Examination of Racial and Gender Disparities in Length of Stay After Lower Extremity Bypass
John Axley, Zdenek Novak, Adam Beck, Benjamin Pearce, Marc Passman, Mark Patterson, Emily Spangler.
University of Alabama at Birmingham, Birmingham, AL, USA.
Objective: Racial disparities in length of stay (LOS) are seen in elective surgical procedures, irrespective of postop complications. Enhanced Recovery Programs (ERP) designed to decrease LOS have been able to mitigate these disparities in the general surgery population. We evaluate patient and perioperative factors from uncomplicated Vascular Quality Initiative Lower extremity bypass (VQI LEB) cases for disparities to identify modifiable targets for ERP pathways and more equitable care.
Methods: From the LEB VQI dataset from 2003-17, a cohort of 28,062 white(non-Hispanic) and black patients without in-hospital complications were assessed. Demographics were compared by parametric and nonparametric testing while multivariable regression was used to model LOS, including interaction between race and gender, with subsequent stratified analyses by gender.
Results: Almost all demographics differed significantly by race, with black patients being younger, more often female, current smokers, diabetic, with ESRD and CHF. Black patients less commonly had COPD, prior PCI/CABG, and were less commonly on preoperative anticoagulation, antiplatelet, or statin therapy. Black patients were less ambulatory(70.2% vs 77.6% ambulatory, and 24.1% vs 18.5% ambulatory with assistance(p<.001)), and had more severe disease indications(26.3% vs 22.3% rest pain, and 36.8% vs 30.2% tissue loss, p<.001, with baseline ABI< 0.4 in 24.9% vs 22.8%, p=.004). Procedural urgency(elective/urgent/emergent) did not differ. Black patients were more likely to have Medicaid or self pay insurance, and less likely to be discharged home. Black patients had greater preoperative vein mapping(63.0% vs 59.1%, p<.001), but less vein use(53.9% vs 59.1%, p<.001). Both median total and postoperative LOS differed significantly by race, most markedly in femoral-tibial bypasses (median total LOS[IQR] of 5[3-8] days in white vs 7[4-12] in black patients(p<.001), and postop LOS of 4[3-6] days in white vs 5[3-7] in black patients(p<.001)). Adjusted analyses demonstrated a significant interaction of race with gender in postoperative, but not total LOS. Variables besides race predictive of postoperative LOS are detailed by gender in Table 1.
Conclusions: In uncomplicated LEB, racial disparities with gender interactions were observed in postoperative, but not total LOS after multivariable adjustment. Modeling in both genders suggests anticoagulation management and preoperative ambulatory status may be modifiable targets for reducing LOS.
Back to 2019 ePosters