Abstract
BACKGROUND: Utilization rates (URs) for implantable cardioverter-defibrillators (ICDs) for primary prevention of sudden cardiac death (PPSCD) are lacking in the community. OBJECTIVE: The purpose of this study was to establish the ICD UR in central Indiana. METHODS: A query run on 2 hospitals in a health information exchange database in Indianapolis identified patients between 2011 and 2012 with left ventricular ejection fraction (EF) ≤0.35. ICD eligibility and utilization were determined from chart review. RESULTS: We identified 1863 patients with at least 1 low EF study. Two cohorts were analyzed: 1672 patients without and 191 patients with International Classification of Diseases, Ninth Revision, Clinical Modification procedure code 37.94 for ICD placement. We manually reviewed a stratified (by hospital) random sample of 300 patients from the no-ICD procedure code cohort and found that 48 (16%) had no ICD but had class I indications for ICD. Eight of 300 (2.7%) actually had ICD implantation for PPSCD. Review of all 191 patients in the ICD procedure code cohort identified 70 with ICD implantation for PPSCD. The ICD UR (ratio between patients with ICD for PPSCD and all with indication) was 38% overall (95% confidence interval [CI] 28%-49%). URs were 48% for males (95% CI 34%-61%), 21% for females (95% CI 16%-26%, P = .0002 vs males), 40% for whites (95% CI 27%-53%), and 37% for blacks (95% CI 28%-46%, P = .66 vs whites). CONCLUSION: ICD UR is 38% among patients meeting class I indications, suggesting further opportunities for improving guideline compliance. This study also illustrates limitations in calculating ICD UR using large electronic repositories without hands-on chart review.