Wrapping our arms around the cost of transfusion therapy.

Health care costs are anticipated to cause ever-increasing financial burdens on patients, families, communities, and employers, as well as state and federal governments. In most developed countries, health care spending increases at a more rapid rate than the gross domestic product (GDP). In the United States, the health spending-to-GDP ratio has reached 17.9%. The Center for Medicare and Medicaid Services estimates that this ratio will reach 19.6% by the year 2021.1, 2 There is great potential for a devastating impact on the overall US economy if this rate of growth in health spending continues unabated. Because of the “cruel financial realities” that must be faced, it is incumbent upon those involved in health care to critically evaluate the suitability and cost-effectiveness of current health care measures and, based on this analysis, make informed decisions regarding whether equal or higher-quality care can be more efficiently provided at lower overall costs. Not surprisingly, the costs of blood transfusion are also on the rise including significant increases in recent decades.3, 4 Transfusion therapy, therefore, is deserving of a comprehensive cost-effectiveness evaluation. The transfusion of allogeneic blood is one of the most common procedures in medical and surgical practice. According to the World Health Organization's Global Database on Blood Safety, 92 million units of blood were donated in 2011.5 In a 2009 survey (based on data from 2008), the mean acquisition cost of a red blood cell (RBC) unit for US hospitals was $223.09.6 However, the costs of running blood banks and transfusion services are much higher than simply the blood acquisition costs. These costs include all resources (human and nonhuman) involved in the ordering of pretransfusion testing, pretransfusion sample collection and transport, pretransfusion testing, cross-match procedures, issuing of blood components, transport of blood components to the bedside, administration and monitoring of transfusion episodes, costs associated with adverse events, and final documentation of transfusions in the medical record. Additional costs necessary to maintain blood banks and transfusion services include supply chain management, information technology, cost accounting and billing, human resource management, cleaning, waste management, quality management, accreditation and regulatory compliance, education and training, and operations management. To accurately and completely calculate the cost of transfusions, direct and indirect costs associated with the process must be incorporated into the equation. Activity-based cost analysis is a technique that can be used to calculate the cost of complex processes such as the true cost of transfusions.7, 8 This approach was utilized to determine the total cost of transfusion of surgical patients in four hospitals in Europe and the United States.9 In the United States, the 2007 total cost of transfusion for a single RBC unit in a surgical patient was calculated to be between $726 and $1183 (data from the two US facilities). According to the National Blood Collection and Utilization Survey Report, 14.9 million RBC units were transfused in the United States in 2009.6 If the mean value for the activity-based cost of an RBC transfusion from the US facilities is used for calculation ($954.50), then the annual cost of RBC transfusions in the United States exceeds $14.2 billion. The underlying assumption of this calculation is, however, that all transfused populations are comparable to the surgical patient populations used in the activity-based cost analysis. This assumption, of course, is not true. The costs associated with the care of patients who have the wide array of disease processes that require transfusion support are much more variable. In 2001, in an attempt to better characterize the variability of costs in various patient populations, Jeffries and colleagues10 published the results of an analysis of transfusion costs associated with individual medical and surgical diagnosis-related groups (DRGs). DRG transfusion cost data from 1995, obtained from 60 university hospital members of the University Health Systems Consortium, was used for the analysis. When all medical and surgical DRGs were considered in aggregate, blood transfusion costs accounted for approximately 1% of total hospital costs. In some patient populations, however, blood costs accounted for notably higher proportions of total hospital costs. For example, transfusion costs accounted for 5, 7.1, and 8.7% of total hospital costs in liver transplant, bone marrow transplant (BMT), and adult leukemia patients, respectively. These three DRGs not only had the highest transfusion costs as a percentage of the total costs of hospitalization, but also ranked 1 through 3 for the highest actual median and mean transfusion costs, with BMT having the highest median and mean transfusion costs ($4444 and $6183) and liver transplant having the second highest median and mean transfusion costs ($3888 and $5527). Furthermore, these same three DRGs had the highest cumulative transfusion costs (cumulative costs calculation—total number of hospital discharges with a particular DRG in this 60-hospital study multiplied by median blood cost for that particular DRG), with BMT, for example, having the highest cumulative cost at $8,185,848 (1842 hospital discharges with the BMT DRG × $4444 median blood cost = $8,185,848) and liver transplant in second place with a cumulative cost of $6,018,624. It is noteworthy that not only are patients in these DRGs among the highest users from the perspective of total number of units transfused per hospitalization, but the care of these patient groups often also results in many additional transfusion costs, such as HLA-matched or cross-matched platelet and leukoreduced and/or irradiated blood components as well as other attendant costs such as RBC serology costs associated with RBC transfusions. The authors of this study proposed that information such as the data they provided might be used to focus on transfusion cost reductions targeted at specific DRGs or groups of DRGs (e.g., specific products, components, or processes). In the activity-based cost analysis of RBC transfusions in surgical patients, blood testing processes in the two US facilities accounted for 11.9 and 8.2% of the annual total cost of transfusions, respectively.9 These results suggest that pretransfusion testing accounts for approximately 10% of the annual expenditure on RBC transfusions for surgical patients in the United States (approx. $95.45 per RBC transfusion). Once again, although this information provides important insight into the cost of pretransfusion testing in a specific population (surgical patients), these data may not necessarily be representative of the cost of testing in other patient populations who might differ significantly from the perspective of RBC serology. These sorts of analyses trigger thoughts regarding whether there are opportunities to reduce the cost of serologic testing in the various patient populations. If cost reduction initiatives are to be effectively implemented, there needs to be good cost data to pinpoint those cost reduction efforts. In this issue of TRANSFUSION, Mazonson and colleagues11 describe the results of the Hospital Investigation of Serologic Testing and Results (HI-STAR) study. The HI-STAR study, using data obtained from four large medical centers in the United States, was designed to describe all the testing processes used to identify antibodies and provide compatible RBC units for patients with positive RBC antibody screens. Additional study objectives were to divide these results into patient subgroups based on diagnosis, transfusion history, and serologic test results and, for these patient subgroups, understand the resource use and costs associated with serologic testing. The HI-STAR study provides financial data (from data collected on 6077 antibody-positive serologic work-ups from 3608 patients) that objectively confirms what has been suspected for decades: not all patient populations are “created equal” when it comes to the extent and cost of serologic testing necessary to provide safe transfusion therapy. In addition, the study provides important information regarding the costs of various serologic tests. Patients with autoimmune hemolytic anemia (AIHA) had the highest mean cost per serologic work-up at $505. This high cost appears to be strongly affected by the fact that the two most costly serologic tests were alloadsorption and autoadsorption procedures at $204 and $75, respectively. In contrast, obstetric and neonatal intensive care unit (NICU) patients had the lowest mean costs per serologic work-up at $69 apiece. In 16 of the 19 categories evaluated, the mean cost per serologic work-up was at least $100. With regard to the mean costs of serologic testing per patient, the most costly diagnostic categories were AIHA, hematologic malignancies, and transplant recipients. The mean cost for patients with AIHA was $1490, which was estimated to be a 285% cost increase (95% confidence interval, 225%-356%) related to this diagnosis while the lowest mean cost was for NICU patients at $69. For 15 of the 19 diagnostic categories, the mean total cost per patient for serologic testing was at least $180. Multivariate regression analysis revealed 12 diagnostic categories as significant predictors of higher serologic costs, with AIHA being the strongest predictor of higher costs. In contrast, obstetric care significantly predicted lower serologic costs. By focusing in on the costs of serologic testing of patients with positive antibody screens, the HI-STAR study adds another significant piece to the puzzle in the quest to understand the true cost of transfusion therapy in various patient populations. This study helps to increase the understanding of the widely variable costs of transfusion support in patients with positive serologic test results, which accounts for a substantial number of transfusion candidates in many major medical centers. Detailed cost data, such as has been reported in the HI-STAR study, will assist health care organizations to better identify cost-saving opportunities associated with the provision of transfusion therapy. Good data, such as that provided by the HI-STAR study, will hopefully result in the design and implementation of patient-centered interventions that will provide more efficient, cost-effective transfusion support of patients with an end result of better overall outcomes at reduced costs. In the “world of transfusion medicine,” there needs to be an acknowledgment that a crossroad has been reached with regard to the delivery of transfusion services. Health care organizations are going to be reimbursed less and differently for the services they provide with an increasing number of reimbursements that will be bundled and that will focus on whole processes of care. To maintain financial viability, health care organizations, including transfusion medicine services, will need to develop approaches to patient care that maintain quality of services at lower overall cost. At the organization where I work, all employees have been challenged to “think differently” and “practice differently” to more cost-effectively maintain the highest standard of health care. For specialists in transfusion medicine, thinking and practicing differently includes devising approaches to reduce the overall costs of transfusion therapy. Thinking and practicing differently in transfusion medicine can take many forms. For example, positive changes may result from new or innovative uses of technology, major practice paradigm shifts, or as a result of identified process improvements. An example of where the innovative use of technology might enable people working in transfusion medicine to practice differently is through further applications of molecular techniques in RBC serologic practices. Such techniques are currently being advantageously used by many centers; however, it is possible that they could play even more important roles in the practice of immunohematology in the future.12 Although the role of molecular techniques in mainstream practice is still being defined and is still very much a “work in progress,” it does hold the potential for minimizing RBC alloimmunization up front (decreasing the overall costs of transfusion care by decreasing the number of complex serologic transfusion patients) as well as being a viable, alternative method of identifying RBC units for safe transfusions in alloimmunized patients (matching donors and recipient extended RBC genotypes) in the absence of repeated expensive serologic evaluations (thereby potentially resulting in cost savings over the long run). Patient blood management (PBM) is poised to assume a widespread and significant role in transfusion medicine, resulting in major paradigm shifts in transfusion practices.13, 14 PBM, which promises to provide patient outcome benefits and significant health care savings, has recently received endorsement by the World Health Assembly as documented in resolution WHA63.12,15 as well as receiving endorsement by the HHS Advisory Committee on Blood Safety and Availability at its June 7 to 8, 2011, meeting. Evidence continues to mount regarding the cost-effectiveness of PBM programs.16 Our institution implemented PBM in 2009, initiating the program in patients undergoing cardiovascular surgery. In addition to seeing better overall outcomes in cardiovascular patients after PBM implementation, the initiative has reduced acquisition costs for blood products by more than $12 million.17 Cost-effective transfusion support can also be achieved through process improvement initiatives. To expedite the flow of patients through “the system” with a minimum of call backs and other patient inconveniences, the practice in hematology patients at our institution has evolved to concurrent ordering of a type and screen (T&S) with a complete blood count (CBC). Although this is a patient-centered approach, there was concern in the Division of Transfusion Medicine that such an approach was resulting in a large and costly volume of unnecessary T&S procedures in hematology patients. A retrospective analysis of 3410 hematology patients over a 6-month period revealed an overall ratio of T&S procedures performed to patients subsequently receiving RBC transfusions of 4.11. Thus, the majority of T&S procedures performed in hematology patients were not subsequently being followed by RBC transfusions. Based on these data, an alternative approach to the serologic management of hematology patients was proposed. It was proposed that a reflex testing scheme be devised that could be triggered by the CBC result in hematology patients. If a hematology patient's CBC revealed a hemoglobin (Hb) value of less than 8.0 g/dL, a reflex testing scheme would be triggered that resulted in the collection of a blood specimen and the performance of a T&S. When the CBC revealed that the patient's Hb was 8.0 g/dL or higher, the reflex T&S would not be performed. The implementation of the reflex T&S approach was projected to decrease the number of T&S procedures performed annually on hematology patients from a preimplementation volume of 10,949 tests to 1812 tests after implementation. This reflex T&S approach will result in significant cost savings and, in addition, result in an annual 76,642-mL decrease in blood volume collected from hematology patients. In conclusion, studies such as the HI-STAR study help those involved in the practice of transfusion medicine to better understand the total cost and variability of costs associated with transfusion therapy. A better understanding of the costs of transfusion therapy allows for more accurate and efficient approaches to reduce the costs of such therapy. These approaches can take many forms. Some of these approaches will require large-scale initiatives while other effective approaches might be relatively small scale. A handful of examples have been provided in this editorial. Once transfusion medicine practitioners get “in tune” with the process of thinking and practicing differently, I am confident that many additional cost-containment strategies will be developed and all of us will be the beneficiaries of this creative thinking. The authors report no conflicts of interest or funding sources.