US20080140452A1 - Transfusion registry network for genetically characterized blood products - Google Patents

Abstract

Disclosed is a registry system, including member institutions, in which transfusion donors and recipients are registered following genotyping, which would typically take place in a member institution, or a member institution would have access to the genotyping information, if performed outside. The registry database can be accessed and searched by members seeking samples of particular type(s). Systems are disclosed for maintaining economic viability of genotyping in connection with transfusions, by maximizing the number of units placed with the minimal number of candidate donors typed. Genotyping of potential donors, and product supply, is matched to forecasted demand. Genotyping can also be limited to the more clinically relevant markers. The registry system can also be integrated with one format of assay which generates an image for analysis, whereby the imaged results can be analyzed and redacted by experts in a central location, and then transmitted back to the patient or their representative.

Description

RELATED APPLICATIONS
• This application claims priority to U.S. application Ser. No. 11/092,420, filed Mar. 29, 2005, which claims priority to U.S. Provisional Application Ser. Nos. 60/586,931 and 60/621,196. Priority to all these applications is hereby claimed.
BACKGROUND
• The prevailing paradigm of organizing the supply of blood units available for transfusion relies on routine typing of transfusion antigens by hemagglutination. Typically, the major transfusion antigen groups, namely A, B, O and D, are typed at collection while a select set of minor group antigens such as RhCE, Kell and Kidd are typed only as needed For blood group antigens other than ABO and D, source material is diminishing, and the cost of FDA-approved commercial reagents is escalating. Many antibodies used for testing for minor blood group antigens (especially when searching for an absence of a high prevalence antigen) are not FDA-approved and are characterized to varying degrees by those who use them. In addition, some antibodies are limited in volume, weakly reactive, or not available. Collectively, the labor-intensive approach limits the number of donors one can test: thereby restricting the supply of antigen-negative RBC products for patients who have produced the corresponding alloantibody, and, more recently, restricting the supply of Rh and K matched RBCs for patients in the Stroke Prevention Trial (STOP) program, which was designed to prevent immunization of such patients.
• Recipients exposed to foreign transfusion antigens generally will form antibodies directed against those antigens. Allo-immunized patients, a subpopulation comprising approximately 2% of transfused patients, and up to 38% of multiply transfused patients, require red blood cell products which do not contain the offending antigen. Such units typically must be found either in the limited available supply or must be found, in real time, by serological typing of such likely candidate units as may be available in inventory. The selection of candidate units for “stat” typing, performed in immunohematology laboratories, is guided largely by empirical factors. The delay introduced by the search for matching units can exacerbate emergency situations and generally will incur substantial cost to hospitals and/or insurance carriers by delaying in-hospital stay. More generally, allo-immunization to red blood cell antigens which are also displayed on other cells (see Table I) and recognized by certain pathogens such as malaria, can introduce unnecessary health risks whose elimination would improve the general health.
• The procurement of matched blood to recipients who either display an uncommon antigen or lack a common antigen, is particularly problematic. While such incidences are considered “rare,” occurring at a rate of one in 1,000 recipients, the supply of matched units is very limited. Thus, existing national collections of special units, including the American Rare Donor Program (ARDP), register donors encountered in the immunohematology laboratories of its members: only 30,000 donors have been registered (see, e.g., the Red Cross Website). In comparison, the National Maarow Donor Program (NMDP), a national registry of prospective bone marrow donors who have been genotyped for polymorphisms in certain loci of the Human Leukocyte Antigen (HLA) gene complex, in the year 2000, comprised 2.7 million fully characterized and 4.1 million known donors to supply matching bone marrow transplants for only ˜2,400 transplantations per year. See the National Marrow Donor Program Website.
• Distribution of the precious few special units available in the program also leaves substantial room for improvement. At present, relying primarily on telephone contacts, only 1,000 special units are placed per year, while up to 2% of the approximately 4-5 million recipients of blood transfusions per year, that is 100,000 recipients, would benefit from improved availability.
• In view of this situation, a method of providing a large and diverse inventory of fully typed blood units, and a method of instant and efficient distribution of units in response to requests posted to a central registry would be desirable in order to improve the public health and to minimize the cost accruing in the health care system in the form of unnecessarily prolonged hospital stays, adverse transfusion reactions (see Hillyer et al., Blood Banking and Transfusion Medicine; published. by Churchill Livingston, Philadelphia Pa.) and other potential complications arising from allo-immunization.
• However, absent substantial government or private funding for such an endeavor, a registry of “critical mass” must be created and operated in a commercially viable manner. The ARDP operation, representative of current practice, illustrates the difficulty: In order to identify a special donor, up to 1,000 donors may have been typed, and from a collection of 30,000 such special units, only 1,000 were placed. While special units fetch a higher price than do “vanilla” red blood cell products, the premium does not come close to covering the cost, in view of the substantial amount of excess typing required. Commercial viability, under these conditions, is doubtful.
SUMMARY
• Described is the efficient organization and operation of a diverse registry of fully characterized blood units. Preferably, donors are characterized by DNA typing of the clinically most relevant genetic markers, including a set of mutations of Human Erythrocyte Antigens (HEA) including genetic variants of Rh, and additional antigens such as HLA and HPA. The registry, also referred to as a Transfusion Network, comprising certain application programs and databases preferably accessible via a web-browser interface, offers essentially instant access to linked inventories of typed units of donor blood (“actual” units) as well as access to genotyped donors who are available on-call (“callable” units), along with requisite information relating to donor status, inventories of actual units or information relating to callable units ran be held by subscribing member organizations, who also may participate in the operation and governance of the registry.
• In a preferred embodiment, the registry network comprises an alliance of dominant regional and national donor centers (such as New York Blood Center and United Blood Services) which would set new standards in transfusion medicine. In another embodiment, regional donor centers and transfusion services are linked so as to create the critical mass of regional centers (both domestic and foreign) to decentralize the market by competing with the dominant national donor centers.
• An “actively managed” registry—Existing registries such a the ARDP largely operate as passive repositories of donors encountered per chance during blood drives. Registries of bone marrow donors operated by the NMDP or comparable organizations around the world (REF), while in some cases actively funding bone marrow drives, operate in essentially the same manner of underwriting the large-scale typing of volunteer donors and collecting results. TO the extent that the population of donors and population of recipients are not balanced, this approach generally will be very inefficient from the point of view of maximizing the probability of a matching a recipient request.
• To overcome this inefficiency and to ensure commercial viability, a preferred strategy is described herein for constructing and maintaining a registry of genotyped donors which maximizes the number of units placed with the minimal number of candidate donors typed. To this end, relevant parameters relating to managing supply and forecasting demand are identified, and methods are described to optimize these parameters so as to maximize revenue and minimize total cost. The registry performs real-time analysis of supply and demand balance and directs its subscribing members to balance their respective donor typing operations.
• A transfusion network, operated as an active registry, permits near-instant selection of prospective donors matching a given recipient by way of implementation on a global network such as the world wide web, thereby also facilitating the efficient distribution of units in inventory, further supported by transaction management including order placement and delivery. The registry will generate revenue from subscription as well as transaction fees, offering a set of products and services as described herein. Thus, a commercially viable registry, the first such in transfusion medicine, is disclosed, to improve clinical outcomes while enhancing economic efficiency.
• In one embodiment, large-scale, rapid and cost-effective DNA typing, also herein referred as genotyping, of prospective donors is performed to permit instant matching of registered donors to recipients of known phenotype or genotype in a manner improving the clinical outcome of transfusion while improving economic efficiencies. To the extent that genotyped donors are retained, the cost of typing is minimized, as discussed herein.
• The registry server preferably executes a “genetic cross-matching, gXM” algorithm to identify actual and callable donors within the registry, A gXM algorithm relating to a selection of the clinically most relevant human erythrocyte antigen (HEA) mutations is described in a co-pending application (see Provisional Application No. 60/621,196, entitled “A Method of Genetic Cross-Matching of Transfusion Recipients to Registered Donors,” as well as applications to be filed claiming priority to it, all of which are incorporated by reference).
DESCRIPTION OF THE FIGURES
• FIG. 1 is an illustration of a transfusion registry network linking multiple donor centers offering blood-derived products to multiple hospitals requesting blood-derived products for transfusion to patients. Participating parties can perform donor genotyping, patient genotyping and patient antibody screening (using, e.g., BeadChip™ assay kits). The registry performs genetic cross-matching and can offer a variety of additional products and services.
• FIGS. 2A to 2D is a graphical depiction of the central components and subsystems of a BeadChip™ format for multiplexed analysis of polymorphisms and profiling of antibodies enabling the large-scale genotyping of donors and patients, as well as detection and identification of antibodies circulating in patient serum.
• FIG. 3 is a graphical depiction of a uniform interface for presentation of data to the registry, preferably by way of a web-enabled Automated Allele Analysis (AAA) program (as disclosed in U.S. application Ser. No. 10/909,638, incorporated by reference), and the connection of the registry to linked inventories
• FIG. 4 is a graph showing the dependence of cost and revenue projections for a multiple donation scenario for various values of the repeat probability, RHO.
• FIG. 5 is an illustration of the concept of optimal resolution (described in text).
DETAILED DESCRIPTION
• In order to maximize the economic efficiency of the transfusion registry, it will be preferable to adopt a strategy of minimizing the total number of donors typed for every recipient request fulfilled. The following exposition refers to a genotype to represent a combination of marker alleles, where, for each marker, the possible values of the allele are Normal (1), Homozygous (−1) or Heterozygous (0), and a specific genotype, representing a combination of alleles, thus has the form of a ternary string.
• Estimating Demand Requests for Special Units—In order to maintain a registry of candidate donors such that the maximal number of requests from prospective recipients for special units can in fact be matched while the number of excess donors typed is kept to a minimum, it will be critical to construct an estimate of anticipated demand.
Denote by:
• • N^Rthe number of requests anticipated (or received);
  • 8 the probability of receiving (logging) a request for a specific genotype;
  • : the probability of matching a request (to a pre-determined level of resolution)
    Available evidence indicates that the incidence of certain genotypes varies substantially between ethnic groups (see G. Hashmi et al., “A Flexible Array Format for Large-scale, Rapid Blood Group DNA Typing,”Transfusion, in press). Therefore, the probability of a request for blood from a donor of specific genotype received from a random sample of a heterogeneous pan-ethnic population in fact represents a weighted average of probabilities, 8_s, for each of multiple constituent homogeneous subpopulations. The population-specific probabilities may be cast in the form:
• 
  8_s˜(N^(Rs)/N^(R))f^(s)Σ(r)
• where f^(s)represents the frequency of occurrence of a certain allele, the ratio (N^(Rs)/N^(R)) represents the relative proportion of individuals in subpopulation s within the pan-ethnic population at large, and Σ(r) represents a function of excess risk associated with a specific subpopulation (relative to the population at large). The function Σ(r), which may assume positive or negative values, reflects actuarial probabilities which in turn reflect genetic risk, e.g., the higher than average incidence of sickle cell anemia in African-Americans, or higher than average incidence of kidney disease in certain native American Indian tribes, requiring multiple transfusions, and environmental risk, e.g., the lower probability of, e.g., the Amish to suffer trauma in automobile accidents.
• The probability, :, of matching a specific request depends on the diversity of the registry and its linked inventories of actual and callable donors.
• Managing Supply Selection of Donors from Stratified Populations—in accordance with the preferred strategy of registry operation, the supply of registered donors will be adjusted to balance the anticipated demand.
• Denote by:
• • N the number of new donors tested;
  • , the fraction of special units encountered in a test population; 0≦, <1;
  • Φ the fraction of special units sold.
• The probability, Φ, of selling any specific unit is determined, for given unit price, by the probability. :, of matching a request for such a unit. Provided that an acceptable price for a unit is agreed upon, then:
• 
  Φ=:
• Preferably, the strategy for balancing the supply of registered donors will reflect the increased probability of finding an acceptable match for a prospective recipient of transfusion within a donor population of similar heritage. The similarity of genotype among individuals of similar heritage has been established for a variety of genetic markers such as those for certain inherited genetic disorders, including so-called Ashkenazi Jewish Diseases and Cystic Fibrosis (see the Jewish Virtual Library Website), as well as for the highly variable human leukocyte gene complex which encodes for the human leukocyte antigens (HLA) determining the compatibility of recipients and donors of solid organs and bone marrow through the National Marrow Donor Program website. For blood group genotypes, but one example is provided by the high incidence in individuals of South Chinese heritage of the Miltenberger mutation within the MNS blood group (see M. Reid, “The Blood Group Antigen FactsBook” (2003)) which is largely absent in individuals of Caucasian heritage.
• As with demand estimation, the probability of encountering a specific genotype in a pan-ethnic and hence genetically heterogeneous donor population will reflect the existence of constituent homogeneous subpopulations displaying varying values of that probability:
• 
  ,=(N⁽¹⁾/N)f⁽¹⁾+(N⁽²⁾/N)f⁽²⁾+ . . . +(N^(s)/N)f^(s)
• where, as before, f⁽¹⁾, f⁽²⁾. . . , f^(s)denote allele frequencies. To balance the supply of registered donors to anticipated demand, it will be desirable to select, for each subpopulation, s, shared among donor and recipient populations, the number of registered donors in accordance with the condition:
• 
  (N^(s)/N)=C(N^(Rs)/N^(R))Σ(r)
• The constant, C, captures factors such as the anticipated number of units required per recipient. This condition dictates that the registry, rather then genotyping alt corners, would accept only a certain continent of donors from each subpopulation.
• Factors determining Profitability—A key aspect of operating a transfusion registry network with an acceptable profit margin concerns the pricing for a test permitting the genotyping a donor sample for a designated number of genetic markers, preferably by invoking elongation-mediated Multiplexed analysis of polymorphisms (“eMAP”; as disclosed in U.S. application Ser. No. 10/271,602, incorporated by reference).
Denote by:
• N_kthe number of new donors tested in year k, where k=0, 1, 2, . . . , n;
• R_kthe number of repeat donors (from year k−1) in year k=1, 2 . . . n;
• Δ the fraction of repeat donors; generally R_k<N_k, and thus Δ<1.
• Δ_Sthe fraction of repeat donors among special donors; Δ_S<1.
• , the fraction of special units encountered in a test population; 0≦, <1;
• c the cost of typing one sample;
• Φ the fraction of special units sold;
• s the excess revenue (over the “vanilla” unit) of a special unit of product.
• The cost of screening in year k is: C_k=cN_k−g(R_k, R_k-1, . . . ), that is, in any year but the first (k=0), the total cost of typing N_kdonor samples will be reduced by a certain portion reflecting the number of repeat donors from previous years. Various assumptions—manifesting themselves in specific forms of the function g(R_k, R_k-1, . . . )—are possible. To the cost of typing must be added the cost of operating the registry—including transaction costs.
• The revenue in year k reflects the sale of special units accumulated in inventory, that is: S_k=h(N_k, N_k-1, . . . ). Various assumptions—manifesting themselves in specific forms of the function h(N_k, N_k-1, . . . )—are possible.
• The profit in year k is given by P_k=S_k−C_k. Break-even, P_k=0, is attained at a certain k.
EXAMPLE 1Single Repeat Donations
• Assume that a certain constant total number of donors, say N₀, is screened every year, and that a (constant) fraction of donors repeat, but repeat only once, namely in the year following their initial donation. Then R_k=ΔN_k-1, and:

• 	Yr0	Yr1	Yr2

  	New donors	N0	N1= N0− R1	N2= N0− R2
  	Repeat donors		R1= ΔN0	R2= ΔN1

  
  where R₂=ΔN₁=Δ(N₀−R₁)=ΔN₀−Δ²N₀and N₂=N₀−ΔN₀+Δ²N₀=N₀(1−Δ+Δ²).
  Generalizing, one finds the expression for N_kto be N_k=N₀(1−Δ+Δ²−Δ³+ . . . ); the alternating series reflects the fact that, as repeat donors stay away, a greater number of new donors must be screened in every even year. For n sufficiently large so that Δ²ⁿ<<1, this expression turns out to be N_k=N₀/(1+Δ), independent of n; for example, with Δ=½, Δ²ⁿ=(½)²ⁿ= 1/256 for n=4.
• Assume further that revenue in any given year reflects the sale of a certain fraction, Φ, of the total units, , N₀, available that year, at an excess sales price, s, per sample, and that the population of repeat donors within the special population equals that within the general population. Then

• 	Yr0	Yr1	Yr2
  	S0= Φ(,N0)s	S1= Φ(,N0)s	S2= Φ(,N0)s

  
  Then P_n=S_n−C_n=[,Φs−c/(1+Δ)]N₀, independent of n. Under these assumptions, to attain break-even, P_n=0, or c/s=(1+Δ),Φ, the cost per unit screened must not exceed a certain fraction of the excess revenue in each year.
• For example, with reported numbers of Δ=½ (see Schreiber, G. B. et al., “Targeting Repeat Blood Donors Can Increase Supply,”Transfusion43: 591-97 (2003)),=0.001 (percentage of “rare” units in pan-ethnic population) and Φ=⅕, reflecting the placement of 1,000 “rare” units (from a stock of 30,000), with approximately 5,000 new rare units acquired per year (See the National Marrow Donor Program website), one obtains c/s= 3/2*0.001*⅕=0.0003. Since s will likely not exceed $1,000, the price per test will have to be negligibly small, not a scenario for a profitable large-scale screening operation. In fact, this is near the worst case scenario in which, along with the low abundance of special samples, and low percentage of placement, donors do not repeat (Δ=0).
• It is anticipated that proper demand projection and inventory management, combined with providing instant access to such inventories by way of a transfusion registry network, as disclosed herein, will provide a basis to attain an operating regime of ,→0.1 and Φ→1 so that, even with Δ=½, c/s= 3/2*0.1*1=0.15.
EXAMPLE 2Multiple Repeat Donations
• In contrast to the previous Example 1, assume that of the total number of donors, say N₀, screened in the first year, a (constant) fraction of donors, once recruited, repeat every year. Given that each donor is genotyped only once, this will have a cumulative effect on cost reduction.
• To illustrate the effect, assume first that the same fraction of general donors and special donors repeat, that is: Δ=Δ_S, and that this fraction is constant.

• 	Yr0	Yr1	Yr2

  Cost

  	C0= cN0	C1	= c(N0− R1)	C2	= c(N0− R2− R1)
  			= cN0(1 − Δ)		= cN0(1 − Δ)2
  	Revenue
  	S0= Φ(,N0)s	S1	= Φ(,N0)s	S2	= Φ(,N0)s

  
  Then P_n=S_n−C_n=[,Φs−c(1−Δ)ⁿ]N₀. The contrast to the model of Example 1 is dramatic: the requisite cost of typing required to attain the same revenue, decreases geometrically with n, the slope of the decrease being set by Δ. Break-even corresponds to c/s=,Φ/(1−Δ)ⁿ, and profit grows rapidly thereafter.
• FIG. 4 illustrates the evolution of projected cost and excess revenue for different values of the repeat probability, Δ. Precedents for relatively high repeat probabilities exist, especially in donors who are aware of their special status. It will be desirable to provide incentives to such donors, as described below.
• Mutually Beneficial Interaction of Registry and Reagent Manufacturer—Provided that the large-scale genotyping of donors and patients is enabled by an efficient methodology, preferably invoking the eMAP-HEA design in conjunction with the BeadChip™ format (US Provisional Application Ser. No. 60/586,931, entitled “Encoded Probe Pairs for Molecular Blood Group Antigen Molecular Typing and Identification of New Alleles” and applications claiming priority thereto (incorporated by reference); G. Hashmi et al., “A Flexible Array Format for Large-scale, Rapid Blood Group DNA Typing,” Transfusion, in press; see also:FIG. 2), the cost of genotyping is reduced by the use of the multiplexed format of analysis and delivery of the assay in a parallel processing format, thereby permitting automation and uniform data management for a large menu of applications, including the typing of multiple antigen groups (FIG. 3).
• In the initial stage, while building its initial donor reservoir, the registry, to the extent that it bears the cost of recruiting and genotyping donors, either directly, or indirectly, by way of subscribing member donor centers, will operate at a loss (FIG. 4). It will be beneficial for the registry to partner with a reagent manufacturer who would underwrite the operations in the initial stage, for example by providing kits at reduced or at no cost to the registry. To the extent that the registry is successful in retaining special donors and hence reduces its cost of typing, the market opportunity for the reagent manufacturer declines. To compensate the reagent manufacturer, the registry could, for example, grant the manufacturer participation in a jointly controlled entity along with a profit sharing arrangement.
EXAMPLE 3Expanding Fire Registry
• The scenario of Example 2 offers several modes of operation. For example, the registry might operate in a “non-profit” regime by setting the ratio c_n/s_nso as to ensure P_n=0. That is, the diagnostic reagent manufacturer, in return for obtaining a designated share of profits after break-even, can subsidize the initial ramp-up of the registry by accepting a lower price per test, corresponding to the condition P_n(c_n/s_n)=0.
• Alternatively, the registry, having attained breakeven, may decide to expand operations by expanding the number of donors screened per year, for example, such that the number of new donors is set by the available profit. This provides a mechanism to compensate the participating reagent manufacturer for the declining sales of tests arising from the successful retention of repeat donors.
• Recruiting and Retaining Special Donors—The single repeat and multiple repeat scenarios indicate the critical role of the effect of the repeat probability on the profitability of the registry.
• As with the recruitment of HLA donors for national registries, genotyping of blood group antigens will permit the identification of prospective future donors—that is, donors who do not have to donate blood until called upon. For example, analysis of DNA extracted from buccal swabs would enable “self-collection” in targeted communities such as churches and synagogues, and simple submission processes, e.g. by mail, to a designated member laboratory (not necessarily a donor center), for DNA analysis. This aspect not only allows the extension of the universe of known special donors, but also would be invaluable in registry management, in order to match the volume of donor typing to projected demand within individual subpopulations.
• To refine this model toward a “best case” scenario, retention efforts would be directed to special donors, not the general donor population. The total number of special donors in each designated subpopulation would be matched to demand projections, as described above. Special donors would be given incentive to repeat by granting them, and donating family members, authorized direct emergency access to the registry.
• Clinical Benefit vs. Cost of Genotyping “Optimal” Panel Size—The clinical outcome of transfusion generally would improve with increasing resolution, that is, with the number of genetic markers included in the determination of patient and donor genotypes. The greatest benefit would derive from matching alleles encoding the clinically most significant blood group antigens (see M. Reid “The Blood Group Antigen FactsBook” (2003)), and the incremental benefit of matching additional alleles would decrease. Ignoring cost, a reasonable criterion for the determining the optimal resolution would be to select this point of diminishing incremental benefit.
• However, significant economic considerations also apply. Thus, the higher the degree of resolution required for genetic cross-matching of a donor genotype to that of a patient, the higher the risk to the registry of not being able to place that unit, and the higher the cost of typing that unit. That is, c, the cost of genotyping, generally will increase with the number, m, of genetic markers included in the set, while the probability, , of matching a request and selling a specific unit will decrease. For example, denoting by f_First, f_Second, f_Third, . . . f_Lastthe relative allele frequencies of markers included in cross-matching, in the order of decreasing clinical significance, :˜f_First×f_Second×f_{Third, . . . ×fLast}, thus :˜1/m^x, where x denotes an exponent, while c˜c₀+a*m^y, where c₀denotes a constant, namely the initial cost of genotyping the first marker, a denotes a constant related to the marginal cost of genotyping additional markers, and y denotes an exponent reflecting the rate of increase in cost: for example, using a single-marker method of genotyping such as Restriction Fragment Length Polymorphism (RFLP) or Allele-specific PCR, one would anticipate cost to increase linearly, y=1, while using the preferred embodiment of elongation-mediated multiplexed analysis of polymorphisms (eMAP), one would anticipate cost to increase in a sublinear form, y<1. In either case, from the cost-benefit point of view, there exists an “optimal” resolution. Unless the market would compensate the registry for the higher cost of matching a donor unit to high resolution, in which case clinical benefit will set the optimal resolution, m*, that value otherwise will be determined by the intersection of the two functions :=:(m) and c=c(m), as illustrated inFIG. 5.
• Implementation of Registry—A co-pending application (U.S. application Ser. No. 10/909,638, incorporated by reference) discloses algorithms and implementations for automated allele analysis, and these methods are useful in connection with genetic cross-matching (gXM) generally, as disclosed in a further co-pending application (Provisional Application No. 60/621,196, noted above, incorporated by reference).
• Using standard software engineering technologies such as MicroSoft.net (“dot-net”), these methods can be implemented in a manner permitting their use in an application-server modality using a standard web browser such as Microsoft Explorer™. Preferably, such an implementation, wAAA™, will invoke an SQL server and provide a uniform interface to Array Imaging Systems generating data for a variety of applications such as multiplex HEA, HLA and HPA analysis as well as patient and donor antibody identification. As disclosed in a Co-pending application (Ser. No. 10/714,203), data records will be uploaded—preferably using transaction protocols preserving donor and patient anonymity—to the wAAA application on a central server permitting review and redaction by, and delivery to authorized users.
• Establishing an Efficient Market: Management of Real-time Transactions—Disclosed is a mode of operating a commercially viable registry in the form a real-time transaction network offering instant access to a diverse collection of characterized donors, that is, both actual donor-derived blood products in linked inventories, and callable candidate donors of desirable genotype. Such a registry will increase demand by extending its reach to a global base of potential customers by providing access to it products and services by way of a standard web browser and permitting applications, notably automated allele analysis, gXM and selection of candidate donors to be performed automatically, in real time. By offering instant transactions, under a variety of pricing arrangements including contracts, notably futures contracts, as well as real-time pricing, for example by way of bid-ask matching (currently available in the context of web-auctions as well as Electronic Communications Networks (ECNs) such as InstiNet (http://www.island.com), the registry will create an efficient market for the global procurement and distribution of matched donor units.
• Collection of Samples, Assay Performance, Analysis of Assay Results, Patient Counseling and Reporting Results to Patients—Also disclosed herein is a model of implementing molecular diagnostics in a mode of “virtual centralization” which permits the steps of actually performing assays, preferably in a standard and universal format such as the Random Encoded Array Detection (READ™) format, and the steps of analyzing, interpreting and reporting assay results including communicating outcomes to the patient or referring physician, to be performed in different locations, such that experts or groups of experts have access, by way of a standard web browser to the wAAA environment to review data generated in a location different from their own physical location.
• Virtual centralization of the data is accomplished by uploading of data relating to assay results, and interpretation/analysis thereof, to a server or other accessible database. It can then be accessed by or securely transmitted to authorized parties to perform additional interpretation or analysis, or to view or report the results. User identification can be secured at all stages of the process so as to preserve confidentiality, such that, for example, only the patient and perhaps his physician will be aware of the patient identity associated with particular results.
• This model is particularly well-suited to the analysis and interpretation of results produced by genetic tests, including, results which can be processed for initial analysis by the web-based AAA program (discussed above) or where these results are in the form of images for which standard formats of network transmission now exist. Assay formats producing suitable images invoke, for example, spatially encoded “dot blot” or “reverse dot blot” formats, including “spotted” probe or protein arrays; arrays of oligonucleotide probes synthesized in-situ on a substrate, or probes (or proteins) associated with encoded beads: see U.S. Pat. No. 6,797,524; U.S. application Ser. No. 10/204,799, filed on Aug. 23, 2002 “Multianalyte Molecular Analysis Using Application-Specific Random Particle Arrays,” (incorporated by reference.
• Networking allows the different parties involved in different steps of the process to perform their respective functions such as collecting samples, performing the assay, or analyzing the results, at the same location, or at different locations. Performing separate functions by different parties at different locations can provide a significant advantage in terms of cost and speed of analysis, as parties do not need to travel to a location to carry out their function. It also allows better control over the confidentiality of the results, as results do not need to be physically transported, by non-secure means, to different locations.
• After sample collection, or self-collection by the patient (e.g., by using a buccal swab), an assay on a patient sample can be performed at a first location. The initial assay results, which may be encoded (see U.S. Pat. No. 6,797,524; US Application Ser. No. 10/204,799, filed on Aug. 23, 2002 “Multianalyte Molecular Analysis Using Application-Specific Random Particle Arrays,” both being incorporated herein by reference) or in the form of an assay image (see “Analysis, Secure Access to and Transmission of Array Images,” Ser. No. 10/714,203, filed Nov. 14, 2003, incorporated herein by reference), can be uploaded to a server or transmitted to an analysis site (which may be the site which sold the assay kit to the remote location). The identity of the patient can be associated with the sample, using, e.g., methods set forth in the co-pending application “Genetic Analysis and Authentication,” Ser. No. 10/238,439 (incorporated herein by reference). The analysis site decodes or interprets or performs a preliminary analysis of the results (and/or the assay image), and may also obtain assistance in interpretation from experts or consultants, who may either be on-site or may transmit an image from the assay, or who have access to the server with such images or results. The analyzed results can then be accessed by, or transmitted to, the patient's physician, or the patient, or the laboratory where the assay was conducted, which in turn provides them to the patient's physician and/or the patient. It is possible to keep the patient identity separated from the results at all stages, so that only the physician and the patient, or even only the patient, can correlate results with a particular patient. This secures the confidentiality of assay information, as is desirable in the case of genetic information, given the growing concern over maintaining confidentiality of individual's genetic data.
• In this manner, the “front-end” of laboratory practice is standardized, preferably by adoption of a BeadChip™ format of performing multiplexed nucleic acid and protein analysis while the “back end”, generally requiring specialized expertise, is moved to a network, preferably by implementation of an application service which provides network protocols to transmit assay results, perform analysis, authorize access to databases for review and result certification, and manage communication between multiple participants in the process.
• It should be understood that the terms and expressions herein are exemplary only, and not limiting, and that the invention is defined only in the claims which follow, and includes all equivalents of the claimed subject matter.

Claims (17)

1. A method for operating a registry offering access to genotyped prospective transfusion donors, for matching to a transfusion recipient of given genotype or phenotype, in a manner increasing the probability of having an acceptable donor for any recipient querying the registry, over said probability where donors are not stratified and typed as set forth below, while reducing the total number of prospective donors to be genotyped over the number required to be genotyped where donors are not stratified and typed as set forth below, comprising:

stratifying prospective donors and recipients into subpopulations; and

for each subpopulation, genotyping sufficient numbers of donors of different genotypes or phenotypes to fulfill anticipated demand from recipients of particular genotypes or phenotypes.

2. The method ofclaim 1 wherein blood units are collected from donors of different genotypes or phenotypes to fulfill anticipated demand for blood units from recipients of particular genotypes or phenotypes.

3. The method ofclaim 2 wherein the registry includes several member institutions, and the inventories of the member institutions include genotyped blood units and genotyped donors who agreed to be available to donate.

4. The method ofclaim 1 wherein following genotyping of the number of donors anticipated to be required to match a preset fraction of the anticipated demand from recipients, additional donors are not genotyped.

5. The method ofclaim 1 wherein genetic markers likely to be present in a subpopulation and indicative of a clinically significant adverse event, if mismatched between donor and recipients, are genotyped.

6. The method ofclaim 1 further including providing incentives to retain donors having particular genotypes.

7. The method ofclaim 6 wherein said incentives include extending authorization to said donors having particular genotypes for preferential access, as against other recipients, to the registry in the event said donors require transfusion.

8. A method for reducing clinically significant adverse clinical events in transfusion recipients to less than the number said events encountered if not using the method herein, comprising ranking the genetic markers which are related to clinically significant events in order of their clinical significance, matching donors and recipients for only those markers which are associated with a clinical significance greater than a threshold value, and providing donors with transfusions from matched recipients.

9. The method ofclaim 8 wherein donors and recipients are not genotyped for other markers.

10. A method for reducing clinically significant adverse clinical events in transfusion recipients to less than the number of said events encountered if not using the method herein, comprising:

stratifying prospective donors and recipients into subpopulations;

determining, in each subpopulation, genetic markers which are associated with clinically significant adverse events;

matching donors and recipients for those markers which have a clinical significance greater than a threshold; and

providing donors with transfusions from matched recipients.

11. The method ofclaim 10 wherein donors and recipients are not genotyped for markers with a significance below the threshold.

12-13. (canceled)

14. The method ofclaim 10 wherein the genetic markers include RhCE, Kidd, Kell, Duffy, Dombrock, MNS, and combinations of markers of RhCE, Kidd, Kell, Duffy, Dombrock and MNS.

15-16. (canceled)

17. A method of centralization of molecular diagnostics within a network of member institutions, comprising:

providing to each member institution an assay delivery system capable of uploading assay data to a central server in a standardized format;

providing, through or on said central server, automated analysis and interpretation of uploaded data to generate preliminary assay results;

providing, through or on said central server, access to and expert review of said preliminary assay results, whereby, following said review, final assay results are generated, which are suitable for release to patients or their representatives.

18. The method ofclaim 17 wherein the communication of preliminary and final assay results maintains patient anonymity.

19. A method of increasing the fraction of repeat donations above that experienced when not using the following method by incentivizing selected genotyped donors, wherein said incentives include extending authorization to said donors having particular genotypes for preferential access, as against other recipients, to the registry in the event said donors require transfusion, and providing transfusions to said genotyped donors requiring transfusion.

Images