Changes in the amplitude of electrocardiogram QRS complexes during follow-up of heart failure patients

Introduction: Changes in the electrocardiogram QRS amplitudes (ECGΔ) during follow-up of heart failure (HF) patients have not been clinically exploited heretofore. Methods: We examined ECGΔ during follow-up of HF patients by employing 42 triplets of ECGs, other laboratory and HF-related clinical data corresponding to clinical stability, worsening, and recovery from 37 HF patients. Results: The % changes (Δ%) in the summed QRS amplitude of all 12 leads (ΣQRS_12L), 6 precordial leads (ΣQRS _V1-V6), 6 limb leads (ΣQRS_6L), leads I+II (ΣQRS_{I + II}), and lead aVR were evaluated. Also relationships between the ECG variables and body weight (BW), percent body-fat, and B-type natriuretic peptide (BNP) were examined. The QRS amplitude(s) in all ECG variables decreased from clinical stability to worsening HF, and returned to baseline at recovery. During HF worsening, Δ% was highest in lead aVR (-15.3 ± 12.3%), followed by Δ% in ΣQRS_6L (-12.9 ± 10.1%) and ΣQRS_{I + II} (-12.1 ± 10.8%). At worsening HF and its recovery, Δ% in ΣQRS_6L correlated with Δ% in percent body-fat (r = 0.333, P = .031; r = 0.308, P = .047). At recovery, Δ% in each ECG variable correlated with Δ% in BW. Receiver operating characteristic (ROC) analysis showed that ≥16% reduction of ΣQRS_6L and ΣQRS_{I + II} discriminated between stable and worsening HF, with a sensitivity of 43% and 40%, and specificity of 98% for both. ECG variables from limb lead(s) had as good area under the curve (AUC) (0.78-0.84) as BNP (AUC: 0.88) for identifying worsening HF. Conclusions: Changes of the QRS amplitudes in ECGs are as useful for monitoring HF patients as BNP.