Stem Cells, Vol. 14, No. 3, 312-319,
May 1996
© 1996 AlphaMed Press
A Review of the Costs, Cost-Effectiveness and Third-Party Charges of Bone Marrow Transplantation
Ilana L. Westermana,
Charles L. Bennetta,b
a Northwestern University, Chicago, Illinois, USA, and
b the Lakeside VA Medical Center, Chicago, Illinois, USA
Key Words. Economics • Cost-effective • Bone marrow transplantation
Dr. Charles L. Bennett, Lakeside VA Medical Center, 333 East Huron Street, Chicago, IL 60611, USA.
 |
Abstract
|
|---|
In an era of health care reform, costs, cost-effectiveness and charges for health care are taking on a larger role in the decision to adopt a new therapy. Hospitals, health maintenance organizations, third-party payers, doctors and patients all have an interest in the economic factors of health care; however, not all new therapies or medications have been analyzed. Bone marrow transplantation is one of the most expensive cancer treatments, costing an average of $193,000 per patient; therefore, many economic studies have focused on the costs of the therapy. However, it is important to note that these findings are not necessarily applicable to other diseases. Cost data are not static and even small changes in protocol can make a large difference in costs. Also, cost data from outside the United States is not always applicable to the United States system due to differences such as socialized medicine, workers' wages and pharmaceutical costs. Many economic analyses focus on the hospitals' charges and extrapolate cost-effectiveness from these figures, yet the amount hospitals charge for services and products is not always relative to the costs. Therefore, third-party payer costs are also important in analyzing the cost-effectiveness of a procedure. This article will review five signature papers which illustrate important issues to consider when trying to determine the costs and cost-effectiveness of bone marrow transplants.
|
Introduction
|
|---|
In 1991, cancer-related deaths in the United States were second only to those due to heart disease, reaching 173.4 per 100,000 people and accounting for 23.7% of the total deaths in the US in 1991 [1 ]. In 1992, the United States spent 817 billion dollars on health care (14% of gross national product), comprising the world's most expensive health care system. Out of this expenditure, bone marrow transplantation (BMT) accounted for $810 million with procedure costs averaging between $100,000 and $300,000 per patient, making it one of the single most costly health care procedures [2 ]. These figures raise the question, "Can we afford to transplant?" When costs were compared over a five-year period for treatment of patients with acute nonlymphocytic leukemia (ANLL), Welch et al. found that on average, allogeneic bone marrow transplantation (alloBMT) cost $193,000, while standard chemotherapy cost $136,000 [3 ]. However, when the cost-per-life-year gained was analyzed, it was determined that alloBMT rated higher at $62,000 per life-year gained when compared to conventional chemotherapy at $64,000 per life-year gained.
Autologous (au)BMT and alloBMT are increasingly utilized for a broader spectrum of diseases, such as breast cancer, and are already the standard procedures for leukemias, some anemias and certain immune diseases. With the growth in unrelated donor allogeneic transplantations, cord blood transplantation and autologous stem cell transplantation, we can expect to see a continual increase in the number of these procedures. Therefore, the costs and cost-effectiveness of these procedures become increasingly important. In this review, we summarize key studies which analyze the economics of transplantation. However, it is important to remember that cost data are not static and that even small changes in protocol, such as the addition of cyclosporin for graft-versus-host-disease (GVHD), can make a large difference in costs. It is also important to consider that cost data from different countries cannot be directly correlated to other countries due to differences such as socialized medicine, workers' wages and pharmaceutical costs. Another important issue, especially in the United States, is charges versus cost. The amount hospitals charge for services and products is not related to their costs. Some hospitals inflate the charges for one department to cover for a less profitable department [4 ]. However, some insightful studies have been conducted on the costs and charges of both allogeneic and autologous bone marrow and peripheral blood transplantations, which can give important clues for reduction of health care costs in the future.
|
Cost-Effectiveness of AlloBMT Versus Chemotherapy for ANLL
|
|---|
One of the first studies to evaluate the costs of alloBMT was conducted by Welch et al. in 1982 [3 ]. This study examined the cost-effectiveness of alloBMT versus chemotherapy for patients with ANLL who were participating in a prospective clinical trial between the years of 1978 and 1982. The study was conducted at the University of Washington and consisted of 41 patients, 17 of whom had a matched donor and BMT, and 19 of whom received two courses of consolidated chemotherapy. All patients had induction chemotherapy at the University of Washington, but subsequent alloBMT or chemotherapy occurred at many hospitals.
Resource-based measurements of costs were collected for what were determined to be the cost drivers between the two therapies: hospital days nonintensive care unit (non-ICU), hospital days intensive care unit (ICU), number of lab tests performed, number of x-rays and number of operating room procedures. Only the number of procedures was counted, not the charges the hospitals billed, thereby reducing discrepancies between hospitals and allowing for uniform data collection. The authors reported a 59% survival rate at five years for BMT patients and only a 26% survival rate at five years for chemotherapy patients.
The study found that chemotherapy patients had an average of seven hospitalizations and alloBMT patients had an average of 4.6 hospitalizations. However, alloBMT patients were more likely to be admitted to an ICU than were chemotherapy patients because of more frequent complications such as GVHD or cytomegalovirus (CMV) infection. The average five-year cost for a chemotherapy patient was $136,000 versus $193,000 for a BMT patient. The biggest difference in cost drivers was time spent in the ICU. Only 5% of chemotherapy patient hospital days were spent in the ICU, compared with 57% for BMT patients. It was also determined that patients who survived five years had lower hospital costs than patients who die. The average cost at five years for a survivor of alloBMT was $166,000 compared with $232,000 for a nonsurvivor, while the average cost at five years for a chemotherapy patient was $79,000 compared with $157,000 for a nonsurvivor. This illustrates the large costs involved with complications. However, it was estimated that alloBMT was cost-effective. This is because most BMT patients were young and the chance of obtaining complete remission was much higher than in chemotherapy patients. Therefore, even though the procedure initially cost $193,000, over the course of a patient's life (assuming 30 years of age at transplantation), the procedure cost only $10,000 per life-year gained.
This study was complete in its analysis of the costs of alloBMT versus chemotherapy for ANLL; however, it is important to note that these findings are not necessarily applicable to other diseases. Also, although the results are statistically significant, the sample size (41 patients) is small, and a larger study might give greater insight into cost drivers. Finally, many advances have been made in alloBMT procedures to help reduce the risk of GVHD and CMV, which caused many patients to spend longer periods of time in the ICU.
|
The Costs of AlloBMT Compared to Chemotherapy for Four Diseases
|
|---|
Viens-Bitker et al. also published a study in 1989 which broadened the Welch et al. study by analyzing the costs of alloBMT compared to chemotherapy for four diseases: acute myelogenous leukemia (AML), severe combined immunodeficiency (SCID), severe aplastic anemia (SAA), and chronic granulocytic leukemia (CGL) [5 ]. The study was conducted in three Parisian hospitals, and patients were only seen at the particular hospital which specialized in their disease. Data were collected for 12 months after diagnosis because it was determined that most complications occur during the six months after transplantation.
As was performed in the Welch et al. study, data were collected for cost drivers: pharmaceutical and blood products, disposable medical supplies, laboratory tests, radiological imaging, medical and nursing care, HLA typing and donor costs, and outpatient care. This study did not collect data on days spent in the hospital, which was the most significant cost driver in the previous study, and the costs were combined for blood products and pharmaceuticals. However, data were collected on outpatient care and donor costs, possible significant factors that the previous study did not address.
The authors concluded that direct standard cost varied greatly depending on disease. For a patient with AML, a BMT cost $40,923; for a patient with SAA, the average cost was twice that amount: $84,537. Patients with SCID averaged $56,814 for a BMT, and for CGL, the average cost was $64,937 (Fig. 1
). For patients with AML, SAA and CGL, the largest cost driver was laboratory and x-rays, which accounted for, respectively, 29.3%, 31.1% and 36.7% of the total cost. For patients with SCID, nursing care accounted for 50.8% of the total cost.
View larger version (32K):
[in this window]
[in a new window]
|
Figure 1. Cost of alloBMT in four diseases [5]. French francs converted to US dollars at the 1989 rate of .16339 $/FF.
|
|
Costs for patients who developed complications were significantly higher (Fig. 2). For patients with AML, an analysis of the most frequent post-transplant complications included thrombocytopenia lasting three months, acute GVHD, interstitial pneumonia and localized zoster virus infection. These post-transplant complications increased the cost of allogeneic transplantation by 36.5%, largely due to the costs of blood products. In patients with SCID, major complications result from HLA haploidentical BMT, which increased the costs of transplantation by more than 148%. This increase was due to the increase in necessary blood products and procedures. In patients with SAA, the most severe complication for transplant patients is acute GVHD. Treatment of this complication increased the cost of transplantation by 274% over direct standard cost, mainly for blood and pharmaceutical products.
View larger version (33K):
[in this window]
[in a new window]
|
Figure 2. Cost of alloBMT in three diseases [5]. French francs converted to US dollars at the 1989 rate of .16339 $/FF.
|
|
This study clearly proves that pathology has a large influence on the average cost of BMT. Because this study was conducted in France under a socialized medical environment, the cost estimates given would not necessarily be proportional to the costs in another country. However, the findings on differences due to disease should be taken into account when trying to forecast the costs of projected BMTs. The study also highlighted areas which accounted for the most significant cost drivers for each disease, thus allowing physicians and pharmaceutical companies to focus on new innovations in these areas.
|
The Costs of Allogeneic or Autologous Transplant Versus Chemotherapy for Patients with AML
|
|---|
Another French study conducted by DuFoir et al. compared the costs of allogeneic or autologous transplant versus chemotherapy for patients with AML [6 ]. Forty patients in first remission were treated in France as a part of two separate clinical trials. Cost data were retrospectively collected for five years after transplantation. Data were collected from hospital records and information from the economics department of the CHR Bordeaux. Data were collected on the following resources: pharmacy charges, blood products, medicotechnical treatments, medical staff expenditures, accommodation (no room charge; only linen charge), amortized charge of laminar air flow (LAF) rooms and donor charges (for allogeneic transplantation).
For patients who had undergone alloBMT, it was determined that the mean number of days that they remained in the hospital was 62 (24 days spent in an LAF room), excluding those who had GVHD. When all patients who underwent alloBMT were accounted for, it was estimated that an average of 101 days was spent in the hospital. For patients who underwent AuBMT, and were not readmitted to the hospital, the average number of days spent hospitalized was 80 (40 of which were spent in LAF rooms). When accounting for patients who were readmitted to the hospital, it was estimated that the average length of hospitalization was 104 days. All patients receiving AuBMT underwent double grafting, which doubled the average length of hospitalization. Chemotherapy patients were hospitalized for an average of 30 days, 22 of which were spent in protected care units. The authors concluded that in this study the number of days spent in the hospital and in LAF units was not a major cost driver. However, in a nonsubsidized US hospital, this would be a significant cost driver [3 ].
The authors broke costs into two categories: costs of care with no relapse and total costs including costs of relapsed patients. For the group of patients who did not relapse, it was estimated that the average cost for chemotherapy was $23,726. The largest cost driver for these patients was blood bank costs, which accounted for 42% of the total cost, followed by medical staff costs, which accounted for 20% of the total cost. For patients receiving alloBMT, the average cost was $72,261, with the largest cost driver being medicotechnical treatments accounting for 30% of the total costs (due to total body irradiation and more biological testing such as virus isolation). The second most expensive component for patients undergoing alloBMT was blood bank costs, which accounted for 25% of the total. AuBMT was estimated to be the most costly of the three procedures, and for patients who did not relapse, the average cost was $72,396. For AuBMT patients, the largest cost driver was blood bank costs, 45% of the total costs, due to prolonged thrombocytopenia and multiple leukaphereses.
It was determined that the risk of relapse at five years was notably lower for patients receiving alloBMT. Only 12% of these patients relapsed compared with 42% of AuBMT patients and 66% of chemotherapy patients. It was estimated that the average cost of relapse was the same for each procedure at $49,024. Therefore, the average cost of chemotherapy with complications was $56,092, for alloBMT $78,144, and for AuBMT $92,987. The criteria used to determine whether a donor was compatible in this study for alloBMT were not presented; therefore, the rate of acute GVHD and relapse could vary considerably with other published studies, which in turn affects the cost of the procedure.
The authors concluded that chemotherapy was most cost-effective at $19,990 per life-year gained at five years, followed by alloBMT at $20,645, and finally AuBMT at $26,253. It was also estimated that the differences in cost between alloBMT and chemotherapy patients were not statistically significant. Therefore, due to the greater survival rate of patients undergoing alloBMT, this is the preferred procedure. It was also determined that there was no statistically significant difference between alloBMT and AuBMT; however the authors stated that AuBMT gave inferior medical results. Without the results of a randomized clinical trial, these results cannot be transferred to other countries.
The estimated costs of a BMT vary greatly with United States studies. These variances can be attributed to differences in labor, pharmaceuticals, overhead and the structure of socialized medicine. The authors concluded that AuBMT was the most costly of the three procedures and had inferior clinical results. However, a clinical trial has not been conducted to validate these results.
|
BMT and the Learning Curve Effect
|
|---|
Bennett et al. conducted a study at the University of Nebraska which illustrates the differences in cost over time at a BMT center [7 ]. In a five-year period, 178 patients with Hodgkin's disease received AuBMT. During this same period, the mortality rate decreased steadily (Fig. 3). The costs of autotransplants also decreased significantly between 1987 and 1991, with the most significant reduction due to the decrease in hospital days per patient (Fig. 4). In 1987, a patient stayed in the hospital an average of 51 days whereas in 1991, patients were hospitalized for an average of only 32 days. It was determined that changes in costs for medications and other factors only slightly fluctuated and those differences were not statistically significant. From 1987 to 1991, 149 AuBMTs were performed for patients with non-Hodgkin's lymphoma. The in-hospital mortality rate decreased significantly over time and the average costs for treatment of non-Hodgkin's lymphoma decreased from $91,000 in 1987 to $74,000 in 1991 (Figs. 3 and 4). Again, hospitalization days were important for these cost reductions; length of stay averaged 45 days in 1987 and 38 days in 1991.
View larger version (11K):
[in this window]
[in a new window]
|
Figure 3. Costs of care and outcomes for high-dose therapy and autologous transplantation for lymphoid malignancies: results from the University of Nebraska 1987 through 1991. [Reproduced from the J Clin Oncol 1995;13:969-973, by copyright permission of W.B. Saunders Company] [7].
|
|
View larger version (11K):
[in this window]
[in a new window]
|
Figure 4. Costs of care and outcomes for high-dose therapy and autologous transplantation for lymphoid malignancies: results from the University of Nebraska 1987 through 1991. [Reproduced from the J Clin Oncol 1995;13:969-973, by copyright permission of W.B. Saunders Company] [7 ].
|
|
Factors associated with decreasing costs were: hospital staff improvements associated with a learning curve, improvements in technology and the process of selective referral. One example of a learning curve effect is that over time the transplant team learned how to better use antibiotics by adding a specialist in infectious diseases and oncology to the transplant team. Also, an example of new technologies that affected costs would be the use of hematopoietic growth factors. Organizations also have a learning curve and become more efficient over time. The transplant service determines which personnel are best suited for the transplant team, which patients should be selected to receive treatment, where to care for patients and what equipment is necessary. These issues can be very different at different hospitals. Many sites need to be surveyed in order to compile a comprehensive list of costs.
|
Third-Party Payer Costs for BMT
|
|---|
While most economic analyses of BMT try to elucidate the cost of a procedure so that results can be generalized for many institutions, Griffiths et al. evaluated the charges that third-party insurers paid for BMT [8 ]. This information is important to policymakers at insurance companies as well as the general public. Data were retrospectively extracted from 413 patient bills for the years 1980 to 1987. All patients had alloBMTs at Johns Hopkins Hospital. Inpatient data were collected for 180 days after transplantation for room and board, pharmacy, radiology, laboratory supplies, physical therapy, and blood and operating room charges. When determining the final charges, an inflation rate of 11.5% was taken into consideration.
Within six months of transplantation, 170 (42.3%) of the 402 patients were rehospitalized. Rehospitalization and survival were largely based on whether the patient contracted acute GVHD or CMV. The estimated overall charge (in 1990 United States dollars) for an alloBMT was $208,410. It was estimated that the patients who contracted CMV and acute GVHD were charged an average of $297,232. Patients who were only diagnosed with acute GVHD had a total patient cost of $195,356, and patients who contracted CMV had total patient charges amounting to $209,365. For patients who were not diagnosed with either CMV or acute GVHD, the total charges were only $142,373.
It was also noted that payer cost increased from 1980 to 1983 and then decreased from 1984 to 1987. This was attributed to two factors— technological advances and cost containment. In 1984, cyclosporine was available to BMT patients, which lowered the risk of acute GVHD from 65% to 52% [9– 11 ]. Also, acyclovir for viral infections became available [12]. The authors concluded that further technological advances in treatments for GVHD and CMV would reduce the overall patient costs for BMT.
|
Conclusion
|
|---|
These five studies give an insight into the costs, cost-effectiveness and third-party charges of AuBMT and AlloBMT. In general, it can be concluded that alloBMT with a matched related donor is the most cost-effective procedure for many diseases. However, the proliferation of autologous peripheral stem cell transplantation programs, improvements in outpatient supportive care programs and the advent of new techniques for expansion of stem cells could change this. Further advances which would minimize GVHD and CMV infections would increase the cost-effectiveness of alloBMT. Also, advances made by transplant teams in streamlining procedures and staffing will reduce the procedural costs. In the United States, the amount of time spent in the ICU has a large impact on the overall costs. Therefore, hospitals and physicians are exploring step-down units and outpatient transplant centers to lower the costs. In Europe, the largest costs are for laboratory and blood products. This type of information can be invaluable to policymakers who try to reduce health care inflation. An ongoing project coordinated by the International Bone Marrow Transplant Registry (C. Bennett, M.D., Ph.D., Principal Investigator) is collecting cost and resource utilization information for both alloBMT and auBMT procedures from major centers in the United States and Canada. A major obstacle to multicenter cost and resource studies is evaluating outpatient medical care where data are often difficult to collect. Studies from integrated delivery systems with automated data sources are needed to completely account for all inpatient and outpatient care, otherwise cost estimators are likely to underestimate the cost of outpatient medical care. Timely reporting and feedback of this data to participating centers is essential if investigators are to use economic data to improve their delivery of health care.
|
Acknowledgments
|
|---|
Dr. Bennett is supported by a Career Development Award from the Veterans Administration Health Services Research and Development program and by the Robert H. Lurie Cancer Center core grant CA 60553-04 from the National Institutes of Health.
|
Footnotes
|
|---|
Provisionally accepted December 20, 1995.
|
References
|
|---|
-
Wingo PA, Tong T, Bolden S. Cancer statistics. CA 1995;45(1):17–18.
-
Gross S. Responsibilities, resources and regulation in bone marrow transplantation. Leukemia 1993;7:1104–1105.[Medline]
-
Welch HG, Larson EB. Cost effectiveness of bone marrow transplantation in acute nonlymphocytic leukemia. N Engl J Med 1989;321:807–812.[Abstract]
-
Bennett CL, Westerman IL. Economic analysis during phase III clinical trials: who, what, when, where, and why. Oncology 1995;9:1–7.
-
Viens-Bitker C, Rery-Lemonnier E, Blum-Boisgard C. Cost of allogeneic bone marrow transplantation in four diseases. Health Policy 1989;12:309–317.[Medline]
-
DuFoir T, Saux MC, Terraza B et al. Comparative costs of allogeneic or autologous bone marrow transplantation and chemotherapy in patients with acute myeloid leukemia in first remission. Bone Marrow Transplant 1992;10:323–329.[Medline]
-
Bennett CL, Armitage JL, Armitage GO et al. Costs of care and outcomes for high-dose therapy and autologous transplantation for lymphoid malignancies: results from the University of Nebraska 1987 through 1991. J Clin Oncol 1995;13:969–973.[Abstract]
-
Griffiths RI, Bass EB, Powe NR et al. Factors influencing third party payer costs for allogeneic BMT. Bone Marrow Transplant 1993;12:43–48.[Medline]
-
Deeg HJ, Storb R, Thomas ED et al. Cyclosporine as prophylaxis for graft-versus-host disease: a randomized study in patients undergoing marrow transplantation for acute nonlymphoblastic leukemia. Blood 1986;54:1325–1334.
-
Santos GW, Tutschka PJ, Brookmeyer R et al. Cyclosporine plus methylprednisolone as prophylaxis for graft-versus-host disease: a randomized double-blind study in patients undergoing allogeneic marrow transplantation. Clin Transplant 1987;1:21–28.
-
Riley WA, Schneckenburger WA. The Maryland experience and a practical proposal to expand existing models in ambulatory primary care. JAMA 1991;1226:1118–1121.
-
Wingard JR. Advances in the management of infectious complications after bone marrow transplantation. Bone Marrow Transplant 1990;6:371–383.[Medline]
Received November 14, 1995;
accepted for publication February 27, 1996.
Top
Abstract
Introduction
