BACKGROUND:
Cardiac Resynchronization Therapy (CRT) is clinically effective and cost-effective, although some patients realize 6-Month (6M) improvements while others do not. Long-term differences in outcomes among short term “responders” and “non-responders” are not reported. We compared survival, burden of Heart Failure (HF) hospitalization, and Average Length of Stay (ALOS), by Clinical Composite Score (CCS) at 6M post-implant, which was used to assess short-term CRT response.
METHODS:
MIRACLE, MIRACLE-ICD, InSync III Marquis, PROSPECT and Adaptive CRT were pooled. Classification of responder status was made at 6M; patients who died post-implant and before CCS assessment at 6M were excluded. We adjusted for differences in baseline characteristics between CCS. HF hospitalizations before 6M were excluded from HF rate calculations. Hospitalizations resulting in death were excluded from the ALOS analysis. Mortality was assessed via Kaplan-Meier curves and Cox regression; HF hospitalization rates using Poisson regression with robust variance estimates; long-term extrapolations used Markov modelling.
RESULTS:
We identified 1,089 Improved (“I”, 67.94%), 235 Unchanged (“U”, 14.66%) and 279 Worsened (“W”, 17.4%) patients. 18M post-implant, 91.5% I patients were alive vs. 89.9% U and 69.9% W (Unadjusted HR I vs. U 0.68 (0.37-1.27), I vs. W 0.24 (0.15-0.37), and U vs. W 0.34 (0.18-0.65), Log-Rank p<0.001). I patients were projected to survive 8.007 years vs. 7.967 U and 1.815 W. Post-CCS assessment HF hospitalization rates in the first year were 0.13 for I, 0.27 for U and 0.90 for W (p<0.0001) leading to a lifetime projection of 1.04, 2.14 and 2.54 total hospitalizations per patient respectively. Adjusted ALOS was directionally favoring I (4.87, 8.06 and 8.38 respectively) but non-significant (p=0.14).
CONCLUSIONS:
Patients who improve or remain unchanged 6M after CRT implant are projected to live longer and consume fewer resources than patients who worsen.
Is cardiac resynchronization therapy cost-effective?
