Societal Costs of Sickle Cell Disease in the United States

Introduction Sickle cell disease (SCD) is a lifelong and costly chronic disease characterized by hemolytic anemia, pain crisis, and multi-end organ damage. Published estimates of SCD prevalence in the United States (US) range from approximately 85,000 to 100,000 people, most of whom are African American or Hispanic American. Individuals with SCD on average live two to three decades less than the general US population (Piel et al 2017). They also have markedly impaired patient-reported health-related quality-of-life (HRQOL) due in part to fatigue, pain, and impaired physical functioning, which leads to a significant reduction in work productivity. However, there are limited data available on the societal costs of SCD, such as lost lifetime earnings, which may lead to an underestimate of the true impact of this disease in a vulnerable population. Objectives We developed a simulation model to estimate the differences in life expectancy measured in years, quality-adjusted life-years (QALY) and income lost due to reduced life expectancy. Results were compared between a population of patients born with SCD in the US and a sex- and race-matched US population born without SCD and to the general US population. Methods To build the model, we (1) generated a Poisson regression from published birth and mortality estimates for SCD supplemented with data from the Centers for Disease Control (CDC) Multiple Cause of Death database to create age-specific life tables for a population of individuals with SCD (SCD population); (2) used published life tables from the CDC to develop age-specific death rates for a population without SCD (non-SCD population); (3) incorporated published utility weights for SCD adolescents and adults, and for the US general population to estimate the impact of the disease on HRQOL; (4) used US Bureau of Labor Statistics Supplemental Survey of Annual Personal Income data to calculate the expected annual personal income based on age, race, and gender; (5) built a cohort simulation model using R (version 3.4.2) to estimate the life expectancy, QALYs, and lost income for the SCD population compared to the non-SCD population, and the US general population. All analyses used Monte Carlo sampling to characterize uncertainty. Results We estimated that there would be 1,950 newborns with SCD born in the US annually. The projected life expectancy at birth is 54 years for the SCD population compared with 76 years for the age- and race-matched non-SCD populatio