Abstract
Introduction:Population-based cancer registries, such as the National Cancer Institute Surveillance, Epidemiology and End Results (SEER) program and the North American Association of Central Cancer Registries (NAACCR) are the largest sources of information for cancer epidemiology and statistics. The most recent acute myeloid leukemia incidence estimate from SEER (2011) is 17.5 per 100,000 (N=7,245) among the US ≥65 year-old population; however, recent studies suggest these registries may underreport cancer rates due to reasons including sequencing of diagnoses and inpatient reporting requirements. For cancers such as myelodysplastic syndrome (MDS) and acute and chronic myeloid leukemia (AML & CML), this is concerning as they are more likely to occur after initial diagnosis of other cancers. A recent study independently calculated MDS, AML and CML cases from 2000-2005 using a Medicare claims-based algorithm and concluded that SEER and NAACCR failed to capture a substantial number of cases and the true incidence was 50-75% greater than reported (Cogle, et al., 2012). Updated AML epidemiology statistics outside of SEER and NAACCR have not been published.Objective:To employ a Medicare claims-based algorithm to estimate gender- and age-specific AML incidence and prevalence among the 2012 US Medicare fee-for-service (FFS) population.Methods:A retrospective analysis of claims using 2012 Centers for Medicare and Medicaid Services (CMS) data included an Institutional sample (100%) and random Non-Institutional Carrier samples (5%) which together represented the healthcare utilization of Medicare Part A & B (Medicare FFS) beneficiaries. AML diagnoses were identified using ICD-9 codes and AML treatments identified using HCPCS J Codes and ICD-9 infusion codes. Prevalent AML patients were defined as having ≥2 AML diagnoses associated with medical claims OR 1 AML medical claim and 1 AML treatment. A sub-population of all prevalent AML patients without historical AML diagnoses or treatments during the prior 2 years were identified as new (incident) AML patients.Analyses were computed by gender and two age-cohorts (<65 and ≥65 years old). Patients in the Institutional 100% dataset were considered census and no weighting was required but appropriate weights were used to project the 5% random carrier sample (<9% of AML patients) to the Medicare FFS population.Results:Of 34.2 million Medicare FFS beneficiaries, 15,976 had AML, a prevalence rate of 0.05% (Table). Most were ≥65 years old (N=11,936; 75%) and prevalence did not vary between age groups; however, women ≥65 years old had a significantly lower prevalence than men ≥65 years old (0.03% vs. 0.06%; z=31.2, p<.001) as men were nearly twice as likely to be diagnosed with AML (RR=1.86, 95% CI: 1.78, 1.95). There were no gender differences in incidence among younger patients (18.6 per 100,000 for men vs. 18.4 per 100,000 for women). A high proportion of AML patients were newly diagnosed (N=9,074; 57%).Conclusions:Our AML incidence estimate for the ≥65 year Medicare FFS cohort of 29.0 per 100,000 (N=7,582) is substantially higher than incidence estimate reported by SEER for this age group. As only 70-80% of the ≥65 year-old population is covered under Medicare FFS, the total number of ≥65 incident patients is likely higher than reported by SEER. Registries may be underreporting AML due to methodological differences. Furthermore, the 15,976 prevalent patients in Medicare FFS alone may be much higher than previously known. Claim-based algorithms may provide higher AML estimates than current SEER methodology. Further research should investigate claims data in the remaining ≥65 year-old population covered under Medicare Advantage and a younger, non-Medicare FFS population sample more representative of persons <65 years of age.TableOne-year Prevalence and Incidence Rates of AML in the Medicare FFS Population, 2012Population, N1-year AML Incident per 100,000, n (%)1-Year AML Prevalence, n (%)Overall34,216,0769,074 (26.5)15,976 (0.05)<65 years8,064,5661,492 (18.5)4,040 (0.05)≥65 years26,151,5107,582 (29.0)11,936 (0.05)Male15,329,0405,181 (33.8)8,854 (0.06)Female18,887,0363,893 (20.6)7,121 (0.04)Male, <65 years4,137,155770 (18.6)2,061 (0.05)Male, ≥65 years11,191,8854,410 (39.4)6,793 (0.06)Female, <65 years3,927,411722 (18.4)1,978 (0.05)Female, ≥65 years14,959,6253,171 (21.2)5,143 (0.03) DisclosuresTurbeville:Sunesis Pharmaceuticals, Inc.: Employment. Morrison:Sunesis Pharmaceuticals, Inc.: Employment.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have