Electrocardiographic characteristics for the prediction of under-sensing in implantable loop recorders

Under-sensing (US) in implantable loop recorders (ILRs) interferes with the accurate diagnosis of arrhythmia, but there is little information available on the details of US of ILRs. The aim of this study was to clarify the frequency of US in patients with ILRs and to investigate the predictors of US in ILRs prior to implantation.

We studied 46 consecutive patients implanted ILR. During the mean follow-up period of 499 ± 363 days, 15 events of US were observed in five patients. There were no significant differences in patient characteristics between patients with and without US. In standard 12-lead electrocardiogram (ECG), QRS complex amplitude in anterolateral chest leads (V2 to V5) were significantly lower in patients with than without US (V2: 0.88 [0.66, 1.22] mV vs. 1.67 [1.23, 2.29] mV, p = .010 V3: 1.25 [1.00, 1.26] mV vs. 1.90 [1.41, 2.29] mV, p = .013; V4: 1.14 [0.96, 1.38] mV vs. 1.93 [1.65, 2.64] mV, p = .023; V5: 0.57 [0.50, 0.75] mV vs. 1.60 [1.20, 1.98] mV, p = .011, respectively). ROC curve analysis showed that cut-off values of 1.30 mV of QRS complex amplitude in V2, 1.26 mV of that in V3, and 0.75 mV of that in V5 had moderate accuracy for predicting US (V2: sensitivity 68%, specificity 100%, area under the curve [AUC] 0.86; V3: sensitivity 85%, specificity 80%, AUC 0.85; V5: sensitivity 98%, specificity 80%, AUC 0.85, respectively).

US was observed in 10.9% patients with an ILR. QRS complex amplitude in anterolateral chest leads (V2 to V5) on ECG might be useful for predicting US in patients with ILRs.