Accuracy of the pacemaker event recorder versus Holter-ECG to detect both symptomatic and asymptomatic ventricular arrhythmias.

BACKGROUND: Although new pacemakers can register cardiac rhythm, few studies were performed evaluating their accuracy in diagnosing ventricular arrhythmias (VA). This study aimed to assess the correlation and agreement between the pacemaker's monitor and the ambulatory Holter in detecting VA.

METHODS AND RESULTS: We studied 129 patients with pacemakers, mean age 68.6 ± 19.1 years, 54.8% female. Once Holter monitoring was connected, the pacemakers' event counters were reset and clocks of both systems were synchronized to register electrocardiograms (ECG) simultaneously. Pacemakers were programmed to detect the lowest ventricular rate and lowest number of sequential beats allowed in their event monitors. After 72 hours, Holter and pacemakers records were analyzed. VA was defined in Holter and event monitor, respectively, as: isolated premature ventricular complexes: "PVC"; pairs: "couplets"; nonsustained ventricular tachycardia (NSVT): "triplets"-3 beats; "runs"-4-8 or > 8 beats, and high ventricular rates ("HVR")-3-4 beats. Spearman correlations evaluated whether pacemaker and Holter identified the same parameters. Intraclass correlation coefficients (ICCs) and respective 95% confidence intervals were calculated to assess the concordance between methods. The agreement between both systems was low, except for "triplet" and three beats NSVT (ICC = 0.984). The correlation for more than 10 PVC/h was moderate (Kappa = 0.483). When the pacemaker was programmed to detect HVR sequences of three beats lower than 140 bpm (< 140/3), the correlation with NSVT was perfect (r = 1) and agreement was also quite high (ICC = 0.800).

CONCLUSIONS: Pacemakers' event monitors underestimate the occurrence of ventricular arrhythmias detected by Holter. Standardization of pacemakers' algorithms is required before using this function for patients' clinical follow-up.