Rates and reasons for device-based guideline eligibility in patients with heart failure.

Background: Implantable cardioverter-defibrillators (ICDs) and cardiac resynchronization therapy (CRT) are recommended by guidelines for patients with heart failure (HF) meeting specific criteria. Uncertainty exists regarding estimates of device eligibility, related in part to the method of assessing for guideline nonadherence.

Objective: The aim of this study was to identify the rates of guideline eligibility and device utilization after accounting for reasons for not receiving an ICD or CRT.

Methods: Patients were identified from 2006 to 2011 in a tertiary Heart Function Clinic in Canada. The chart-level data were collected that would indicate guideline eligibility and nonadherence.

Results: A total of 762 patients with HF were included (mean age 66 years; 527 (69%) were males; median left ventricular ejection fraction 33%). Over follow-up, 331 patients (43%) were never guideline eligible whereas 431 (57%) were guideline eligible for a device. Yearly rates for ICD and CRT adherence in "guideline-eligible" patients ranged from 59% to 68% and from 66% to 81%, respectively. "Patient preference" was the most commonly documented reason for guideline nonadherence in eligible patients. After removal of patients with reasons for nonadherence, rates of ICD and CRT adherence in the "truly eligible" patients were found to be higher (70%-81% and 71%-88%, respectively) than those in guideline-eligible patients. Independent predictors of device nonadherence in truly eligible patients were age >75 years, QRS duration <120 ms, left ventricular ejection fraction <30%, and female sex.

Conclusion: Based on chart-level data, utilization rates of device-based therapies in patients with HF appear much higher than those of prior registry-based estimates. Given the importance of patient preferences for lack of device use, future quality-of-care metrics based on guideline adherence should capture detailed chart-level data and patient preferences.