Trends In Racial Disparities In Ventricular Assist Device Implantation And Survival Rates Among Patients With Ambulatory Heart Failure

Introduction Allocation of ventricular assist devices (VAD) has been associated with patient race and a presumed higher severity of heart failure (HF) in African American patients. Data has demonstrated that patients with ambulatory HF (INTERMACS profiles 4-7) may benefit from VAD. It is unclear if there are racial/ethnic differences in VAD implantation rates and survival among patients with ambulatory HF. Hypothesis We hypothesized that VAD implantation rates among White patients with ambulatory HF were higher than rates among African American and Hispanic patients and that survival post VAD was similar across groups. Methods Using the INTERMACS database from 2012-2017, we examined annual census-adjusted VAD implantation rates using negative binomial models with quadratic effect of time among patients with ambulatory HF by race and ethnicity. Survival was evaluated via Kaplan-Meier estimates as well as Cox models adjusted for demographics, device strategy, comorbidities, and laboratory values and an interaction of time with race/ethnicity to account for non-proportionality of the hazards. Results Rates of implantation peaked between 2013-2015 before declining in all populations (Figure 1A). From 2012-2017 implantation rates were higher for African American and White patients than Hispanic patients, but rates were similar for African American and White patients. Kaplan-Meier curves were significantly different among the three groups (log rank test, p-value=0.0067), with estimated survival among African American patients higher than among White patients by nine months (9-month survival: African American 92% [95% CI:88-94%]; White 86% [95% CI:84-87%]. Low sample size for Hispanic patients resulted in imprecise survival estimates (9-month survival 86% [95% CI:78-91%]. Cox models demonstrated that Hazard Ratios for African American and Hispanic vs. White patients declined over time, with estimated Hazard Ratios less than 1 for both groups by 4 months (Figure 1B). Conclusions Rates of VAD implantation for patients with ambulatory HF are lower than expected, particularly for African American and Hispanic patients who have higher prevalence of HF than White patients. Estimated survival was higher among African American compared to White patients by nine months. Further investigation is needed to understand the drivers of lower VAD implantation rates in African American and Hispanic patients. Allocation of ventricular assist devices (VAD) has been associated with patient race and a presumed higher severity of heart failure (HF) in African American patients. Data has demonstrated that patients with ambulatory HF (INTERMACS profiles 4-7) may benefit from VAD. It is unclear if there are racial/ethnic differences in VAD implantation rates and survival among patients with ambulatory HF. We hypothesized that VAD implantation rates among White patients with ambulatory HF were higher than rates among African American and Hispanic patients and that survival post VAD was similar across groups. Using the INTERMACS database from 2012-2017, we examined annual census-adjusted VAD implantation rates using negative binomial models with quadratic effect of time among patients with ambulatory HF by race and ethnicity. Survival was evaluated via Kaplan-Meier estimates as well as Cox models adjusted for demographics, device strategy, comorbidities, and laboratory values and an interaction of time with race/ethnicity to account for non-proportionality of the hazards. Rates of implantation peaked between 2013-2015 before declining in all populations (Figure 1A). From 2012-2017 implantation rates were higher for African American and White patients than Hispanic patients, but rates were similar for African American and White patients. Kaplan-Meier curves were significantly different among the three groups (log rank test, p-value=0.0067), with estimated survival among African American patients higher than among White patients by nine months (9-month survival: African American 92% [95% CI:88-94%]; White 86% [95% CI:84-87%]. Low sample size for Hispanic patients resulted in imprecise survival estimates (9-month survival 86% [95% CI:78-91%]. Cox models demonstrated that Hazard Ratios for African American and Hispanic vs. White patients declined over time, with estimated Hazard Ratios less than 1 for both groups by 4 months (Figure 1B). Rates of VAD implantation for patients with ambulatory HF are lower than expected, particularly for African American and Hispanic patients who have higher prevalence of HF than White patients. Estimated survival was higher among African American compared to White patients by nine months. Further investigation is needed to understand the drivers of lower VAD implantation rates in African American and Hispanic patients.