Medicare coverage of Auto HSCT for patients > 65 years with multiple myeloma (MM) began in 2000. In the decade that followed, both the number and cost of Auto HSCT increased dramatically. By 2009, patients over the age of 60 accounted for 40% of the Auto HSCT with MM being the leading indication (Center for International Blood and Marrow Research, 2011). The cost implications of the increased use of Auto HSCT are not well described. The majority of studies looking at the economics of HSCT are single institutional studies focusing on younger patients for which a wide range of costs and indications are reported. Using the SEER-Medicare database, we examined the costs and life year benefit of Auto HSCT for MM patients over age 65 by evaluating the difference over time relative to comorbidity burden.
Patients were identified as having an Auto HSCT based on ICD-9 or HCPCS codes (i.e., ICD-9 41.00, 41.01, 41.04, 41.07, and 41.09; or HCPCS 38241) and were 66 years of age or older at the time of diagnosis. Two time periods were used: 10/00-6/04 (Early) and 7/04-1/08 (Late). Costs and survival were calculated for the first 100 days (d), 1 year (yr), and 3 yrs of follow-up post-Auto HSCT. Data on comorbidities, based on the Charleston Comorbidity Index (CCI), were recorded at the time of diagnosis and HSCT. Patients were classified as CCI0 if they had no comorbidity or CCI1+ if they had one or more comorbid conditions. All cost data were reported in 2009 dollars, adjusted using the medical care component of the Consumer Price Index.
10,832 patients were found to have multiple myeloma, with 275 receiving an Auto HSCT between 10/00 and 1/08. 129 (47%) Auto HSCTs occurred in the Early and 146 (53%) in Late periods. Patients in both time periods were primarily Caucasian; the majority were male; and the median age was 70 yrs (range 66-93). Median time from diagnosis to transplant was 368d and 373 d, with 100d post transplant survival 91% and 97% (p=0.07), 1-yr 83% and 89% (p=0.14), and 3-yr 53% and 60% (p=0.28), respectively. Median costs for each follow-up time increased between Early and Late, even after accounting for inflation. While the distribution of CCI at Auto HSCT was similar in both time periods (CCI0, 62% and 64%, respectively), two key differences were noted in post transplant survival. One-yr survival in the early period was significantly higher for patients with CCI0 than CCI1+ (p=0.006). Similarly, 100-day survival in the Late period was significantly better for those without comorbidities (p=0.035). Furthermore, 3 yr survival for CCI1+ increased from 41% to 60% (p= 0.061) between the Early and Late periods. Median 100-day costs were greater for CCI0 than CCI1+ in both Early ($85696 vs $74381) and Late ($84774 vs $77081); however, by 3 years, the cumulative cost for CCI1+ was greater ($213,757 vs $228,974 in Early and $208,410 vs $339,827 in Late).
Over the past decade, there has been an increase in the utilization of Auto HSCT for MM in the elderly population along with an overall increase in health care cost. With a similar distribution of comorbidity over time, there was significantly increased survival in the later period with an associated increase in cost. Timing and components of the cost will be identified and may help in determining care decisions. This analysis highlights the interconnected nature of utilization, cost, and survival for patients receiving Auto HSCT, and any evaluation of one of these outcomes must also consider the others.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.