Primary Prevention ICD in a Contemporary Underserved Population: Factors Associated With Underutilization and Mortality
Justin J. Song, Sarah Xu, Ariel Gidon, Carlos Matos, Mohammed Kallash, Samuel J. Apple, Girish Jayant, Rani Marhaba, Jacob Borck, Serena Zhang, Sonya Kalani, Yawen Yuan, Carlos J. Rodriguez, Luigi Di Biase, Nils A. Guttenplan, John D. Fisher, Kevin J. FerrickAbstract
Background
Registry studies from 2000–2010 demonstrated underutilization of primary prevention ICD therapy. As heart failure therapies have reduced sudden cardiac death, investigators have questioned guidelines derived from trials conducted two decades prior, particularly in non‐ischemic cardiomyopathy (NICM) and elderly subgroups. We characterized contemporary referral determinants, decision‐making processes, and outcomes of primary prevention ICD therapy in a diverse, under‐represented population.
Methods
In this retrospective single‐center cohort study, we reviewed records of all patients meeting Class I criteria for primary prevention ICD in 2019. Multivariable logistic regression identified factors associated with ICD referral, and time‐varying Cox models compared all‐cause mortality by implantation status through 2022.
Results
Of 559 eligible patients, 345 (62%) were referred and 227 (41%) underwent implantation. Referral was associated with male gender, care coordination by a cardiologist rather than a generalist, and lower social vulnerability index (SVI) score. No decision‐making rationale was documented for a plurality of non‐referred and non‐implanted patients; documented rationales identified potential intervention targets. ICD implantation was not associated with all‐cause mortality (aHR 0.74, 95% CI 0.49–1.12). In exploratory analyses stratified by cardiomyopathy type, implantation was associated with lower mortality in ischemic cardiomyopathy (aHR 0.51, 95% CI 0.27–0.95) but not in NICM (aHR 1.11, 95% CI 0.63–1.95).
Conclusion
Many eligible patients are never referred for ICD therapy. Female gender, generalist care coordination, and high SVI were associated with non‐referral. The exploratory association of implantation with lower mortality in ischemic cardiomyopathy was not significant in propensity score overlap‐weighted analyses. The NICM estimate was imprecise.