Abstract
The primary aim of this study was to describe healthcare costs and utilization during the first year after a diagnosis of acute myeloid leukemia (AML) for privately insured non-Medicare patients in the United States (US) aged 50-64 years who were treated with either chemotherapy or chemotherapy and allogeneic hematopoietic cell transplantation (alloHCT). MarketScan (Truven Health Analytics) adjudicated total payments for inpatient, outpatient and prescription drug claims from 2007-2011 were used to estimate costs from the health system perspective. Stabilized inverse propensity score weights were constructed using logistic regression to account for differential selection of alloHCT over chemotherapy. Weighted generalized linear models (GLM) adjusted costs and utilization (hospitalizations, inpatient days and outpatient visit-days) for differences in age, sex, diagnosis year, region, insurance plan type, Elixhauser Comorbidity Index (ECI) and 60-day pre-diagnosis costs. Because mortality data were not available, models could not be adjusted for survival times. Among 29,915 patients with a primary diagnosis of AML, a total of 985 patients met inclusion criteria (774 [79%] receiving chemotherapy alone and 211 [21%] alloHCT). Adjusted mean one-year costs were $280,788 for chemotherapy and $544,178 for alloHCT. Patients receiving chemotherapy alone had a mean of 4 hospitalizations, 52.9 inpatient days and 52.4 outpatient visits in the year following AML diagnosis; patients receiving alloHCT had 5 hospitalizations, 92.5 inpatient days and 74.5 outpatient visits. Treating AML in the first year after diagnosis incurs substantial health care costs and utilization with chemotherapy alone and with alloHCT. Our analysis informs health care providers, policy makers and payers so they can better understand treatment costs and utilization for privately-insured patients age 50-64 with AML.
Keywords: Hematopoietic cell transplantation, acute myeloid leukemia, cost analysis, resource utilization, claims data
Introduction
Acute myeloid leukemia (AML) comprises an assorted group of hematopoietic neoplasms characterized by excessive clonal proliferation of myeloid precursor cells in the peripheral blood and bone marrow.1 AML primarily affects older adults, with an incidence of 10.8 per 100,000 for patients age 50 years and older versus 1.3 per 100,000 in patients younger than 50.2 Treatment for AML typically involves induction chemotherapy followed by consolidation chemotherapy (chemotherapy alone) or induction chemotherapy followed by allogeneic hematopoietic cell transplantation (alloHCT), based on disease-related risk factors (e.g., cytogenetic and molecular findings), patient-related variables (e.g., age, performance status), and donor availablity.3,4 Reduced-intensity conditioning regimens now allow alloHCT for many older patients with acceptable risks of morbidity and non-relapse mortality.3,5–9
Among patients who undergo chemotherapy alone or alloHCT, healthcare costs associated with treatment may create additional barriers to patients' healthcare access.10–16 Previous studies on costs for patients with AML generally focused on single-institution experiences15,17 or on patients covered by Medicare.18,19
The primary aim of this study was to explore health care costs and utilization during the first year after AML diagnosis for privately insured non-Medicare patients age 50-64 years in the United States (US), who received either chemotherapy alone or who subsequently underwent alloHCT. Better understanding of healthcare costs and utilization for this age group, where AML is more prevalent and individuals are primarily covered by private insurance, can assist payers and providers with improving access to appropriate care while ensuring wise use of limited resources.
Methods
Data Source
MarketScan (Truven Health Analytics) comprises medical and prescription drug claims for over 115 million patients broadly representing the US privately insured population. Our dataset included claims from 2007-2011 for 29,915 patients who had at least one claim containing an International Classification of Disease-9 (ICD-9) diagnosis or procedure code indicating AML (Appendix A). This research was considered exempt by the National Marrow Donor Program® Institutional Review Board.
Patient Selection
Three primary inclusion criteria were imposed: 1) a primary diagnosis code indicating AML on at least one inpatient claim or at least 2 outpatient claims within 3 months of each other; 2) first claim date with primary diagnosis of AML (“diagnosis date”) between March 2007 and December 2010; and 3) age 50-64 years on diagnosis date. Exclusion criteria were used to improve specificity of patient selection and treatment (Figure 1). Our goal was to select patients who underwent initial treatment of chemotherapy alone or alloHCT for AML with curative intent, excluding patients whose treatment was for recurrent disease, secondary AML or for blood disorders other than AML.20 Patients with total first year post-diagnosis costs below extreme value thresholds based on gamma distributions fitted to observed costs were independently reviewed for eligibility by clinician co-authors (LB, NM, WS).
Figure 1. Patient selection20.
Healthcare cost and utilization
Inpatient, outpatient, prescription drug and total costs from the health system perspective were estimated from adjudicated claims (including insurance plan payments, patient copayment/co-insurance, and deductibles) for up to one year after diagnosis date. Resource utilization per patient was calculated by summing the number of inpatient hospitalizations, inpatient days and outpatient visits for up to one year after the AML diagnosis date. When outpatient claims indicated a date of service occurring during an inpatient stay (e.g., for lab tests), they were counted as part of the inpatient stay. For patients without prescription drug coverage, prescription drug costs were imputed from mean drug costs of all patients with prescription drug coverage. Prescription drug costs were included with outpatient costs. Time trends were accounted for by including year of diagnosis in regression analyses.
Statistical analysis
Patient characteristics were analyzed by treatment group (chemotherapy alone or alloHCT), reported as count and proportion or median and range, and compared using t-tests for continuous variables and χ2 tests for categorical variables.
Unadjusted mean (with 95% confidence intervals) healthcare cost and utilization were estimated in the first year after diagnosis by treatment group. Because treatment was not randomly assigned, propensity scores were used to account for potential bias due to observable characteristics affecting both treatment selection and cost/utilization. Propensity to receive alloHCT (vs. chemotherapy alone) was estimated using logistic regression conditional on patient age, gender, year of diagnosis, US Census region, insurance plan type, inpatient, outpatient and prescription drug utilization during the 60 days prior to diagnosis date, and a categorized version of the Elixhauser comorbidity index (ECI)21,22 as well as a set of indicators of 23 constituent comorbidities present in at least 1.5% of the sample, based on presence 60-days prior to AML diagnosis. The ECI identifies comorbidities as separate, independent measures based on ICD-9 codes, and excludes comorbidities related to the primary diagnosis,21 and is one of the most widely applied measures for use with administrative data for comorbidity risk adjustment.22 Disease characteristics, such as cytogenetics or response to induction therapy, were not available.
Stabilized inverse probability of treatment weights23,24 calculated from the propensity scores were included in weighted generalized linear models (GLM) of healthcare costs (gamma with log-link) and utilization (negative binomial with log-link). Inverse probability weighting increases the weight of individuals least likely to select the treatment that they actually received, thus effectively balancing the sample. Separate models were constructed for total costs, inpatient costs and outpatient costs. Given high costs in the AML population, zero-inflated models were not needed to account for an excessive number of observations with no costs or utilization; rather, in the small number of cases having either zero inpatient or outpatient expenditures, the value was set to $1 as a continuity correction. For each outcome, a sequence of GLM models was constructed including interaction effects of treatment and other model covariates (age category, gender, year of diagnosis, region, insurance plan type, a continuous measure of the ECI, insurance type, and healthcare costs for the 60 days prior to the defined AML diagnosis date). Final model selection was based on model parsimony and comparison of the corrected Akaike information criterion (AICC). Adjusted costs and utilization were predicted from the final models evaluated at covariate sample means and can be interpreted as the expected outcomes for a population in equipoise between selecting chemotherapy alone or alloHCT based on observed variables.
Because inclusion criteria were based on diagnosis and procedure data coded for administrative purposes, sensitivity analyses were conducted to evaluate the robustness of the total cost estimates in different selected samples based on relaxing or tightening inclusion criteria. The primary sensitivity analysis restricted the chemotherapy alone group to those who had a claim for human leukocyte antigen (HLA) typing, and therefore were presumed to be under consideration for HCT. Other sensitivity analyses were completed using patients under 60 years of age only; determining comorbidities and pre-diagnosis costs based on 90 days of claims prior to AML diagnosis as opposed to 60; assigning members of the chemotherapy alone cohort who went on to receive alloHCT more than one year after AML diagnosis to the HCT cohort; and restricting the alloHCT cohort to include only patients who received HCT fewer than 180 days post-diagnosis.
Survival data were not available for analysis. In order to investigate potential relationships between survival time and cumulative costs, an exploratory analysis was conducted in which continued enrollment status was considered as a proxy for survival. Mean adjusted total costs by treatment group were assessed at 3, 6, 9, 12, 18 and 24 months after AML diagnosis for patients who were still enrolled and observable in the insurance plan at those time points (e.g., those diagnosed in 2010 were not observable at 24 months).
All analyses were conducted using SAS Enterprise Guide 6.1 (SAS Institute Inc.).
Results
Patient characteristics
A total of 991 patients satisfied inclusion criteria (780 chemotherapy alone and 211 alloHCT). Gamma distributions fitted to total costs in each cohort identified five alloHCT patients with total costs less than $204,000 and 13 chemotherapy alone patients with total costs less than $28,000 for adjudication. After review, it was determined that 0 of 5 HCT patients and 6 of 13 chemotherapy alone patients did not receive treatment for AML with curative intent and were excluded from the final analysis.
Patient demographics are presented in Table 1. The distribution of patient ages differed between treatments, with patients receiving alloHCT being younger. A higher proportion of patients receiving alloHCT were male compared to patients receiving chemotherapy alone. Geographic region and insurance plan type did not differ significantly by treatment. Pre-existing comorbidity measured by the ECI did not differ significantly by treatment group, though diabetes mellitus was more prevalent in patients receiving alloHCT.
Table 1. Patient demographics for patients age 50-64 years with AML by treatment group.
| Variable | AlloHCT N (%) | Chemotherapy Alone N (%) | P-Value |
|---|---|---|---|
| N | 211 | 774 | |
| Age group (years) | 0.0002 | ||
| 50-54 | 78 (37.0) | 184 (23.8) | |
| 55-59 | 69 (32.7) | 262 (33.9) | |
| 60-64 | 64 (30.3) | 328 (42.4) | |
| Gender | 0.0138 | ||
| Male | 127 (60.2) | 392 (50.7) | |
| Female | 84 (39.8) | 382 (49.4) | |
| Region* | 0.226 | ||
| Northeast | 38 (18.0) | 124 (16.0) | |
| North Central | 68 (32.2) | 221 (28.6) | |
| South | 68 (32.2) | 317 (41.0) | |
| West | 31 (14.7) | 90 (11.6) | |
| Unknown | 6 (2.8) | 22 (2.8) | |
| Plan type | 0.6074 | ||
| PPO & POS | 155 (73.5) | 580 (74.9) | |
| Comprehensive | 7 (3.3) | 31 (4.0) | |
| EPO & HMO | 20 (9.5) | 69 (8.9) | |
| CDHP & HDHP | 13 (6.2) | 29 (3.8) | |
| Unknown | 16 (7.6) | 65 (8.4) | |
| Elixhauser Co-morbidity Index (ECI) | 0.1673 | ||
| Mean (SD) | 3.36 (2.3) | 3.57 (2.3) | |
| Range | 0 - 12 | 0 - 12 | |
| Mean time to first chemotherapy (days) | 0.3743 | ||
| Mean (SD) | 22.8 (37.9) | 20.2 (37.6) | |
| Range | 0 - 213 | 0 - 349 |
Note: Characteristics based on AML diagnosis date; EPO-Exclusive Provider Organization, HMO-Health Maintenance Organization; POS-Point-of-Service; PPO-Preferred Provider Organization; CDHP-Consumer Driven Health Plan; HDHP-High Deductible Health Plan
Regions and states include: Northeast: CT, ME, MA, NH, RI, VT, NJ, NY, PA; Midwest: IN, IL, MI, OH, WI, IA, KS, MN, MO, NE, ND, SD; South: DE, DC, FL, GA, MD, NC, SC, VA, WV, AL, KY, MS, TN, AR, LA, OK, TX; West: AZ, CO, ID, NM, MT, UT, NV, WY, AK, CA, HI, OR, WA
Unadjusted cost and utilization
Unadjusted results are presented in Table 2. For patients receiving alloHCT, mean unadjusted total costs, inpatient costs, outpatient costs, hospitalizations, inpatient days, and outpatient visits within the first year of AML diagnosis were greater than for patients receiving chemotherapy alone.
Table 2. Unadjusted and adjusted healthcare costs and utilization within 1 year of AML diagnosis by treatment group.
| AlloHCT (n = 211) | Chemotherapy Alone (n = 774) | |||
|---|---|---|---|---|
| Unadjusted Mean (95% CI) | Adjusted Mean (95% CI) | Unadjusted Mean (95% CI) | Adjusted Mean (95% CI) | |
| Costs | ||||
| Inpatient | $454,584 ($412,549 - $496,618) | $412,252 ($355,370 - $478,238) | $237,534 ($222,896 - $252,173) | $217,380 ($193,578 - $244,109) |
| Outpatient* | $125,065 ($113,237 - $136,894) | $123,764 ($101,931-$150,272) | $58,994 ($54,574 - $63,415) | $60,526 ($51,841-$70,666) |
| Total* | $579,649 ($535,521 - $623,778) | $544,178 ($489,994-$604,353) | $296,529 ($280,574 - $312,483) | $280,788 ($258,473-$305,030) |
| Utilization | ||||
| Number of Hospitalizations | 5.2 (4.8-5.5) | 5.1 (4.7-5.6) | 4.1 (3.9-4.3) | 4.0 (3.8-4.4) |
| Number of Inpatient Days | 88.0 (82.7-93.3) | 93.5 (82.6-105.9) | 57.1 (54.8-59.4) | 52.4 (48.3-56.8) |
| Number of Outpatient Visits | 72.3 (67.4-77.2) | 74.5 (64.0-86.8) | 47.4 (45.0-49.9) | 49.5 (43.9-55.7) |
Outpatient and total costs included prescription drug costs
Adjusted by: patient age, sex, diagnosis year, region, insurance plan type, 60-day-prior comorbidity and medical costs.
Note: Adjusted total does not equal sum of adjusted inpatient and outpatient because separate models were used for adjustment.
Propensity-weighted GLM models
Detailed propensity models are available in Appendix A1-A2. Receipt of alloHCT (vs. chemotherapy alone), West region (vs. Northeast), higher comorbidity measured by ECI and higher pre-diagnosis outpatient costs were associated with higher total costs in the year after AML diagnosis. Age 60-64 (vs. 50-54), female gender and higher pre-diagnosis inpatient costs were associated with lower total costs (Table 3). The interaction between ECI and treatment type was negative and statistically significant, attenuating the relationship between higher comorbidity and higher cost in the alloHCT group. Coefficients on year of diagnosis did not indicate a significant trend in total costs from 2007-2011.
Table 3. Generalized Linear Regression Model: First year costs (total costs, inpatient costs and outpatient costs) for patients age 50-64 with AML who received of chemotherapy alone or alloHCT from 2007-2010.
| Parameter | Total Costs Overall | Inpatient Costs | Outpatient Costs | |||
|---|---|---|---|---|---|---|
| Estimate | Standard error | Estimate | Standard error | Estimate | Standard error | |
| Intercept | 12.3431 | 0.084** | 12.0941 | 0.1173** | 10.8079 | 0.1595** |
| Treatment | ||||||
| AlloHCT | 0.8779 | 0.0835** | 0.8282 | 0.1179** | 1.0299 | 0.1557** |
| Chemotherapy alone (ref) | - | - | - | - | - | - |
| Age group, years | ||||||
| 55-59 | -0.0501 | 0.0502 | -0.046 | 0.0705 | -0.0740 | 0.0940 |
| 60-64 | -0.1813 | 0.0484** | -0.2126 | 0.0678* | -0.0718 | 0.0913 |
| 50-54 (ref) | - | - | - | - | - | - |
| Gender | ||||||
| Female | -0.1077 | 0.0386* | -0.0893 | 0.0540 | -0.1534 | 0.0729* |
| Male (ref) | - | - | - | - | - | - |
| Region | ||||||
| North Central | -0.0705 | 0.0607 | -0.0638 | 0.085 | -0.0993 | 0.1143 |
| South | 0.0066 | 0.0584 | -0.0249 | 0.0817 | 0.1271 | 0.1107 |
| West | 0.2803 | 0.0738** | 0.2535 | 0.1032* | 0.2825 | 0.1403* |
| Unknown | -0.2202 | 0.1241 | -0.2887 | 0.1744 | -0.0121 | 0.2315 |
| Northeast (ref) | - | - | - | - | - | - |
| Insurance plan type | ||||||
| Comprehensive | -0.0604 | 0.1031 | -0.1632 | 0.1444 | 0.2391 | 0.1943 |
| EPO & HMO | -0.0092 | 0.0691 | -0.0153 | 0.0967 | 0.0020 | 0.1301 |
| CDHP & HDHP | -0.0144 | 0.0963 | 0.0057 | 0.1348 | -0.0884 | 0.1805 |
| Unknown | -0.0907 | 0.0714 | -0.1308 | 0.1002 | 0.0693 | 0.1336 |
| PPO & POS (ref) | - | - | - | - | - | - |
| AML Diagnosis Year | ||||||
| 2008 | 0.0884 | 0.0623 | 0.1116 | 0.0871 | 0.0084 | 0.1181 |
| 2009 | 0.0624 | 0.0589 | 0.0832 | 0.0827 | -0.0131 | 0.1110 |
| 2010 | 0.0292 | 0.0605 | 0.0372 | 0.0848 | -0.0038 | 0.1144 |
| 2007 (ref) | - | - | - | - | - | - |
| ECI | 0.0846 | 0.0097** | 0.0974 | 0.0136** | 0.0438 | 0.0184* |
| ECI*alloHCT | -0.0612 | 0.0197* | -0.0533 | 0.0278 | -0.0891 | 0.0366* |
| 60 days prior -Inpatient costs | -2.2E-06 | 6.97E-07** | -3.50E-06 | 1.01E-06** | 8.98E-07 | 1.35E-06 |
| 60 days prior -Outpatient costs | 1.3E-05 | 4.75E-06** | 1.10E-05 | 6.57E-06 | 2.26E-05 | 9.29E-06* |
| Scale | 2.783 | 0.1181 | 1.4195 | 0.0576 | 0.797 | 0.0308 |
p<0.05;
p<=0.001 Note: Characteristics based on AML diagnosis date; EPO-Exclusive Provider Organization, HMO-Health Maintenance Organization; POS-Point-of-Service; PPO-Preferred Provider Organization; CDHP-Consumer Driven Health Plan; HDHP-High Deductible Health Plan
The signs and significance of coefficients in the models for inpatient and outpatient costs were similar to those for total costs. However, female gender was not associated with inpatient costs, and pre-diagnosis outpatient costs were not associated with inpatient costs. Higher age and pre-diagnosis inpatient costs were not associated with lower outpatient costs, while higher pre-diagnosis outpatient costs were associated with higher outpatient costs.
Adjusted cost and utilization
Adjusted cost and utilization results presented in Table 2. Mean adjusted total costs were $280,788 and $544,178 for chemotherapy alone and alloHCT, respectively. Mean adjusted inpatient and outpatient costs and utilization within the first year of AML diagnosis for patients receiving alloHCT were higher than for patients receiving chemotherapy alone (Table 3).
Receipt of alloHCT and higher comorbidity were associated with a higher number of both inpatient hospitalizations and days (Table 4). Age 60-64 (vs. 50-54) and higher pre-diagnosis inpatient costs were associated with a lower number of both inpatient hospitalizations and days. An interaction between age 60-64 and alloHCT was positive and statistically significant. Fewer inpatient days were predicted for the North Central and Unknown regions compared to the Northeast. In the model for outpatient visits, alloHCT was associated with more outpatient visits; no other predictor had a statistically significant relationship.
Table 4. Generalized Linear Regression Model: First year utilization (inpatient days, inpatient visits and outpatient visits) for patients age 50-64 with AML who received treatment of chemotherapy alone or alloHCT from 2007-2010.
| Parameter | Inpatient Days | Inpatient visits | Outpatient visits | |||
|---|---|---|---|---|---|---|
| Estimate | Standard Error | Estimate | Standard Error | Estimate | Standard Error | |
| Intercept | 3.9518 | 0.0829** | 1.1811 | 0.0792** | 3.7922 | 0.1238** |
| Treatment | ||||||
| AlloHCT | 0.4024 | 0.1082** | 0.1737 | 0.0712* | 0.4099 | 0.0665** |
| Chemotherapy alone (ref) | - | - | - | - | - | - |
| Age group, years | ||||||
| 55-59 | -0.0019 | 0.0474 | -0.0126 | 0.051 | 0.0313 | 0.0722 |
| 60-64 | -0.1225 | 0.0459* | -0.2401 | 0.0509** | 0.0378 | 0.0698 |
| 50-54 (ref) | - | - | - | - | - | - |
| age 55-59*alloHCT | - | - | -0.0575 | 0.1008 | - | - |
| age 60-64*alloHCT | - | - | 0.2442 | 0.1013* | - | - |
| Gender | ||||||
| Female | -0.015 | 0.0367 | 0.0082 | 0.0351 | -0.0527 | 0.0562 |
| Male (ref) | - | - | - | - | - | - |
| Region | ||||||
| North Central | -0.14 | 0.0657* | 0.0341 | 0.0555 | -0.0593 | 0.0897 |
| South | -0.1033 | 0.0631 | -0.0221 | 0.0544 | -0.0329 | 0.0864 |
| West | -0.1418 | 0.0796 | 0.0386 | 0.0687 | 0.1318 | 0.1088 |
| Unknown | -0.3934 | 0.1343* | 0.0293 | 0.1126 | -0.0283 | 0.1804 |
| Northeast (ref) | - | - | - | - | - | - |
| North Central*alloHCT | - | - | - | - | ||
| South*alloHCT | -0.037 | 0.1281 | - | - | - | - |
| West*alloHCT | -0.0559 | 0.164 | - | - | - | - |
| Unknown*alloHCT | 0.7996 | 0.2740* | - | - | - | - |
| Insurance plan type | ||||||
| Comprehensive | -0.1112 | 0.0978 | -0.0498 | 0.0951 | 0.1916 | 0.1477 |
| EPO & HMO | -0.0151 | 0.0660 | 0.058 | 0.0618 | -0.0388 | 0.1010 |
| CDHP & HDHP | 0.0655 | 0.0918 | -0.0206 | 0.0886 | -0.1093 | 0.1408 |
| Unknown | -0.1218 | 0.0694 | -0.0396 | 0.0665 | 0.1334 | 0.1054 |
| PPO & POS (ref) | - | - | - | - | - | - |
| AML Diagnosis Year | ||||||
| 2008 | -0.047 | 0.0592 | -0.0662 | 0.0566 | 0.0317 | 0.0900 |
| 2009 | -0.095 | 0.0559 | -0.0653 | 0.0536 | -0.0777 | 0.0852 |
| 2010 | -0.0783 | 0.0574 | -0.0355 | 0.0545 | -0.0857 | 0.0876 |
| 2007 (ref) | - | - | - | - | - | - |
| ECI | 0.0867 | 0.0082** | 0.0932 | 0.0074** | 0.0192 | 0.0123 |
| 60 days prior - Inpatient costs | -0.0036 | 0.0007** | -0.0026 | 0.0008** | 0.0012 | 0.0011 |
| 60 days prior - Outpatient costs | 0.0068 | 0.0044 | 0.007 | 0.0042 | 0.0111 | 0.0071 |
| Dispersion | 0.3047 | 0.0147 | 0.0525 | 0.0123 | 0.7257 | 0.0334 |
p<0.05;
p<=0.001
Note: Characteristics based on AML diagnosis date; alloHCT- Allogeneic Hematopoietic Cell Transplantation; EPO-Exclusive Provider Organization, HMO-Health Maintenance Organization; POS-Point-of-Service; PPO-Preferred Provider Organization; CDHP-Consumer Driven Health Plan; HDHP-High Deductible Health Plan; ECI: Elixhauser Comorbidity Index
Sensitivity Analysis
Mean adjusted total costs were substantially higher based on the model estimated with the subset of 95 chemotherapy alone patients who had HLA typing performed: $354,267 (95% CI: $301,383-$416,410) for chemotherapy alone and $627,161 (95% CI: $550,054-$715,076) for alloHCT. There was a larger magnitude for the negative coefficient on female gender. Other sensitivity analyses produced only modest changes in adjusted total costs (Appendix A3-A4).
Exploratory analysis
Results of the exploratory analysis are presented in Figure 2. The difference in mean adjusted total costs between continuously enrolled patients receiving chemotherapy alone or alloHCT increased from 3 months to 18 months before declining at 24 months. Total unadjusted costs among those enrolled at 12 months were similar to those for all patients: $298,964 vs. $296,529 (p=0.8463) for chemotherapy alone and $605,039 vs. $579,649 (p=0.5051) for alloHCT.
Figure 2. Adjusted costs at 3, 6, 9, 12, 18 and 24 months after AML diagnosis for patients who received chemotherapy alone or alloHCT who were continuously enrolled up to that time interval.
Discussion
This study is the first to use nationally representative data to assess healthcare costs and utilization of chemotherapy alone and alloHCT among privately-insured US patients age 50-64 years with AML. The findings demonstrate that adjusted healthcare costs and utilization during the first year after AML diagnosis were substantially less for patients treated with chemotherapy alone than for patients treated with alloHCT.
Our cost estimates are substantially higher than previously published results in older US populations.18,19 This difference may be due to the different payer type, age group of the population studied, type of costs included (adjudicated costs vs. charges), or more recent time frame of data. In a study examining chemotherapy and supportive care treatment, survival and costs for a Medicare population (age ≥65) with primary AML from 1997-2007, Meyers et al. found that after AML diagnosis, mean health costs were $96,078.18 Menzin et al., in a study that used Medicare data from 1991-1996, found that costs for patients age ≥65 exceeded $41,000 (1998 dollars) in the two years after diagnosis.19
Our findings that patients who are older, female or who had higher pre-diagnosis inpatient costs had lower expected total costs in the year after diagnosis contrasts with the broader risk adjustment literature, which tend to report female gender, older age, and higher pre-diagnosis utilization to be associated with higher subsequent costs25,26 We cannot rule out the presence of one or more confounding variables, such as inability to tolerate aggressive therapy, and we are unable to account for the competing risk of mortality, which may be associated with these variables.
Although differences in observable characteristics were accounted for through inverse propensity of treatment weighting and GLM, propensity scores cannot account for full treatment selection bias due to unobserved characteristics (such as disease stage), which are likely to affect treatment selection and healthcare utilization and costs. Our primary sensitivity analysis, restricting the chemotherapy alone group to patients who had HLA typing (and therefore must have been at least considered as potential candidates for alloHCT) supported this. The magnitudes of coefficients for ECI and the interaction between ECI and receipt of alloHCT were smaller than coefficients for the full chemotherapy group, suggesting an attenuated association between comorbidity and total costs. Potential survival benefits for older patients receiving alloHCT27,28 might contribute to differences in costs between treatment groups. Patients not able to undergo transplantation may die sooner, possibly leading to lower costs in the chemotherapy alone group. Since mortality data was not available, we treated continuous enrollment in insurance as a proxy for survival in an exploratory analysis. However, disenrollment may be due to reasons other than death, including a change in insurance plans or loss of insurance coverage. We found that for patients who received chemotherapy alone, costs for the continuously enrolled were lower than for all patients, suggesting that patients receiving chemotherapy alone who die sooner may not necessarily incur lower costs. Our exploratory analysis also suggested that cost differences between chemotherapy alone and alloHCT may be lower among patients who survive to two years or more. Both results merit additional research.
In previously published work, we have discussed challenges associated with using administrative claims data to accurately identify cohorts with newly-diagnosed AML and their treatment groups.20 Given these limitations, caution must be exercised in interpreting any results as comparative between chemotherapy alone and alloHCT. Notably, administrative claims data are generally unable to provide detailed clinical data, such as performance status, cytogenetic and molecular profiles, remission status, response to induction therapy, and level of HLA-match, all of which are known to influence both treatment selection and outcomes.29–31 Though not an option for MarketScan data, the possibility of linking claims databases to medical records or registries such as the Center for International Blood and Marrow Transplant Research (CIBMTR) may allow for clinical risk factors and outcomes of transplant recipients to be explored in the future within the context of costs.32,33
This analysis assessed all health care utilization and costs during the first year of AML diagnosis. A longer-term analysis would provide a more comprehensive view of the costs of chemotherapy alone and alloHCT. This study was restricted to patients age 50-64; there is limited research on privately insured individuals for this age group, and patients 65 and older are often eligible for Medicare.
This analysis included adjudicated claims; it did not represent indirect patient costs or activity-based costs incurred by institutions. Even privately insured patients incur both substantial medical and non-medical costs that place enormous burden on patients and their families and may even contribute to poorer health outcomes as a consequence of “financial toxicity.”34 Patients may need to take a non-formulary medication, or a brand versus generic medication; out-of-pocket costs for these types of medications may not be in claims data. Future work is needed to examine such out-of-pocket expenses.
Conclusion & Implications
These data provide a first step to allow transplant centers, policy makers and payers to better understand treatment costs and utilization for privately-insured patients age 50-64 with AML. Administrative claims databases are useful for such an analysis, but have limitations, including lack of data on relapse and later events that impact total costs of care, and clinical and survival data. Analyses in which these factors can be captured are required to fully understand the impact of different therapeutic approaches.
Supplementary Material
Highlights.
Costs of AML therapy in first year for privately insured patients aged 50-64 years
As expected, total costs were lower for patients receiving chemotherapy alone
Limitations of administrative claims data preclude direct comparisons between therapies
Acknowledgments
Disclosure: CIBMTR® (Center for International Blood and Marrow Transplant Research®) is a research collaboration between the National Marrow Donor Program®/Be The Match® and Medical College of Wisconsin. The CIBMTR is supported by Public Health Service Grant/Cooperative Agreement U24-CA76518 from the National Cancer Institute (NCI), the National Heart, Lung and Blood Institute (NHLBI) and the National Institute of Allergy and Infectious Diseases (NIAID); a Grant/Cooperative Agreement 5U01HL069294 from NHLBI and NCI; a contract HHSH234200637015C with Health Resources and Services Administration (HRSA/DHHS); two Grants N00014-06-1-0704 and N00014-08-1-0058 from the Office of Naval Research; and grants from AABB; Allos, Inc.; Amgen, Inc.; Anonymous donation to the Medical College of Wisconsin; Astellas Pharma US, Inc.; Be the Match Foundation; Biogen IDEC; BioMarin Pharmaceutical, Inc.; Biovitrum AB; BloodCenter of Wisconsin; Blue Cross and Blue Shield Association; Bone Marrow Foundation; Buchanan Family Foundation; CaridianBCT; Celgene Corporation; CellGenix, GmbH; Children's Leukemia Research Association; ClinImmune Labs; CTI Clinical Trial and Consulting Services; Eisai, Inc.; Genentech, Inc.; Genzyme Corporation; Histogenetics, Inc.; HKS Medical Information Systems; Hospira, Inc.; Kirin Brewery Co., Ltd.; The Leukemia & Lymphoma Society; Merck & Company; The Medical College of Wisconsin; Millennium Pharmaceuticals, Inc.; Miller Pharmacal Group; Milliman USA, Inc.; Miltenyi Biotec, Inc.; National Marrow Donor Program; Nature Publishing Group; Novartis Oncology; Oncology Nursing Society; Osiris Therapeutics, Inc.; Otsuka America Pharmaceutical, Inc.; Pall Life Sciences; Pfizer Inc; Schering Corporation; Sigma-Tau Pharmaceuticals; Soligenix, Inc.; StemCyte, Inc.; StemSoft Software, Inc.; Sysmex America, Inc.; THERAKOS, Inc.; Vidacare Corporation; ViraCor Laboratories; ViroPharma, Inc.; and Wellpoint, Inc. The views expressed in this article do not reflect the official policy or position of the National Institutes of Health, the Department of the Navy, the Department of Defense, or any other agency of the U.S. Government.
The Health Services Research program is supported in part by Health Resources and Services Administration Contract No. HHSH234200637018C. The views expressed in this article do not reflect the official policy or position of the Health Resources and Services Administration or the National Marrow Donor Program/Be The Match®.
We thank Mary Horowitz, MD, MS, Chief Scientific Director of the Center for International Blood and Marrow Transplant Research; Michael Boo, JD, Chief Strategy Officer of the National Marrow Donor Program/Be The Match; Jackie Foster, MPH, RN, OCN, Patient Education Specialist of the National Marrow Donor Program/Be The Match, and Jill Randall, Team Lead, Patient Services Coordinator of the National Marrow Donor Program/Be The Match for their critical review of the draft manuscript.
Footnotes
- Poster presentation: BMT Tandem Meetings, Honolulu, HI; February 14, 2016
- Oral presentation: Minnesota Health Services Research Conference, St. Paul, MN; March 1, 2016
- Manuscript: Preussler JM, Mau LW, Majhail NS, et al. Administrative Claims Data for Economic Analyses in Hematopoietic Cell Transplantation: Challenges and Opportunities. Biol Blood Marrow Transplant. October 2016. 22(10), 1738–1746. http://doi.org/10.1016/j.bbmt.2016.05.005
Conflicts of Interest: None of the authors has any financial conflicts of interest to report
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Schiffer CA, Anastasi J. Classification of acute myeloid leukemia. [Accessed January 14, 2016];UpToDate. http://www.uptodate.com/contents/classification-of-acute-myeloid-leukemia?source=search_result&search=AML&selectedTitle=1∼150. Published December 17, 2015.
- 2.Acute Myeloid Leukemia - SEER Stat Fact Sheets. [Accessed June 5, 2013]; http://seer.cancer.gov/statfacts/html/amyl.html.
- 3.Ustun C, Lazarus HM, Weisdorf D. To transplant or not: a dilemma for treatment of elderly AML patients in the twenty-first century. Bone Marrow Transplant. 2013 May; doi: 10.1038/bmt.2013.67. [DOI] [PubMed] [Google Scholar]
- 4.Roboz GJ. Current treatment of acute myeloid leukemia. Curr Opin Oncol. 2012;24(6):711–719. doi: 10.1097/CCO.0b013e328358f62d. [DOI] [PubMed] [Google Scholar]
- 5.Majhail NS, Brunstein CG, Shanley R, et al. Reduced-intensity hematopoietic cell transplantation in older patients with AML/MDS: umbilical cord blood is a feasible option for patients without HLA-matched sibling donors. Bone Marrow Transplant. 2012;47(4):494–498. doi: 10.1038/bmt.2011.114. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.McClune BL, Weisdorf DJ, Pedersen TL, et al. Effect of Age on Outcome of Reduced-Intensity Hematopoietic Cell Transplantation for Older Patients With Acute Myeloid Leukemia in First Complete Remission or With Myelodysplastic Syndrome. J Clin Oncol. 2010;28(11):1878–1887. doi: 10.1200/JCO.2009.25.4821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sorror ML, Sandmaier BM, Storer BE, et al. Long-term outcomes among older patients following nonmyeloablative conditioning and allogeneic hematopoietic cell transplantation for advanced hematologic malignancies. JAMA. 2011;306(17):1874–1883. doi: 10.1001/jama.2011.1558. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Chevallier P, Szydlo RM, Blaise D, et al. Reduced-intensity conditioning before allogeneic hematopoietic stem cell transplantation in patients over 60 years: a report from the SFGM-TC. Biol Blood Marrow Transplant. 2012;18(2):289–294. doi: 10.1016/j.bbmt.2011.07.013. [DOI] [PubMed] [Google Scholar]
- 9.Koreth J, Aldridge J, Kim HT, et al. Reduced-intensity conditioning hematopoietic stem cell transplantation in patients over 60 years: hematologic malignancy outcomes are not impaired in advanced age. Biol Blood Marrow Transplant. 2010;16(6):792–800. doi: 10.1016/j.bbmt.2009.12.537. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Department of Health and Human Services. Access to Health Services. [Accessed January 15, 2016];HealthyPeople gov. http://www.healthypeople.gov/2020/topics-objectives/topic/Access-to-Health-Services?topicid=1. Published January 14, 2016.
- 11.Stranges E, Russo A, Friedman B. Procedures with the Most Rapidly Increasing Hospital Costs, 2004-2007 - Healthcare Cost and Utilization Project (HCUP) Statistical Briefs -Statistical Brief #82. [Accessed January 9, 2012]; http://www.ncbi.nlm.nih.gov/books/NBK53597/
- 12.Denzen EM, Thao V, Hahn T, et al. Financial impact of allogeneic hematopoietic cell transplantation on patients and families over 2 years: results from a multicenter pilot study. Bone Marrow Transplant. 2016 Apr; doi: 10.1038/bmt.2016.103. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Majhail NS, Rizzo JD, Hahn T, et al. Pilot study of patient and caregiver out-of-pocket costs of allogeneic hematopoietic cell transplantation. Bone Marrow Transplant. 2012 Dec; doi: 10.1038/bmt.2012.248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Majhail NS, Mau LW, Denzen EM, Arneson TJ. Costs of autologous and allogeneic hematopoietic cell transplantation in the United States: a study using a large national private claims database. Bone Marrow Transplant. 2013;48(2):294–300. doi: 10.1038/bmt.2012.133. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Preussler JM, Denzen EM, Majhail NS. Costs and cost-effectiveness of hematopoietic cell transplantation. Biol Blood Marrow Transplant. 2012;18(11):1620–1628. doi: 10.1016/j.bbmt.2012.04.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Majhail NS, Mothukuri JM, Brunstein CG, Weisdorf DJ. Costs of hematopoietic cell transplantation: comparison of umbilical cord blood and matched related donor transplantation and the impact of posttransplant complications. Biology of Blood and Marrow Transplantation. 2009;15(5):564–573. doi: 10.1016/j.bbmt.2009.01.011. [DOI] [PubMed] [Google Scholar]
- 17.Khera N, Zeliadt SB, Lee SJ. Economics of hematopoietic cell transplantation. Blood. 2012;120(8):1545–1551. doi: 10.1182/blood-2012-05-426783. [DOI] [PubMed] [Google Scholar]
- 18.Meyers J, Yu Y, Kaye JA, Davis KL. Medicare Fee-for-Service Enrollees with Primary Acute Myeloid Leukemia: An Analysis of Treatment Patterns, Survival, and Healthcare Resource Utilization and Costs. Applied Health Economics and Health Policy. 2013;11(3):275–286. doi: 10.1007/s40258-013-0032-2. [DOI] [PubMed] [Google Scholar]
- 19.Menzin J, Lang K, Earle CC, Kerney D, Mallick R. The outcomes and costs of acute myeloid leukemia among the elderly. Arch Intern Med. 2002;162(14):1597–1603. doi: 10.1001/archinte.162.14.1597. [DOI] [PubMed] [Google Scholar]
- 20.Preussler JM, Mau LW, Majhail NS, et al. Administrative Claims Data for Economic Analyses in Hematopoietic Cell Transplantation: Challenges and Opportunities. Biol Blood Marrow Transplant. 2016 May; doi: 10.1016/j.bbmt.2016.05.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8–27. doi: 10.1097/00005650-199801000-00004. [DOI] [PubMed] [Google Scholar]
- 22.Healthcare Cost and Utilization Project (HCUP) Definitions of Comorbidities, Comorbidity Software Version 37/2012. http://www.hcup-us.ahrq.gov/toolssoftware/comorbidity/Table2-FY12-V3_7.pdf.
- 23.Rosenbaum PR. Model-Based Direct Adjustment. Journal of the American Statistical Association. 1987;82(398):387–394. doi: 10.1080/01621459.1987.10478441. [DOI] [Google Scholar]
- 24.Lunceford JK, Davidian M. Stratification and weighting via the propensity score in estimation of causal treatment effects: a comparative study. Stat Med. 2004;23(19):2937–2960. doi: 10.1002/sim.1903. [DOI] [PubMed] [Google Scholar]
- 25.Pope GC, Ellis RP, Ash AS, et al. Principal inpatient diagnostic cost group model for Medicare risk adjustment. Health Care Financ Rev. 2000;21(3):93–118. [PMC free article] [PubMed] [Google Scholar]
- 26.Kautter J, Pope GC, Ingber M, et al. The HHS-HCC risk adjustment model for individual and small group markets under the Affordable Care Act. Medicare Medicaid Res Rev. 2014;4(3) doi: 10.5600/mmrr2014-004-03-a03. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Farag SS, Maharry K, Zhang MJ, et al. Comparison of reduced-intensity hematopoietic cell transplantation with chemotherapy in patients age 60-70 years with acute myelogenous leukemia in first remission. Biol Blood Marrow Transplant. 2011;17(12):1796–1803. doi: 10.1016/j.bbmt.2011.06.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Kurosawa S, Yamaguchi T, Uchida N, et al. Comparison of allogeneic hematopoietic cell transplantation and chemotherapy in elderly patients with non-M3 acute myelogenous leukemia in first complete remission. Biol Blood Marrow Transplant. 2011;17(3):401–411. doi: 10.1016/j.bbmt.2010.07.013. [DOI] [PubMed] [Google Scholar]
- 29.Armand P, Kim HT, Zhang MJ, et al. Classifying Cytogenetics in Patients with AML in Complete Remission Undergoing Allogeneic Transplantation: A CIBMTR Study. Biol Blood Marrow Transplant. 2012;18(2):280–288. doi: 10.1016/j.bbmt.2011.07.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Armand P, Gibson CJ, Cutler C, et al. A disease risk index for patients undergoing allogeneic stem cell transplantation. Blood. 2012;120(4):905–913. doi: 10.1182/blood-2012-03-418202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Slovak ML, Kopecky KJ, Cassileth PA, et al. Karyotypic analysis predicts outcome of preremission and postremission therapy in adult acute myeloid leukemia: a Southwest Oncology Group/Eastern Cooperative Oncology Group study. Blood. 2000;96(13):4075–4083. [PubMed] [Google Scholar]
- 32.Hashmi SK, Bredeson C, Duarte RF, et al. Biol Blood Marrow Transplant. Oct 2016, National Institutes of Health Blood and Marrow Transplant Late Effects Initiative: The Healthcare Delivery Working Group Report. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Shaw BE, Hahn T, Martin PJ, et al. National Institutes of Health Hematopoietic Cell Transplantation Late Effects Initiative: The Research Methodology and Study Design Working Group Report. Biol Blood Marrow Transplant. 2017;23(1):10–23. doi: 10.1016/j.bbmt.2016.08.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Khera N. Reporting and Grading Financial Toxicity. JCO. 2014;32(29):3337–3338. doi: 10.1200/JCO.2014.57.8740. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.