Stem Cell Transplants Appear Less Costly Than BMT in Breast Cancer, But Not in NHL

CHICAGO—The first US multicenter cost analyses comparing autologous peripheral blood stem cell transplant (PBSCT) and autologous bone marrow transplant (BMT) in breast cancer and non-Hodgkin’s lymphoma (NHL) showed a cost advantage for stem cells in breast cancer, but not NHL. Tammy Stinson, project manager and analyst for Health Services Research, Northwestern University and VA Chicago Healthcare System, presented the results at the ASH meeting.

She said that this information could be useful for supporting the introduction of new technologies and for assisting hospitals and insurance companies in case rate negotiations. “Oncology services are now more often covered by managed care. Clinically based cost evaluations are important to assure appropriate reimbursement,” Ms. Stinson said.

The current study looked at 609 patients (260 NHL, 349 breast cancer) from four academic transplant centers treated between 1994 and 1996 with high-dose chemotherapy and transplant: 60% PBSCT and 40% BMT.

The Autologous Blood and Marrow Transplant Registry (ABMTR) provided clinical and demographic data on the patients, while the participating centers provided inpatient and outpatient billing data for each patient.

Ms. Stinson said that the study investigators, including Charles Bennett, MD, and Teresa Waters, of Northwestern, have had years of experience deciphering costs.

The methodology used was a cost-identification analysis. The researchers calculated direct medical care costs associated with transplant, including inpatient and outpatient use of facilities up to 100 days post-transplant. Hospital charges were converted to costs utilizing department-specific cost-to-charge ratios.

A 3% discount rate and adjustments for inflation based on the Consumer Price Index were taken into account over the course of the 3-year study. Median values were used to identify significant differences between the two methods of transplantation.

Sampling of Bills Reviewed

To make sure that all associated costs were included, Ms. Stinson said she reviewed a sampling of the bills from each site to make sure that the key transplant procedures were included.

“I know what charges are typically there, and we developed methods to estimate missing charges,” she said. “For instance, if a center left off a chemotherapy charge, I would use ABMTR information to find a treatment profile match and estimate the costs from there. However, this didn’t happen very often.”

Physician charges were not included, she said, because “it was often difficult to get this information and determine if it was correct.” Physician charges are commonly not included in economic analyses of treatments, she said.

 The study found that overall, for breast cancer patients, the median cost of PBSCT was $8,809 lower than the cost of BMT ($58,645 vs $67,454). For NHL patients, the $5,719 difference ($67,999 for PBSCT vs $73,618 for BMT) favored PBSCT but was not significant.

The main cost drivers for both types of cancers and in both transplant methods were pharmacy, room, blood bank products and lab testing. The lack of a cost difference between BMT and PBSCT for NHL may be due to the high variance in the median values of these cost drivers, she said.

“The non-Hodgkin’s costs are more variable than breast cancer costs,” Ms. Stinson explained, “because for breast cancer patients whether the transplant therapy is bone marrow or stem cell, it is often first-line therapy.” In NHL patients, it may be second-, third, or even fourth-line therapy. In this study, 34% of the NHL patients were transplanted after primary induction failure, 17% in CR, and 47% following relapse.

A New Cost Study

Currently, a new cost study of PBSCT vs BMT is under review with the National Cancer Institute through the NIH. It will be done in collaboration with the ABMTR.

Dr. Mary Horowitz, of the International Blood and Marrow Transplant Registry (IBMTR), Dr. Bennett, Ms. Waters, and Ms. Stinson will be the chief investigators. Eighteen institutions will be included, and the researchers hope to accrue 9,000 to 10,000 patients. They also hope to look at hospital organizational behavior and how it relates to cost.