Introduction:
There has been interest in quality improvement programs based on screening of LV function in order to increase rates of adherence to guideline-based recommendations for implantable cardioverter defibrillators (ICDs) for primary prevention (PP) of sudden cardiac death.
Methods:
We queried the echocardiography database at a tertiary academic medical center for patients with LVEF ≤ 35% on a transthoracic echocardiogram (TTE) and analyzed clinical data from the electronic medical record. We then determined the proportion receiving ICDs, reasons ICDs were not implanted, mortality rates, and our risk score for death (JACC 2012; 60:1647-55).
Results:
Of the 406 patients reviewed with LVEF 35% or less, 22.9% had an ICD at the time of the TTE and 10.8% had an ICD implanted during a median follow-up of 4.0 years (ICD-DURING-FU). With respect to patients without ICDs, 6.2% (of total group) did not survive the index hospitalization, 16.4% did not meet primary prevention criteria due to life expectancy less than 1 year or other guideline-based criteria, and 23.4% had spontaneous LVEF improvement to above 35% (LVEF-SPONT-IMPROVED), while the remaining 20.0% met guideline-based criteria for a primary prevention ICD (NO-ICD-DESPITE-CRITERIA) (Table). The NO-ICD-DESPITE-CRITERIA group had a greater mortality (35.8%) than the LVEF-SPONT-IMPROVED group (21.1%; p=0.03) and the ICD-DURING-FU group (20.5%; trend with p=0.07). In a multivariable logistic model, LVEF-SPONT-IMPROVED (v. ICD-DURING-FU) patients had OR=0.50 for death (p=0.049) after adjustment for the risk score.
Conclusions:
In a large cross section of patients with TTEs and extended follow-up, 1 in 5 patients did not have ICDs implanted despite meeting standard PP criteria. Another significant group did not have ICDs implanted because they had improvement in LV function during follow-up, and these patients had a similar survival probability to patients who did have ICDs implanted during follow-up.