AbstractAbstract 1524▪The websites of the WMDA and BMDW list almost 100 registries providing volunteer unrelated donor (VUD) stem cells to the transplant community. A minority of these registries have completed a formal accreditation process although the majority of donors (>80%) in BMDW are contained within WMDA accredited registries. As a user of these registry services we were interested to assess variations in the efficiency of services in order to optimise our own search process and maximise our ability to transplant patients who might benefit in a timely manner. Within a national health service we are careful to contain costs and are aware that not only does the level of HLA typing of individual donors differ between registries but the costs of, and time to, obtaining confirmatory samples and the costs of eventual stem cell procurement also vary considerably. Since January 2004 we have maintained a comprehensive database to follow the progress of VUD searches in our institution. During this time searches were commenced for 193 patients (139 Caucasian, 15 African-Caribbean, 24 Asian and 15 mixed-race), and donors identified for 149 (77%). After high resolution typing of the potential recipient our initial request goes to our national hub which then provides a list of possible donors worldwide, including known degree of match and registry source. The simple parameter of time to donor identification is heavily influenced by the transplant centre which often categorises recipients in degrees of urgency and was not thought to be a useful measure of efficiency. Instead we looked at the availability of, time to obtain, and value of (as defined by degree of matching), confirmatory tissue typing samples once the request had been initiated, in addition to normal measures of the quality of the received stem cell product. For the 193 patients we requested further samples from 1226 donors with a median number of requests per patient of 7 (range 1–20). We utilised 30 registries (median number of requests per registry (R/R) of 3, range 1–387), although some 50% of the requests went to only 5 registries with a median of 220 R/R, range 43–387. 806 confirmatory samples were received from 1226 requests (66%). Donor availability and the time to sample arrival were compared in the 5 most frequently utilised registries, hereafter known as registries A-E. The percentage of the requests that resulted in samples and the percentages of samples found to be HLA-matched were 76%, 72%, 70%, 67% and 53% and 31%, 41%, 40%, 46% and 28% from registries A,B,C,D and E respectively. The chance of obtaining a suitable HLA-matched donor was highest from registry D at 31% and lowest from registry E at 15%. Failure to supply a confirmatory sample is due to inability to contact the donor, temporary unavailability or donor refusal and was unacceptably high in registry E at 47%. The median times from request to sample arrival ranged from 15–22 days. We confirmed the difficulties in finding donors for non-Caucasian patients with fewer potential donors provided initially by our hub and a reduced chance of finding fully HLA matched donors (defined generously as 8/8 antigens matched). Such donors were found for 62%, 26%, 46% and 40% of Caucasian, African-Caribbean, Asian and mixed-race patients. 61 patients have been transplanted to date. No differences were found between the 5 registries in the hours to infusion, viability at infusion, volume collected, total nucleated cell counts, CD34+cell numbers or time to engraftment. Of the 88 patients for whom donors were found but no transplant has been performed, 72 remain alive on treatment. Many of these have chronic myeloid leukemia (CML) and are responding to second generation tyrosine kinase inhibitors. The costs of obtaining samples varies from approximately $500-2000 per sample resulting in a range of 500 to several thousand dollars per patient. Cost cannot be recouped in our country if the transplant is not performed. As a result of this audit we have refined our search process using the initial search as a surrogate for the ease of finding a donor subsequently, so that full searches are only initiated in patients failing imatinib if they have few donors and/or donors with low resolution typing and focussing on those registries most likely to provide HLA-matched blood samples. Registries also have a responsibility to improve their service and minimise costs by focussing on donor retention and high resolution typing. Disclosures:No relevant conflicts of interest to declare.