percent are from Hispanic donors. Fifty-six units from this bank have been transplanted, but a disproportionately high percentage (14 percent) of units from Hispanic donors were chosen for transplantation.
One of the most challenging aspects of cord blood acquisition is the selection of an appropriate unit for transplantation. The lack of an agreed-upon search algorithm creates a challenge for transplant physicians searching for treatment options for their patients. At present, physicians must search several unrelated databases to identify all units that might be compatible. They must also compare these results of these searches with information on potential donors in adult bone marrow registries to see whether a suitable marrow match is available. Figures 4-3 and 4-4 show outlines of the typical decision-making procedures that a physician must perform when he or she is searching for HPCs for transplantation. Multiple factors, such as cell dose requirements, HLA match requirements, and the geography of the transplant center are important in the search process.
An ideal search algorithm would encompass a scalable system that is capable of searching every available cord blood unit banked by all accred-