LVEF Cutoffs Leave a Wide Clinical Gap in ICD Eligibility

The implantable cardioverter-defibrillator ICD has long sat at the center of primary prevention for sudden cardiac death. In routine practice, the device is typically reserved for patients with a left ventricular ejection fraction LVEF of 35% or less, a threshold that grew out of randomized trials showing survival benefit in that population. The problem is easy to state and difficult to solve: sudden cardiac arrest does not respect that boundary. Many patients who suffer malignant ventricular arrhythmias have far better systolic function than the patients who qualify for an ICD under current guideline logic.

This tension is the core clinical issue raised by the abstract. LVEF remains a useful marker of risk, but it is not a complete one. A patient can have a moderately preserved ejection fraction and still carry enough arrhythmic substrate to die suddenly, especially when myocardial scar, electrical instability, or other unmeasured factors are present. By contrast, some patients with more depressed ventricular function never experience ventricular tachyarrhythmias at all. The result is a screening framework that is biologically convenient but clinically incomplete.

Sudden cardiac death is not a binary event, and neither is arrhythmic risk.

Why the 35% Threshold Became Dominant

The 35% cutoff has endured because it is simple, reproducible, and tied to randomized evidence. Trials that established ICD benefit for primary prevention focused on patients with more substantial left ventricular dysfunction, and professional guidance followed that evidence base. In that setting, lower LVEF captured a group with high enough event rates to justify an invasive device and the procedural burdens that come with implantation.

Yet the threshold also reflects the limitations of the evidence that built it. A cutoff can guide treatment, but it can also compress a continuous risk spectrum into a yes-or-no decision. That approach works when the biomarker closely matches disease biology; it becomes less reliable when the biomarker is only one piece of a broader arrhythmic phenotype. The abstract makes clear that this is exactly the case for sudden cardiac arrest, where the risk curve extends across a range of ventricular function rather than stopping at 35%.

In practical terms, this means clinicians are often forced to choose between overinclusion and underprotection. Broadening ICD eligibility would inevitably expose more patients to device-related complications, inappropriate shocks, lead problems, and long-term follow-up demands. Keeping the current cutoff preserves specificity but leaves a substantial number of vulnerable patients outside the protection of a defibrillator.

The Bedside Problem Hidden by a Simple Number

The central challenge is that LVEF is a measure of pump performance, not a direct measure of electrical instability. A patient may have relatively preserved contractility and still possess a scarred or irritable myocardium capable of initiating ventricular fibrillation. Conversely, depressed LVEF often signals a vulnerable substrate, which is why it has remained such an effective trial-enrollment tool. But the abstract’s premise is that the strongest population-level predictor is not the same thing as the best individual-level selector.

That mismatch matters at the bedside. A clinician evaluating a patient after myocardial injury, cardiomyopathy, or another structural heart disease process may be reassured by an LVEF above 35%, even when the overall arrhythmic profile is troubling. Under current practice, those patients are frequently managed without an ICD unless another qualifying indication emerges. The abstract argues, at minimum, that this leaves open a major gap between guideline thresholds and the real-world biology of sudden death.

What the Abstract Establishes—and What It Does Not

The abstract does not present a new trial result, a comparator cohort, or a fresh risk model. It lays out the rationale for questioning a longstanding standard. That is still clinically important, because the logic of prevention depends on how accurately we identify high-risk patients before the first arrest occurs.

The authors’ framing also points toward the kind of evidence that would be needed to move beyond LVEF alone. A better selection strategy would have to identify risk across the full continuum of ventricular function, not only in the lowest stratum. Such a strategy might incorporate scar burden, electrocardiographic markers, ventricular remodeling patterns, or other measures of arrhythmic susceptibility. The abstract itself does not test those alternatives, but it makes the case that the current cutoff cannot fully capture who is truly at risk.

For now, the clinical takeaway is constrained but clear. ICD therapy remains well supported in patients with LVEF of 35% or less, yet the majority of sudden cardiac arrests occur outside that boundary. Any prevention strategy that relies too heavily on a single number will miss part of the population it is meant to protect. The unresolved issue is not whether the 35% cutoff has evidence behind it; it does. The unresolved issue is whether that evidence can carry the full weight of modern sudden death prevention without refinement.


Reporting by Dr. Aishwarya N., MBBS for Medicine World Foundation, with additional data provided by JAMA (Journal of the American Medical Association).