NYC H+H/Bellevue & NYU Grossman School of Medicine New York, New York, United States
Background/Case Studies: Blood transfusion is a life-saving therapy, and it is critical to ensure that all patients are treated equitably in their access to this care. However, some patients face difficulties in accessing healthcare or even face bias within healthcare organizations. The decision to transfuse is often predicated on a patient’s hemoglobin concentration – differences in transfusion threshold may indicate specific difficulties faced by patient populations or biases in transfusing certain patients. The municipal public health system maintains records on patient transfusions and their pre-transfusion hemoglobin concentrations. Socioeconomic determinants of health are also included in the patient’s medical record demographics. We investigate the differences in transfusions between patient populations via hemoglobin, a quantifiable and unbiased variable.
Study
Design/Methods: Transfusion orders for public hospital system were collected for the years 2020 and 2024 as well as patients’ pre-transfusion hemoglobin concentrations, if available. The hemoglobin levels then stratified by the following criteria: domicile status, sex, and whether English was the patient’s preferred language (Equity Factors). We ran a Two-Way ANOVA using the Year and Equity Factors as independent variables and the pre-transfusion hemoglobin as the dependent variable, then performed a Tukey’s Post-Hoc Multiple Comparisons test to identify significant differences and interaction effects.
Results/Findings: Compared to 2020, in 2024 the pre-transfusion hemoglobin was lower for all patients, females, English-preferring patients and higher for Non-English preferring patients. Within 2020, the pre-transfusion hemoglobin concentrations were higher for men than for women and were lower for Non-English preferring versus English-preferring patients. In 2024, those differences remained, but also the pre-transfusion hemoglobin concentration was lower for non-domiciled patients compared to domiciled patients. The results are shown in Figure 1. Conclusions: These results suggest that non-domiciled patients present with a low average and large variation in their pre-transfusion hemoglobin levels, possibly due to their insecure housing situation. Additionally, the threshold to transfuse women appears to have lowered from 2020 to 2024 while men have remained at nearly the exact same pre-transfusion concentrations. In both 2020 and 2024, the pre-transfusion hemoglobin was lower in women compared to men, likely reflecting known variance. Curiously, patients who prefer English are transfused at higher hemoglobin levels than patients who are non-English preferring, possibly reflecting a pre-hospital barrier or difficulty in communicating with providers.
