Decreased Inappropriate Shocks by Remote Monitoring of ICD
Decreased Inappropriate Shocks by Remote Monitoring of ICD
We randomly assigned 433 patients to RM (n = 221; active group) versus standard ambulatory follow-up (n = 212; control group) at 43 French medical centers (Appendix). The characteristics of the 2 study groups, including the use of cardiovascular drugs, were similar (Table 2). Medications to slow the sinus rate and atrioventricular conduction included β-adrenergic blockers in 76%, amiodarone in 23%, and digitalis in 2% of patients. With respect to the ICD programming, the lower limit of the VF zone was 214 bpm in 71% (69% in the active vs. 73% in the control group, P = 0.35) and 230 bpm in 31% of patients (29% vs. 33%, P = 0.36). The threshold of the first VT zone was <150 bpm in 4% of patients (4% vs. 3%, P = 0.46) in whom a slow VT had been documented and >176 bpm in 15% of patients (16% vs. 15%, P = 0.76), while its range was between 150 and 176 bpm in 68% of patients (66% vs. 70%, P = 0.35). In 4.8% of patients who presented with idiopathic VF or Brugada syndrome, the ICD was programmed with a VF zone only and a threshold of 200 bpm. The numbers of intervals needed to detect tachyarrhythmias were usually the nominal setting chosen by the device manufacturer, that is, 26 cycles in the first VT zone, 16 in the second, and 12 cycles in the VF zone.
Over a 27-month follow-up, 22 patients (10.4%) received ≥1 inappropriate shock in the control versus 11 patients (5.0%) in the active group, representing a 52% decrease (HR 0.47; 95% CI 0.23 to 0.97; P = 0.04), and a total of 283 versus 28 inappropriate shocks in the control versus the active groups. The ratio of inappropriate/total number of shocks was 283/657 (43.1%) in the control and 28/193 (14.5%) in the active group (P < 0.001).
The number (%) of patients and inappropriate shocks delivered in each group are detailed in Table 3. No patient in the active group received >10 inappropriate shocks, compared with 4 patients in the control group, with a maximum of 87 inappropriate shocks delivered to 1 patient. In a subgroup analysis, the proportion of patients who received inappropriate shocks was lower in the active than in the control subgroups in 18 of 19 comparisons (Fig. 1). The first inappropriate shock occurred at a mean of 7.4 ± 7.7 months in the control and 6.8 ± 8.2 months in the active group (P = 0.41). Figure 2 shows the Kaplan–Meier survival estimates of time to first inappropriate shock in the 2 study groups.
(Enlarge Image)
Figure 1.
Risk of inappropriate shocks (IS) according to selected characteristics in each study group. The values are hazard ratios (HR) with 95% confidence interval (CI) and the numbers of patients with IS/number (%) of patients in each study group. LVEF = left ventricular ejection fraction; NYHA = New York Heart Association.
(Enlarge Image)
Figure 2.
Cumulative estimate of first occurrence of inappropriate shock in each study group.
The inappropriate shocks were prompted by SVTA in 48.5%, ventricular oversensing in 18.2%, T wave oversensing in 15.2%, lead dysfunction in 15.2% and surgical interventions in 3% of patients, with no significant between-groups difference. RM decreased by 74% the number of inappropriate shocks prompted by SVTA and by 98% those due to lead dysfunction (Table 4 and Fig. 3).
(Enlarge Image)
Figure 3.
Causes of inappropriate shocks—proportions of patients with ≥1 inappropriate shock (A) or number of inappropriate shocks (B) in each study group for supraventricular tachycardia, noise oversensing, lead dysfunction or T-wave oversensing. IS = inappropriate shock; SVTA = supraventricular tachycardia.
Among the 33 recipients of inappropriate shocks, 11 were hospitalized in the control group for inappropriate shocks prompted by SVTA (n = 6), T wave oversensing (n = 1), and lead dysfunctions (n = 4), versus 3 patients in the active group hospitalized for inappropriate shocks prompted by SVTA (n = 1), T wave oversensing (n = 1), and lead dysfunction (n = 1), representing a 72% lower rate of hospitalizations for inappropriate shocks in the active group (P = 0.02). In addition, among these 33 patients, other episodes of inappropriate shocks were managed by changes in medical treatment or in ICD programming during scheduled ambulatory follow-ups (6 in the control versus 1 in the active group), during unscheduled follow-ups triggered by patient calls (5 in the control versus 2 in the active group) or by a RM event (2 in the active group) or during a hospitalization for another cause (3 in the control versus 2 in the active group).
In the active group, over the 27 months, additional ambulatory follow-ups or hospitalizations were triggered by 54 RM events in 46 patients, including oversensing (n = 16), ventricular arrhythmias (n = 24), R wave amplitude safety margin (n = 8), pacing mode switches (n = 2), abnormal lead impedance (n = 2), and <90% intrinsic ventricular rhythm (n = 2). Of these 54 RM events, 45 were actionable leading to 7 lead repositioning or replacements, 33 ICD reprogramming or changes in medical treatment, 1 external electric cardioversion of AF, and 4 radiofrequency ablation of ventricular tachycardia.
We retrospectively compared the median and mean number of days elapsed in both groups between RM reportable adverse events that might cause inappropriate shock (SVTA, inappropriate diagnosis and lead dysfunctions) and subsequent medical interventions. The median and mean delay was 5 [2–21] and 14 ± 20 days in the active, versus 25 [4–104] and 51 ± 59 days in the control group, representing a 20-day gain in the event detection (P = 0.04).
In the active group, lead dysfunctions occurred in 8 patients before the delivery of inappropriate shocks in 7. All lead dysfunctions were detected following a RM event. In the control group, lead dysfunction was the cause of 82 inappropriate shocks delivered to 5 patients. Therefore, the mean number of inappropriate shocks per patient-month was significantly lower in the active than in the control group (P = 0.02). Lead dysfunction events occurred at a mean time of 27 ± 22 months after lead implantation. Two of them occurred after an ICD replacement with no new lead implantation. Besides the repositioning or replacement of the dysfunctional leads, 4 ICD were replaced for premature battery depletion in the control versus 0 in the active group.
In the control group, 41.5% of patients experienced ≥1 major adverse event, versus 38.5% in the active group. Syncope occurred in 3 patients (1.4%) in the control group and 4 patients (1.8%) in the active group, due to arrhythmias in 2 patients in the control and 1 patient in the active group.
Results
Study Population and ICD Programming
We randomly assigned 433 patients to RM (n = 221; active group) versus standard ambulatory follow-up (n = 212; control group) at 43 French medical centers (Appendix). The characteristics of the 2 study groups, including the use of cardiovascular drugs, were similar (Table 2). Medications to slow the sinus rate and atrioventricular conduction included β-adrenergic blockers in 76%, amiodarone in 23%, and digitalis in 2% of patients. With respect to the ICD programming, the lower limit of the VF zone was 214 bpm in 71% (69% in the active vs. 73% in the control group, P = 0.35) and 230 bpm in 31% of patients (29% vs. 33%, P = 0.36). The threshold of the first VT zone was <150 bpm in 4% of patients (4% vs. 3%, P = 0.46) in whom a slow VT had been documented and >176 bpm in 15% of patients (16% vs. 15%, P = 0.76), while its range was between 150 and 176 bpm in 68% of patients (66% vs. 70%, P = 0.35). In 4.8% of patients who presented with idiopathic VF or Brugada syndrome, the ICD was programmed with a VF zone only and a threshold of 200 bpm. The numbers of intervals needed to detect tachyarrhythmias were usually the nominal setting chosen by the device manufacturer, that is, 26 cycles in the first VT zone, 16 in the second, and 12 cycles in the VF zone.
Incidence and Causes of Inappropriate Shocks
Over a 27-month follow-up, 22 patients (10.4%) received ≥1 inappropriate shock in the control versus 11 patients (5.0%) in the active group, representing a 52% decrease (HR 0.47; 95% CI 0.23 to 0.97; P = 0.04), and a total of 283 versus 28 inappropriate shocks in the control versus the active groups. The ratio of inappropriate/total number of shocks was 283/657 (43.1%) in the control and 28/193 (14.5%) in the active group (P < 0.001).
The number (%) of patients and inappropriate shocks delivered in each group are detailed in Table 3. No patient in the active group received >10 inappropriate shocks, compared with 4 patients in the control group, with a maximum of 87 inappropriate shocks delivered to 1 patient. In a subgroup analysis, the proportion of patients who received inappropriate shocks was lower in the active than in the control subgroups in 18 of 19 comparisons (Fig. 1). The first inappropriate shock occurred at a mean of 7.4 ± 7.7 months in the control and 6.8 ± 8.2 months in the active group (P = 0.41). Figure 2 shows the Kaplan–Meier survival estimates of time to first inappropriate shock in the 2 study groups.
(Enlarge Image)
Figure 1.
Risk of inappropriate shocks (IS) according to selected characteristics in each study group. The values are hazard ratios (HR) with 95% confidence interval (CI) and the numbers of patients with IS/number (%) of patients in each study group. LVEF = left ventricular ejection fraction; NYHA = New York Heart Association.
(Enlarge Image)
Figure 2.
Cumulative estimate of first occurrence of inappropriate shock in each study group.
The inappropriate shocks were prompted by SVTA in 48.5%, ventricular oversensing in 18.2%, T wave oversensing in 15.2%, lead dysfunction in 15.2% and surgical interventions in 3% of patients, with no significant between-groups difference. RM decreased by 74% the number of inappropriate shocks prompted by SVTA and by 98% those due to lead dysfunction (Table 4 and Fig. 3).
(Enlarge Image)
Figure 3.
Causes of inappropriate shocks—proportions of patients with ≥1 inappropriate shock (A) or number of inappropriate shocks (B) in each study group for supraventricular tachycardia, noise oversensing, lead dysfunction or T-wave oversensing. IS = inappropriate shock; SVTA = supraventricular tachycardia.
Management of Inappropriate Shock and RM Event
Among the 33 recipients of inappropriate shocks, 11 were hospitalized in the control group for inappropriate shocks prompted by SVTA (n = 6), T wave oversensing (n = 1), and lead dysfunctions (n = 4), versus 3 patients in the active group hospitalized for inappropriate shocks prompted by SVTA (n = 1), T wave oversensing (n = 1), and lead dysfunction (n = 1), representing a 72% lower rate of hospitalizations for inappropriate shocks in the active group (P = 0.02). In addition, among these 33 patients, other episodes of inappropriate shocks were managed by changes in medical treatment or in ICD programming during scheduled ambulatory follow-ups (6 in the control versus 1 in the active group), during unscheduled follow-ups triggered by patient calls (5 in the control versus 2 in the active group) or by a RM event (2 in the active group) or during a hospitalization for another cause (3 in the control versus 2 in the active group).
In the active group, over the 27 months, additional ambulatory follow-ups or hospitalizations were triggered by 54 RM events in 46 patients, including oversensing (n = 16), ventricular arrhythmias (n = 24), R wave amplitude safety margin (n = 8), pacing mode switches (n = 2), abnormal lead impedance (n = 2), and <90% intrinsic ventricular rhythm (n = 2). Of these 54 RM events, 45 were actionable leading to 7 lead repositioning or replacements, 33 ICD reprogramming or changes in medical treatment, 1 external electric cardioversion of AF, and 4 radiofrequency ablation of ventricular tachycardia.
Medical Intervention Delay
We retrospectively compared the median and mean number of days elapsed in both groups between RM reportable adverse events that might cause inappropriate shock (SVTA, inappropriate diagnosis and lead dysfunctions) and subsequent medical interventions. The median and mean delay was 5 [2–21] and 14 ± 20 days in the active, versus 25 [4–104] and 51 ± 59 days in the control group, representing a 20-day gain in the event detection (P = 0.04).
Lead Dysfunction
In the active group, lead dysfunctions occurred in 8 patients before the delivery of inappropriate shocks in 7. All lead dysfunctions were detected following a RM event. In the control group, lead dysfunction was the cause of 82 inappropriate shocks delivered to 5 patients. Therefore, the mean number of inappropriate shocks per patient-month was significantly lower in the active than in the control group (P = 0.02). Lead dysfunction events occurred at a mean time of 27 ± 22 months after lead implantation. Two of them occurred after an ICD replacement with no new lead implantation. Besides the repositioning or replacement of the dysfunctional leads, 4 ICD were replaced for premature battery depletion in the control versus 0 in the active group.
Adverse Events
In the control group, 41.5% of patients experienced ≥1 major adverse event, versus 38.5% in the active group. Syncope occurred in 3 patients (1.4%) in the control group and 4 patients (1.8%) in the active group, due to arrhythmias in 2 patients in the control and 1 patient in the active group.