Abstract

This study highlighted the importance of the need to address observed demographic biases in patients diagnosed with systemic mastocytosis (SM) to ensure equity of trained prediction models. There was considerable variability in how often tryptase testing was performed, but this variability did not appear to translate into large differences in the frequency of SM diagnosis across practice sites. Preliminary findings suggest there is potentially a substantial set of patients with SM from which to learn an electronic health recording (HER) prediction model, which has the potential to detect SM-associated symptoms earlier, support providers in diagnosis of SM, and ultimately decrease SM morbidity and mortality. Preliminary findings suggest underutilization of high-sensitivity peripheral blood mutational testing (KIT D816V).