Association of global and disease-specific health status with outcomes following continuous-flow left ventricular assist device implantation

Association of global and disease-specific health status with outcomes following continuous-flow left ventricular assist device implantation

BMC Cardiovascular Disorders, March 2017

28288574

Kelsey M. Flint, John A. Spertus, Fengming Tang, Philip Jones, Timothy J. Fendler, Larry A. Allen

The prognostic value of heart failure specific and global health status before and after left ventricular assist device (LVAD) implantation in the usual care setting is not well studied. We included 3,836 continuous-flow LVAD patients in the INTERMACS registry. Health status was measured pre-operatively and 3 months post-LVAD using the Kansas City Cardiomyopathy Questionnaire (KCCQ) and EuroQol visual analog scale (VAS). Primary outcomes were mortality/rehospitalization. Inverse propensity weighting was used to minimize bias from missing data. Pre-operative global and heart failure-specific health status were very poor: KCCQ median 34.6 (IQR 21.4-50.5); VAS median 43 (interquartile range (IQR) 25-65). Health status measures improved 3 months after LVAD placement: KCCQ median 69.3 (IQR 54.2-82.3); VAS median 75 (IQR 60-85). Pre-operative health status was not associated with death (unadjusted HR for lowest vs. highest score quartiles: 1.09 (0.85-1.41) KCCQ; 1.12 (0.85-1.49) VAS) or rehospitalization (unadjusted HR 0.83 (0.72-0.96) KCCQ; 0.99 (0.85-1.16) VAS). Three-month KCCQ was associated with mortality (unadjusted HR 2.17 (1.47-3.21); VAS was not (1.43 (0.94-2.17). Three-month KCCQ added incremental discriminatory value to the HeartMate II Risk Score for death (c-stat 0.60 to 0.66); VAS did not (c-stat 0.59 to 0.60). Three-month health status was associated with rehospitalization (unadjusted HR 1.31 (1.15-1.57) KCCQ; 1.24 (1.05-1.46) VAS), but did not add incremental discriminatory value (c-stat 0.52 to 0.55 and 0.54, respectively). These real-world data suggest that pre-operative health status has limited association with outcomes after LVAD. However, persistently low health status after surgery may independently signal higher risk for subsequent death. Further study is needed to determine the clinical utility of routinely collected health status data after LVAD implantation.