Is QRS ‘Micro’-Fragmentation an Effective Marker For Mortality?

ECG, blood pressure machine
ECG, blood pressure machine
A study of 3 populations with different clinical characteristics was conducted to determine mortality risk factors related to QRS “micro”-fragmentation.

QRS ‘micro’-fragmentation (QRS-mf) derived from data collected by standard 12-lead electrocardiogram (ECG) was found to be a powerful predictor for mortality among 3 populations of patients with differing clinical characteristics. These findings were published in the European Heart Journal.

ECG data were sourced from the EU-CERT-ICD, VA Washington, and Whitehall II studies. QRS ‘macro’-fragmentation (QRS-Mf) was evaluated using filtered median beat images of individual leads and defined as the additional QRS local maxima. QRS-mf was defined as the sum of the fourth, fifth, and sixth decomposition fractions of the QRS complex signal. Patient outcomes were related with QRS-mf.

Among a total study population of 7779 patients, 504 did not survive during the 5-year follow-up.

In all 3 cohorts, non-survivors had a QRS-mf cumulative frequency which differed significantly from survivors (Kolmogorov-Smirnov test [K-S] range, 1.724-3.606; all P £.0052). The probability of death despite implantable cardioverter-defibrillator protection was significantly associated with abnormal QRS-mf among prophylactic recipients (c2, 44.04; P <.0001) and patients with atrial fibrillation (c2, 5.562; P =.018).

A QRS-mf cutoff of 3.5% differentiated survival probabilities among disease subgroups of patients.

For each separate cohort, mortality was associated with the continuous risk factors of age (hazard ratio [HR] range, 1.029-1.072; all P £.005), QRS-mf (HR range, 1.367-1.555; all P £.024), and total cosine R to T (HR range, 1.004-1.013; all P £.038). Additional predictors not reaching significance among all 3 cohorts included heart rate and left ventricular ejection fraction (LVEF).

For dichotomized risk factors, mortality was associated with heart rate of greater than 75 bpm (HR range, 1.645-2.214; all P £.007), QRS-mf greater than 3.5% (HR range, 1.578-2.214; all P £.019), and aged older than 65 years (HR range, 1.542-2.097; all P £.025). Additional predictors not reaching significance among all cohorts were LVEF less than 25%, total cosine R to T greater than 110°, and QTc greater than 450 ms.

This study may have been limited by using the 3.5% QRS-mf cutoff, which was determined using healthy individuals. Additional study is needed to optimize QRS-mf cutoffs prospectively.

“The presented analyses confirm that QRS-μf is a new potent risk indicator available from objective analysis of standard 12-lead ECGs,” the study authors noted.  “[I]n a number of clinically defined sub-populations, including atrial fibrillation patients, we found this risk factor to be a predictor of mortality independent of several other previously established risk indices.”

Disclosure: Multiple authors declared affiliations with industry. Please refer to the original article for a full list of disclosures.


Hnatkova K, Andršová I, Novotný T, et al. QRS micro-fragmentation as a mortality predictor. Eur Heart J. Published online February 21, 2022. doi:10.1093/eurheartj/ehac085