Abstract 9618: Real-World Effectiveness of ICDs Implanted During Unplanned Medicare Hospitalizations
Background: Benefits of ICDs for primary prevention are well established in ambulatory HF population.
Aim: To examine clinical effectiveness of primary ICDs in Medicare patients(pts) receiving the device during unplanned hospitalizations for HF or other acute conditions.
Methods: Linking data from the CMS/National Cardiovascular Data Registry (NCDR®) ICD Registry™, a national HF registry, and the Medicare files (04-09), we identified 23,111 hospitalized pts ≥ 66 yrs with EF ≤35% who were eligible for primary ICDs from the ICD registry (ICD recipients) or the HF registry (non-ICD recipients). To account for potential bias due to ICD recipients being healthier, we employed latency analysis conditioning on the 1st 6 months survival and adjusted for high-dimension propensity score (hdPS). Cox models were used to derive adjusted hazard ratios (HRs) for all-cause mortality and sudden cardiac death (SCD).
Results: The mean follow-up was 2.8 yrs. ICD recipients were younger, had lower EF, more prior cardiac hospitalizations, yet, non-ICD recipients had lower eGFR, more prior non-cardiac hospitalizations and non-CV comorbidities. ICD recipients had lower crude mortality than non-ICD recipients (Figure), but, the difference was most notable within the 1st 3-6 months of follow-up, likely reflecting selection of healthier pts for ICDs. Conditioning on the 1st 6-month survival and with hdPS adjustment, ICD use was not associated with a significant decrease in mortality (HR=0.91 95% CI=0.81, 1.00) or SCD (HR=0.95, 95% CI=0.78, 1.17). No significant treatment heterogeneity was observed among various demographic subgroups.
Conclusion: ICD use did not appear to be associated with a lower mortality risk among non-ambulatory Medicare patients receiving primary ICD during unplanned hospitalization for HF or other acute conditions after accounting for selection bias and confounding. The effectiveness of primary ICD therapy likely did not differ among demographic subgroups.
- © 2013 by American Heart Association, Inc.