Canadian Journal of Cardiology



      Palpitations in children can be caused by benign or malignant heart rhythms. Documenting the rhythm during symptoms (symptom-rhythm correlation) can be diagnostic but may be challenging to achieve. The AliveCor Kardia monitor is an event recorder developed to detect atrial fibrillation in adults, with its utility in children remaining less rigorously studied. We compared use of the Kardia to the Cardiocall event recorder, our institution’s current standard event recorder for children with palpitations, through a prospective, randomized study of children presenting to pediatric cardiology for investigation of palpitations who require an event recorder for symptom-rhythm correlation.

      Methods and Results

      Pediatric patients were randomized to receiving the Kardia or Cardiocall event recorder for rhythm documentation. Diagnostic tracings were defined as recording one pathologic arrhythmia or 3 sinus rhythm tracings. We assessed tracing quality, diagnoses obtained, and time to diagnosis between groups. Patients were provided surveys to assess their perceptions of using the devices. Differences between groups were assessed using chi square, Mann-Whitney U, and Fisher exact analysis. Eighty-four participants were enrolled: 43 (51%) receiving Kardia and 41 (49%) Cardiocall devices. There were 148 tracings recorded (84 from Kardia and 64 from Cardiocall devices). Seventy-three (87%) Kardia tracings were of adequate quality for interpretation compared to 58 (91%, p=0.48) from Cardiocall. Diagnostic tracings were achieved in 51% vs 34% (p=0.11) in the Kardia vs Cardiocall group at medians of 15 (6-39) and 8 (3-21) days, respectively (p=0.23). Diagnoses obtained using Kardia vs Cardiocall tracings were sinus rhythm in 67 (80%) vs 57 (81%) tracings, AVRT or AVNRT in 5 (7%) vs 3 (4%), atrial tachycardia in 2 (3%) vs 0, atrial fibrillation in 1 (1%) vs 0, and indeterminate in 3 (4%) vs 6 (9%), respectively. Patients who used the Kardia monitor were more often willing to use the device again (90% vs 42%, p=0.012), with no differences between groups in finding episodes easy to record (74% vs 100%, p=0.13), easy to transmit (79% vs 46%, p=0.11), or overall satisfaction (75 vs 58%, p=0.44).


      Our preliminary data suggest the Kardia device provided adequate quality for rhythm strip interpretation with no difference compared to the standard Cardiocall monitor. Families who used the Kardia monitor were more willing to use the device again which should be considered when a symptom-rhythm correlation is needed.