Background: The effects of left ventricular assist device (LVAD) therapy on quality of life (QoL) are typically assessed with a combination of heart failure (HF)-specific and generic QoL measures. A generic QoL measure is believed to be necessary to fully capture the QoL effects of LVADs, as non-cardiac factors, such as comorbidities or LVAD complications, may limit generic QoL despite improvement in HF QoL after LVAD. However, the frequency and drivers of this expected discrepancy between HF-specific and generic QoL measures after LVAD implant are not known.
Methods: Patients who underwent LVAD implantation as part of the INTERMACS registry from 2012-14 were assessed prior to and 6 months after LVAD with the Kansas City Cardiomyopathy Questionnaire (KCCQ) and EQ-5D Visual Analog Scale (VAS) to measure HF-specific and generic QoL, respectively. Scores range from 0-100, and ~5 points indicates a clinically relevant change for both measures. Changes in scores were standardized by dividing by the standard deviation of the corresponding baseline score and were categorized as: <0 (worsening), 0 to <0.5 (small improvement), 0.5 to <1 (moderate improvement), and ≥1 (large improvement). Discordance was defined as a ≥2-category difference in standardized changes. Among patients who reported improvement in HF-QoL, multivariable logistic regression was used to examine factors associated with KCCQ-VAS discordance.
Results: Among 1888 patients, the majority reported improvement in both HF-specific and generic QoL after LVAD, with mean changes in the KCCQ of 32.7 ± 25.0 and in the VAS of 27.6 ± 27.4. Among the 1539 patients (81.5%) with moderate/large improvement in KCCQ, 334 (21.7%) had discordant changes in generic QoL (i.e., VAS did not substantially increase despite improvement in KCCQ). In multivariable modeling, baseline VAS scores were the strongest predictor of future KCCQ-VAS discordance (OR 2.17 per +10 VAS, 95% CI 1.98-2.38). Post-LVAD complications were not associated with discordance.
Conclusion: In a multicenter cohort of patients undergoing LVAD implant, most patients had substantial improvements in both CHF-specific and generic QoL. Discordance occurred in only ~1/5 of patients and was primarily observed among patients who reported good generic QoL prior to their LVAD (despite low KCCQ). Contrary to our expectations, non-cardiac comorbidities and LVAD complications were not associated with discordance. These results support the continued use of the KCCQ to monitor QoL before and after LVAD implantation but also highlight potential limitations of the VAS in fully capturing the QoL effects from LVADs.