The New England Journal of Medicine April 14, 2005

The Effect of Cardiac Resynchronization on Morbidity and Mortality in Heart Failure (CARE-HF)

John G.F. Cleland, Jean-Claude Daubert, Erland Erdmann, et al. (CARE-HF Study Investigators)

Bottom Line

In patients with severe heart failure, left ventricular systolic dysfunction, and cardiac dyssynchrony, the addition of cardiac resynchronization therapy (CRT) to optimal medical management significantly reduced the risk of all-cause mortality and major cardiovascular hospitalization.

Key Findings

1. The primary composite endpoint (death from any cause or unplanned hospitalization for a major cardiovascular event) occurred in 39% (159/409) of the CRT group compared to 55% (224/404) of the medical-therapy group (Hazard Ratio 0.63; 95% CI 0.51 to 0.77; P<0.001) [3.2.1].
2. All-cause mortality was significantly lower in the CRT group, occurring in 20% (82/409) of patients compared to 30% (120/404) in the medical-therapy group (Hazard Ratio 0.64; 95% CI 0.48 to 0.85; P<0.002).
3. CRT resulted in significant favorable reverse remodeling, including reductions in interventricular mechanical delay and end-systolic volume index, and increases in left ventricular ejection fraction (P<0.01 for all comparisons).
4. Patients receiving CRT experienced significant improvements in NYHA class and quality of life as assessed by the Minnesota Living with Heart Failure questionnaire.

Study Design

Design
RCT
Open-Label
Sample
813
Patients
Duration
29.4 mo
Median
Setting
Multicenter, Europe
Population Patients with NYHA class III or IV heart failure, LVEF ≤35%, and cardiac dyssynchrony (QRS ≥120 ms) despite receiving optimal pharmacological therapy.
Intervention Cardiac resynchronization therapy (CRT-P, biventricular pacing without a defibrillator) in addition to standard medical therapy.
Comparator Standard medical therapy alone.
Outcome Composite of death from any cause or an unplanned hospitalization for a major cardiovascular event.

Study Limitations

The open-label study design inherently carries a risk of performance or detection bias, though objective hard endpoints like all-cause mortality mitigate this risk.
The trial utilized CRT-pacemakers (CRT-P) rather than CRT-defibrillators (CRT-D), so the combined or independent survival benefit of simultaneous ICD therapy could not be addressed.
Racial and ethnic diversity was limited, and the proportion of female participants was relatively low (26%).
Echocardiographic criteria used for enrollment (especially for borderline QRS duration 120-149 ms) can be operator-dependent and challenging to standardize across all centers.

Clinical Significance

CARE-HF established biventricular pacing as a fundamentally life-saving intervention. Prior to this trial, CRT was known to improve symptoms and reduce hospitalizations, but CARE-HF definitively proved that CRT alone (without a backup defibrillator) yields a dramatic 36% relative risk reduction in all-cause mortality. This cemented CRT as a foundational pillar in the treatment algorithms for patients with HFrEF and wide QRS complexes.

Historical Context

During the late 1990s and early 2000s, trials like MUSTIC and MIRACLE demonstrated that CRT could improve functional capacity, quality of life, and structural remodeling in dyssynchronous heart failure. The COMPANION trial (2004) showed that CRT with or without an ICD reduced the composite of death or hospitalization, but the reduction in mortality for CRT-P alone narrowly missed statistical significance (P=0.06). CARE-HF (2005) was the landmark study that resolved this ambiguity, demonstrating a robust and definitive overall survival benefit for CRT-P, which revolutionized heart failure device therapy guidelines globally.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

What is the pathophysiological concept of 'cardiac dyssynchrony' in heart failure, as evidenced by a widened QRS complex, and how does biventricular pacing physiologically improve cardiac output in these patients?

Key Response

In patients with severe heart failure, a widened QRS (often a Left Bundle Branch Block) indicates delayed left ventricular activation. This electrical delay causes mechanical dyssynchrony: the septum contracts before the lateral wall, causing inefficient pumping, increased end-systolic volume, and worsened mitral regurgitation. Cardiac Resynchronization Therapy (CRT) uses a right ventricular lead and a coronary sinus lead (stimulating the LV lateral wall) to restore simultaneous contraction of both ventricles, directly improving stroke volume and reducing functional mitral regurgitation.

Resident
Resident

Based on the CARE-HF trial and subsequent clinical practice, what are the core clinical and electrocardiographic criteria a heart failure patient must meet to be considered for CRT?

Key Response

To qualify for CRT based on the foundational CARE-HF criteria, a patient typically must have severe symptomatic heart failure (NYHA class III or ambulatory IV) despite optimal guideline-directed medical therapy, severe left ventricular systolic dysfunction (LVEF <= 35%), and evidence of electrical dyssynchrony (QRS duration >= 120 ms, though current guidelines strongly favor >= 150 ms with LBBB morphology). Recognizing these criteria is essential for appropriate patient referral and ward management.

Fellow
Fellow

The CARE-HF trial utilized CRT-P (pacemaker only) without a defibrillator backup, demonstrating a mortality benefit. How does this compare to the COMPANION trial, and how does it complicate the modern EP decision of whether to implant a CRT-P versus a CRT-D in patients with non-ischemic cardiomyopathy?

Key Response

COMPANION evaluated both CRT-P and CRT-D, but CARE-HF was landmark because it proved that the pacing component alone (CRT-P) significantly reduced all-cause mortality, primarily by reversing heart failure progression rather than just treating sudden cardiac death (SCD). For EP fellows, deciding between CRT-P and CRT-D in non-ischemic cardiomyopathy is highly nuanced today (especially post-DANISH trial), requiring consideration of patient age, comorbidities, and competing risks of non-SCD mortality.

Attending
Attending

Approximately 30% of patients who meet the CARE-HF criteria and receive a CRT device are 'non-responders.' As an attending cardiologist, what is your systematic approach to evaluating and managing a patient who fails to improve clinically post-implantation?

Key Response

Managing CRT non-responders requires a step-wise approach: 1) Verifying that the percentage of biventricular pacing is >98% (frequent PVCs or atrial fibrillation with rapid ventricular response can inhibit pacing), 2) Checking device interrogation for LV lead dislodgement or high thresholds, 3) Optimizing AV and VV delays via echocardiography to maximize diastolic filling, and 4) Re-optimizing neurohormonal medical therapy. High-yield teaching focuses on the fact that the device only works if it is actively pacing near 100% of the time.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

CARE-HF was an open-label trial regarding the intervention, as patients and treating physicians were aware of the device implantation. How does the absence of a sham-control arm in device trials bias secondary endpoints like NYHA class or quality of life, and what statistical or methodological safeguards were used to protect the primary endpoint?

Key Response

The lack of a sham procedure introduces severe placebo effects for subjective endpoints (e.g., 6-minute walk test, NYHA class, QoL scores). To protect the validity of the primary endpoint (all-cause mortality and major cardiovascular hospitalization), CARE-HF utilized a blinded Clinical Events Committee (CEC) to adjudicate hospitalizations, and all-cause mortality provides a hard, objective endpoint that is highly resistant to performance or detection bias.

Journal Editor
Journal Editor

The CARE-HF trial was terminated early by the Data and Safety Monitoring Board (DSMB) due to a highly significant mortality benefit. As an editor, what are the inherent threats to validity when a trial is stopped early for benefit, and how does it impact the confidence interval of the estimated treatment effect?

Key Response

Trials stopped early for benefit often suffer from 'random high' bias, where the treatment effect is systematically overestimated because stopping rules are triggered at peaks of random statistical fluctuation. A rigorous peer reviewer would flag that while the outcome is clearly positive, the reported hazard ratio might exaggerate the true clinical benefit, and long-term safety data might be truncated, demanding cautious interpretation of the effect size.

Guideline Committee
Guideline Committee

CARE-HF included patients with a QRS >= 120 ms. However, recent ACC/AHA/HFSA and ESC Heart Failure guidelines have modified the Class I recommendation for CRT. How has the evidence evolved since CARE-HF to justify current guidelines, specifically regarding QRS morphology and duration?

Key Response

While CARE-HF established the foundational benefit of CRT in QRS >= 120 ms, subsequent trials (like EchoCRT) and subgroup meta-analyses demonstrated that the benefit is overwhelmingly driven by patients with Left Bundle Branch Block (LBBB) morphology and a very wide QRS (>= 150 ms). Consequently, current guidelines have restricted the Class I recommendation to patients with LBBB and QRS >= 150 ms, downgrading non-LBBB or shorter QRS (120-149 ms) durations to Class IIa or IIb due to lack of consistent mortality benefit or potential harm.

Clinical Landscape

Noteworthy Related Trials

2004

COMPANION Trial

n = 1,520 · NEJM

Tested

CRT pacemaker (CRT-P) or CRT defibrillator (CRT-D)

Population

Patients with advanced heart failure (NYHA III/IV), LVEF <= 35%, and QRS >= 120 ms

Comparator

Optimal medical therapy alone

Endpoint

Composite of death from or hospitalization for any cause

Key result: Both CRT-P and CRT-D reduced the risk of the primary endpoint compared to medical therapy, with CRT-D significantly reducing the risk of death from any cause.
2009

MADIT-CRT Trial

n = 1,820 · NEJM

Tested

CRT with ICD (CRT-D)

Population

Patients with ischemic or nonischemic cardiomyopathy, LVEF <= 30%, QRS >= 130 ms, and NYHA class I or II HF

Comparator

ICD alone

Endpoint

Death from any cause or nonfatal heart failure event

Key result: CRT-D significantly reduced the risk of heart-failure events in patients with mild heart failure symptoms compared to ICD alone.
2010

RAFT Trial

n = 1,798 · NEJM

Tested

CRT with ICD (CRT-D)

Population

Patients with NYHA class II or III heart failure, LVEF <= 30%, and QRS >= 120 ms

Comparator

ICD alone

Endpoint

Composite of death from any cause or hospitalization for heart failure

Key result: The addition of CRT to an ICD and optimal medical therapy significantly reduced rates of death and heart failure hospitalization.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis