Relation of Sociodemographic Factors With Primary Cause of Hospitalization Among Patients With Left Ventricular Assist Devices (from the National Inpatient Sample 2012 to 2017)

Left ventricular assist devices (LVADs) are an established intervention for end-stage heart failure (HF). Rehospitalization for serious complications remains common during the continuous-flow LVAD era. Whether sociodemographic factors are associated with differences in the frequency of severe complications leading to hospitalization remains unclear. Using data from the National Inpatient Sample, we identified all hospitalizations from 2012 to 2017 of adults aged ≥18 years with previous LVAD placement. We categorized the primary cause of hospitalizations into key adverse diagnoses, including bleeding, HF, arrhythmias, LVAD complications, stroke, and a composite of device-related infection or sepsis. We assessed the association of sociodemographic markers with primary diagnoses using modified Poisson regression. We identified 62,630 hospitalizations during the study period (41% aged ≥65, 77% men, 26% Black, 5% Hispanic). Bleeding (18%), infections (15%), and HF (15%) were the most common primary diagnoses. In the multivariable analyses, gastrointestinal bleeding was more likely among older adults (relative risk [RR] 95% confidence interval [CI] 4.69 [3.57 to 6.16]; age ≥65 vs 18 to 44 years), among Black than White patients (RR 95% CI 1.17 [1.04 to 1.32]), and less likely for the highest income quartile than the lowest (RR 95% CI 0.79 [0.69 to 0.91]). Device-related infection/sepsis was also less likely for patients with higher income (RR 95% CI 0.80 [0.67 to 0.96]). Ventricular arrhythmias were less frequent diagnoses for women than men (RR 95% CI 0.59 [0.46 to 0.75]). LVADs complications were less likely in older adults than younger adults (RR 95% CI 0.70 [0.50 to 0.98]). In conclusion, after LVAD implantation, the frequency in which specific adverse events are the primary cause of rehospitalization varies significantly by sociodemographic factors. Further study is needed to determine if there are opportunities for targeted preventive measures based on sociodemographic markers.