Abstract 4143815: Structural, Hospital, and Clinical Factors Do Not Completely Explain Major Lower Limb Amputation Disparities among Patients Hospitalized with Peripheral Artery Disease

Background: Racial and ethnic disparities in major lower limb amputation (MLLA) in patients with peripheral artery disease (PAD) are well documented. Rural/urban differences have also been observed. Disparities in the prevalence of PAD, progression to chronic limb-threatening ischemia, and access to vascular centers and revascularization may explain observed differences. Research Question/Hypothesis: Are MLLA disparities still present after accounting for clinical presentation, hospital characteristics, and social drivers of health? We hypothesized that race, ethnicity and rurality differences in MLLA rates would disappear after accounting for these factors. Methods: We conducted a retrospective cohort study of 2017-19 PAD-related hospitalizations in adults ≥40 years old living in Florida, Georgia, Maryland, Mississippi, and New York using HCUP State Inpatient Databases, which include all community hospital admissions in each state, regardless of insurance status. Patients were excluded if they were not a resident or if they were admitted for lower limb trauma, malignancy, or non-vascular wounds. We used machine learning (LASSO) to predict the probability of revascularization and MLLA for each hospitalization, using clinical presentation, chronic conditions, hospitalizations and surgery within past year, hospital characteristics, and social drivers of health. We calculated observed-to-expected (O/E) ratios across race, ethnicity, and rurality. Bootstrapping (n=1000) was used to estimate confidence intervals. Results: 1,577,061 hospitalizations (990,152 patients) were included; there were 100,549 revascularizations (6.4%) and 21,233 MLLA (1.4%). No meaningful disparities were seen for revascularization. More than expected MLLA were performed in Black patients, both rural (O/E:1.26, 95% CI:0.98-1.54) and non-rural (O/E:1.12, 95% CI:1.09-1.15), and rural Hispanic patients (O/E:1.50, 95% CI:0.89-2.12), Fig . Conclusions: Even after accounting for clinical, hospital, and structural factors, MLLA disparities were observed among Black and rural Hispanic patients. These results suggest other factors, such as implicit bias in healthcare, play a role in the elevated limb amputation rates in minority and rural patients.