Mortality and healthcare costs in Medicare beneficiaries with AL amyloidosis

We used 2011–2016 data from the 5% Medicare Limited Data Set (LDS) for this retrospective, matched case–control study. The LDS contains a 5% random sample of Medicare beneficiaries, which is representative of the overall population covered by Medicare, the largest insurer of people aged 65 years and older in the USA. Data included information about year of birth, sex, race or ethnic origin, date of death (if deceased), reasons for Medicare entitlement and Medicare Advantage Plan (health maintenance organization) enrollment.

Medicare Part A and Part B files contain Medicare reimbursed claims from the following settings for fee-for-service enrollees: inpatient, outpatient, emergency room, physician office, skilled nursing facility, hospice, home health agency and durable medical equipment. Specifically, we used both beneficiary-level and claim-level data from the following individual files: Master Beneficiary Summary File, Inpatient File, Home Health Agency File, Outpatient File, Carrier (Physician/Supplier Part B Claims) File, Durable Medical Equipment File, Hospice File, Skilled Nursing Facility File.

We linked beneficiary information across Medicare files using encrypted identification numbers that are included on all patient- and claim-level files.

The Medicare claims data were created in accordance with the principles of the Health Insurance Portability and Accountability Act (HIPAA). The analysis of the LDS is exempt from a Privacy Board review.

The study population was comprised of the following two cohorts of Medicare beneficiaries: newly diagnosed AL amyloidosis beneficiaries, and matched, disease-free beneficiaries.

In the 2011–2016 data from the 5% Medicare LDS, 249 (33.3%) newly diagnosed AL amyloidosis Medicare beneficiaries were identified and matched to 747 (66.7%) disease-free controls. As shown in Table 1, 2566 patients had at least one inpatient or two outpatient claims for AL amyloidosis between 01/01/2012 and 12/31/2016, with the first diagnosis date occurring during the ID period (01/01/2012–12/31/2015). Of those 2566 patients, 2220 patients had no AL amyloidosis diagnosis in the baseline (1-year disease free) period; 460 patients received an AL amyloidosis treatment within 90 days of the index date; 279 patients were continuously enrolled during baseline in a qualified health plan and were continuously enrolled for at least 1 year following the index date (except those who died within 1 year). This left 249 patients with AL amyloidosis who were 66 or older in the study.

The mean (SD) age of newly diagnosed AL amyloidosis beneficiaries and disease-free controls was 76 (6.2) years; 41.8% of both newly diagnosed AL amyloidosis beneficiaries and disease-free controls were female; all US regions were represented; and the majority (90.4%) of both newly diagnosed AL amyloidosis beneficiaries and disease-free controls were white. Significantly more newly diagnosed AL amyloidosis beneficiaries were over 65 years old with end-stage renal disease compared with disease-free controls (6.0 vs 0.8%; p < 0.001). Enrollment length (from index date to the end of enrollment in the qualified health plan or study end [12/31/2016]) was significantly shorter in newly diagnosed AL amyloidosis beneficiaries than disease-free controls (mean [SD] 801 [488.0] days vs 933 [430.1] days; p < 0.001). Significantly more newly diagnosed AL amyloidosis beneficiaries died during follow-up than disease free controls (36.9 vs 15.3%; p < 0.001); and significantly more newly diagnosed AL amyloidosis beneficiaries died less than 1 year (≤364 days) after the index date than disease free controls (19.7 vs 5.4%; p < 0.001; Table 2).

Individuals with AL amyloidosis had a substantial burden of comorbidities or concurrent conditions, as determined by having at least one claim with a relevant ICD-9/ICD-10 code. The mean (SD) of CCI in AL amyloidosis patients was 5.0 (3.6) compared with 2.1 (2.5) in disease-free controls. Common individual comorbidities in the CCI in AL amyloidosis patients were congestive heart failure (36.1% in AL amyloidosis patients compared with 11.2% in controls), cerebrovascular disease (27.7%; 16.1%), chronic pulmonary disease (37.8%; 22.5%), moderate or severe liver disease (29.7%; 10.2%), renal disease (35.7%; 13.8%), diabetes without chronic complications (42.6%; 29.2%) and malignancy/lymphoma/leukemia (35.7%; 15.1%). Furthermore, hypothyroidism (cases: 32.1%; controls: 22.2%) and hyperlipidemia (77.5%; 65.6%) were common among newly diagnosed AL amyloidosis beneficiaries (Table 3).

Based on the Kaplan–Meier survival analysis, 3 months after the index date, 7.6% of newly diagnosed AL amyloidosis beneficiaries had died compared with 1.2% of disease-free controls. 6 months after the index date, 13.7% of newly diagnosed AL amyloidosis beneficiaries had died compared with 2.7% of disease-free controls. One year (≤365 days) after the index date, 19.7% of newly diagnosed AL amyloidosis beneficiaries had died compared with 5.5% of disease-free controls. 18 months after the index date, 26.6 and 8.2% of newly diagnosed AL amyloidosis beneficiaries and disease-free controls had died, respectively. 2 years after the index date, 30.6 and 11.8% of newly diagnosed AL amyloidosis beneficiaries and disease-free controls had died, respectively (all p < 0.001; Figure 2; Table 4).

Mean (SD) total annual all-cause Part A and Part B healthcare costs in the 1-year follow-up period were statistically significantly higher (p < 0.001) in newly diagnosed AL amyloidosis beneficiaries than in disease-free controls ($71,040 [65,766] vs $13,722 [27,493]). The majority of total costs among newly diagnosed AL amyloidosis beneficiaries accrued from inpatient hospital costs ($28,126 [40,409]) and non-ED outpatient service costs (mean [SD] $37,137 [37,363]; Table 5).

Death within 1 year after the index date was almost four-times higher in newly diagnosed AL amyloidosis beneficiaries than disease-free beneficiaries. To date, no studies have provided nationally representative data on survival in Medicare beneficiaries with newly diagnosed AL amyloidosis. However, several European and clinical studies have examined survival in these patients. For example, in a Swedish population-based study that used the nationwide Swedish Patient Registry, Weiss et al. (2016) found that 30% of patients diagnosed with AL amyloidosis between 2010 and 2013 died within 1 year of diagnosis; and 39% of patients died within two years [6]. A study by Muchtar et al. on patients with AL amyloidosis diagnosed and seen at Mayo Clinic between 2010 and 2014 found a 2-year death rate of 40% [18]. These death rates are higher than those in our analysis likely due to differences in patient populations and data used, as patients seen at the Mayo Clinic are probably sicker than average Medicare patients with AL amyloidosis.

The mortality rate found in our study may be lower than those found in other studies because patients whose disease is so far progressed that they do not receive treatment are excluded from our study.

The incremental cost of AL amyloidosis is very high. In the 1-year follow-up period, healthcare costs were more than five times higher in AL amyloidosis beneficiaries than in disease-free beneficiaries. To date, no studies have provided nationally representative data on healthcare costs in Medicare beneficiaries with newly diagnosed AL amyloidosis. However, in a study by Hari et al. that used administrative claims data on adult patients with relapsed/refractory AL amyloidosis, total average healthcare costs in the 1 year following AL amyloidosis relapse or refractory state were $139,143. The authors state that these costs were similar to those for newly diagnosed AL amyloidosis patients, but the exact cost is not provided [19]. Patients in the Hari et al. study differed from those included in our study because patients with relapsed/refractory AL amyloidosis lived long enough to receive a second line of treatment. In comparison, Quock et al. performed a study on newly diagnosed AL amyloidosis patients ≥18 years old using Truven MarketScan^® Commercial and Medicare Supplement Databases (Truven Health Analytics, MI, USA) and found a mean total cost of $122,180 in the 1-year post-diagnosis period [10]. The average cost found in this previous study may be higher than what we found in the current study because we were unable to include costs associated with prescription drugs covered by Medicare Part D in the current study due to lack of access to these data. For example, in the Quock et al. study, mean total outpatient pharmacy costs for all prevalent AL amyloidosis patients between 2007 and 2015 were $13,054; if these pharmacy costs were not included in this previous study, mean total costs would have been similar to those found in the current study [10].

This study has limitations. First, there is no diagnosis code in ICD-9-CM or ICD-10-CM specific to AL amyloidosis and no generally accepted or clinically validated method for identifying this condition using health insurance claims data; the codes for this study were selected with clinical expert input to eliminate as many non-AL amyloid patients as possible (e.g., by excluding 277.31, familial Mediterranean fever and E85.3, secondary systemic amyloidosis). We further required patients to have received treatment consistent with expert recommendations for AL amyloidosis. This requirement would be expected to decrease the sensitivity but increase the specificity of our identification algorithm. Nonetheless, patients with transthyretin-related hereditary amyloidosis would likely still have been included in our sample. We are planning to validate the algorithm we used in a future study using data collected from medical records. Second, the administrative claims used in this study were collected for reimbursement purposes and the completeness and accuracy of medical coding is subject to data coding restrictions and data entry error. Third, presence of comorbidities was determined based upon having at least one claim with a relevant ICD-9/ICD-10 code, not a clinical diagnosis, so misclassification, diagnostic uncertainty or coding errors were possible. Fourth, the healthcare cost estimates may be underestimated as claims for patients who die outside of the hospital were not available, and services not covered by insurance or rendered ‘out-of-network’ were not included. Fifth, cost estimates include direct healthcare costs only, and do not take into account important indirect costs associated with caregiver burden, loss of productivity or reduced quality of life. Sixth, we did not match based upon clinical factors, such as cardiovascular disease, as the goal of our study is to describe the incremental burden of illness AL amyloidosis patients face compared with population-based, disease-free controls. We would expect that patients with amyloidosis would have many more comorbid illnesses than matched controls. If we matched by comorbidity, we would be unable to estimate the burden of the disease, as the comparison group would (because of matching) have an abnormally high burden of disease for their age. Lastly, due to lack of access to Part D data, patients treated with prescription drugs covered under Part D were not included in our study. However, we plan to replicate the study using the Medicare 100% data file, which includes Part D data, in the near future.