ARIMA prediction model for mortality caused by Acute Myeloid Leukemia in the United States population up to the year 2050: A CDC wonder database analysis.

Abstract

Abstract Background: Acute Myeloid Leukemia is one of the leading causes of cancer-related mortality in the United States. In this mortality database study of death certificates, obtained from a data registry spanning from 1999 to 2024, we examined mortality trends using national mortality datasets among individuals suffering from Acute Myeloid Leukemia in the United States. Objective: In this retrospective study, our aim was to analyze these trends in mortality among US residents by demographic characteristics including gender, race/ethnicity, and geographic characteristics like urbanization and census region, and prediction of future years based on the existing data by Autoregressive Integrated Moving Average (ARIMA) forecasting modeling. Methods: The national mortality data from the multiple causes of death files in the CDC WONDER Database were queried by applying the ICD-10 codes as C92.0 for Acute Myeloid Leukemia to identify deaths among the US population from 1999 to 2024. Trends in age-adjusted mortality rate (AAMR) were assessed for age groups, gender, race/ethnicity, urbanization, and census region. Results were expressed as average annual percentage changes (AAPC), and 95% confidence intervals (CI). The data was analyzed on statistical software i.e., the Joinpoint regression by National Institutes of Health (NIH version 5.0.2). Results: There were a total of 255,680 deaths due to Acute Myeloid Leukemia in the United States from 1999 to 2024, with a crude mortality rate of 3.16 per 100,000 population, with 57% of them being males while 43% females. Mortality rates (AAMR) continued to trend up from 1999-2024 overall (AAPC: +0.43 [+0.24 to +0.57]). ARIMA predictive model shows a steady rise in the mortality rates approaching 2050 up to 3.04 per 100,000 population especially in the male gender (AAMR: 3.97), White race (AAMR: 3.42), Midwest region (AAMR: 3.55), medium-small metropolitan areas (AAMR: 3.48) and non-metropolitan areas (AAMR: 3.48). States in top percentile for mortality included Iowa, Vermont, Wisconsin, Florida, West Virginia, Maine and North Dakota. Conclusion: These findings provide valuable insights into age-adjusted mortality patterns among the United States population with Acute Myeloid Leukemia. Mortality rates are predicted to trend gradually upwards among the US population with certain demographic and geographic characteristics, such as males, White population, medium, small and non-metropolitan areas, and those belonging to the Midwest region, having higher future predictions of AAMR.