Prevalence and Incidence of Primary Autoimmune Hemolytic Anemia and Cold Agglutinin Disease in the United States, 2016-2022: A Retrospective Study in Administrative Claims

Abstract

Background Autoimmune hemolytic anemia (AIHA) is a decompensated acquired hemolysis caused by autoantibodies that act against red blood cells leading to anemia. Cold agglutinin disease (CAD) is a rare subtype of AIHA mediated by classical complement pathway activation, characterized by chronic hemolysis, severe anemia, fatigue, weakness, dizziness, and circulatory problems. The aim of this study was to update the epidemiology of AIHA and CAD in the United States using administrative claims data that includes commercially insured patients, as well as those with Medicare Advantage, Medicare Fee for Service (FFS) and Managed Medicaid plans. Methods Administrative closed claims from Optum CDM database (commercially insured single-payer and Medicare Advantage), Inovalon MORE 2 Registry (commercially insured multi-payer, Medicare Advantage, and Managed Medicaid), and 100% Medicare FFS (Part A/B services) data were used for this analysis. AIHA cases were defined as patients with at least 2 claims with ICD-9/10 diagnosis codes for AIHA (283.0, D59.1, D59.10, D59.11, D59.12, D59.13, D59.19) at least 30 days apart. CAD cases were defined as patients with at least 2 claims with ICD-10 diagnosis codes for CAD (D59.12) or at least 1 claim with diagnosis code for AIHA followed by 1 claim with diagnosis code for CAD (D59.12) at least 30 days apart. Index date was assigned to the first observed claims with the AIHA/CAD diagnosis. Analyses included patients 18 years of age and above with at least 180 days of continuous enrollment in their health plan prior to index for the assessment of the exclusion criteria. Patients with secondary causes of AIHA/CAD (e.g., lymphoma, leukemia, myeloma, cytomegalovirus, or Epstein-Barr virus) were excluded. US census data from years 2016-2021 was used to standardize the incidence and prevalence estimates by age and sex. Point prevalence was calculated as of January 1 st of each year using an all-data lookback. Period prevalence was calculated using a one calendar year assessment period. Incidence estimated new cases in a given year among patients with no prior evidence of AIHA in all data lookback. For CAD, estimates were presented for years 2022/23 due to the recency of the ICD-10 code specific to CAD (active as of 1 October 2020). Results Age and sex standardized point prevalence of AIHA per 1,000,000 persons was approximately 210 in the Medicare FFS across years 2016-2021, 160 in the Optum claims across years 2016-2022, and 50 in the MORE 2 database across years 2016-2022. One-year prevalence per 1,000,000 persons was approximately 190 in the Medicare FFS, 70 in the Optum claims, and 50 in the MORE 2 database across study years. Incidence per 1,000,000 was 50-60 in the Medicare FFS, 20-30 in the Optum claims, and 15 in the MORE 2 database. While overall standardized estimates differed across databases, incidence and prevalence estimates were similar when stratified by age group, indicating that standardization did not fully account for the underlying differences in the composition and representativeness of the overall US population in the three databases. Incidence and prevalence were higher among females than males and increased with age as expected. Incidence per 1,000,000 for year 2021 ranged from roughly 10-80 among those under the age of 45 years to 30-80 among those over the age of 65 years across the three databases. Point prevalence per 1,000,000 for year 2021 had a broader range from 30 to 220 among those under the age of 45 years and from 120 to 360 among those over the age of 65 years. The standardized point prevalence per 1,000,000 for CAD was 31 in Optum claims (year 2023), 33 in Medicare FFS (year 2021), and 14 in MORE 2 (year 2021). Prevalence of CAD was higher among females and increased with age. Conclusion This multi-database analysis provides updated epidemiology estimates for AIHA and CAD in the US. Age and sex standardized point prevalence for AIHA ranged from 60 to 210 per 1,000,000 across the 3 databases, with higher prevalence in the Medicare FFS database which primarily captures population over the age of 65 and was lower in the two databases that capture younger age groups, such as Managed Medicaid and commercially insured populations. Similarly, incidence of AIHA ranged from 15 to 60 per 1,000,000. Point prevalence of CAD ranged from 14 to 33 per 1,000,000. These estimates are in line with those previously reported for North America and Western Europe.