Long-term Mortality With CRT in the CARE-HF Trial
The patient characteristics, study methods, and results have been published in detail. In brief, the study was a randomized, open-label study to assess the effects of adding or not adding a CRT device without a defibrillator function to guideline-indicated pharmacological therapy. Major inclusion criteria were a ventricular ejection fraction (EF) ≤35%, a QRS duration ≥150 ms or QRS 120–149 ms associated with echocardiographic criteria for dyssynchrony, and the presence of New York Heart Association (NYHA) class III/IV symptoms at the screening visit despite the use of loop diuretics and neuroendocrine antagonists. Patients who were not in sinus rhythm were excluded. The study was approved by all relevant ethical bodies. All patients provided written informed consent prior to transmission of data to the central administrative office. Once eligibility had been verified and at least 24 h after their original consent, patients were verbally re-consented in order to reduce subsequent rates of refusal to accept their randomly assigned treatment group. Once re-consent was given, randomly assigned treatment was disclosed to the investigator and patient. All analyses were conducted on an intention-to-treat basis, with patients remaining in their assigned group regardless of subsequent treatment. Medtronic sponsored the study.
The first patient was enrolled in the CARE-HF study at the end of 2000; recruitment was complete by March 2003 and follow-up in September 2004. After reporting the study in 2005, investigators were asked to inform patients of the results, and encouraged, if they felt it appropriate, to implant a CRT device with or without a defibrillator function according to local practice and after discussion with the patient. Investigators were subsequently asked, in 2006, to consent patients for further follow-up through routine clinics and telephone contacts to ascertain survival and device status until September 2009. Only information on device implantation and NYHA class was requested at the time of re-consent.
The study was designed by the Steering Committee. Analyses were pre-specified and conducted according to the intention-to-treat principle by N.F. independently of the sponsor, which did not have access to the study database. Tests were considered significant at the 2.5% level on one-sided tests. The primary outcome for the extension phase was all-cause mortality alone. No adjustments were made for multiple testing. Time-to-event was described using Kaplan–Meier curves and analysed using Cox proportional hazards models. The NYHA class was included as the only covariate as randomization was stratified on this characteristic. The assumption of constant proportional hazards was assessed using an explanatory time-dependent variable. The potential heterogeneity of effect by subgroup was investigated using an interaction term in the Cox proportional hazards model.
Methods
The patient characteristics, study methods, and results have been published in detail. In brief, the study was a randomized, open-label study to assess the effects of adding or not adding a CRT device without a defibrillator function to guideline-indicated pharmacological therapy. Major inclusion criteria were a ventricular ejection fraction (EF) ≤35%, a QRS duration ≥150 ms or QRS 120–149 ms associated with echocardiographic criteria for dyssynchrony, and the presence of New York Heart Association (NYHA) class III/IV symptoms at the screening visit despite the use of loop diuretics and neuroendocrine antagonists. Patients who were not in sinus rhythm were excluded. The study was approved by all relevant ethical bodies. All patients provided written informed consent prior to transmission of data to the central administrative office. Once eligibility had been verified and at least 24 h after their original consent, patients were verbally re-consented in order to reduce subsequent rates of refusal to accept their randomly assigned treatment group. Once re-consent was given, randomly assigned treatment was disclosed to the investigator and patient. All analyses were conducted on an intention-to-treat basis, with patients remaining in their assigned group regardless of subsequent treatment. Medtronic sponsored the study.
The first patient was enrolled in the CARE-HF study at the end of 2000; recruitment was complete by March 2003 and follow-up in September 2004. After reporting the study in 2005, investigators were asked to inform patients of the results, and encouraged, if they felt it appropriate, to implant a CRT device with or without a defibrillator function according to local practice and after discussion with the patient. Investigators were subsequently asked, in 2006, to consent patients for further follow-up through routine clinics and telephone contacts to ascertain survival and device status until September 2009. Only information on device implantation and NYHA class was requested at the time of re-consent.
Statistical Analysis
The study was designed by the Steering Committee. Analyses were pre-specified and conducted according to the intention-to-treat principle by N.F. independently of the sponsor, which did not have access to the study database. Tests were considered significant at the 2.5% level on one-sided tests. The primary outcome for the extension phase was all-cause mortality alone. No adjustments were made for multiple testing. Time-to-event was described using Kaplan–Meier curves and analysed using Cox proportional hazards models. The NYHA class was included as the only covariate as randomization was stratified on this characteristic. The assumption of constant proportional hazards was assessed using an explanatory time-dependent variable. The potential heterogeneity of effect by subgroup was investigated using an interaction term in the Cox proportional hazards model.
SHARE
