Findings of an observational investigation of pure remote follow-up of pacemaker patients: is the in-clinic device check still needed?

Artificial intelligence to detect noise events in remote monitoring data

Background

Remote monitoring (RM) of cardiac implantable electrical devices (CIEDs) can detect various events early. However, the diagnostic ability of CIEDs has not been sufficient, especially for lead failure. The first notification of lead failure was almost noise events, which were detected as arrhythmia by the CIED. A human must analyze the intracardiac electrogram to accurately detect lead failure. However, the number of arrhythmic events is too large for human analysis. Artificial intelligence (AI) seems to be helpful in the early and accurate detection of lead failure before human analysis.

Objective

To test whether a neural network can be trained to precisely identify noise events in the intracardiac electrogram of RM data.

Methods

We analyzed 21 918 RM data consisting of 12 925 and 1884 Medtronic and Boston Scientific data, respectively. Among these, 153 and 52 Medtronic and Boston Scientific data, respectively, were diagnosed as noise events by human analysis. In Medtronic, 306 events, including 153 noise events and randomly selected 153 out of 12 692 nonnoise events, were analyzed in a five-fold cross-validation with a convolutional neural network. The Boston Scientific data were analyzed similarly.

Results

The precision rate, recall rate, F1 score, accuracy rate, and the area under the curve were 85.8 ± 4.0%, 91.6 ± 6.7%, 88.4 ± 2.0%, 88.0 ± 2.0%, and 0.958 ± 0.021 in Medtronic and 88.4 ± 12.8%, 81.0 ± 9.3%, 84.1 ± 8.3%, 84.2 ± 8.3% and 0.928 ± 0.041 in Boston Scientific. Five-fold cross-validation with a weighted loss function could increase the recall rate.

Conclusions

AI can accurately detect noise events. AI analysis may be helpful for detecting lead failure events early and accurately.

Investigating the Cost-Effectiveness of Telemonitoring Patients With Cardiac Implantable Electronic Devices: Systematic Review

BACKGROUND

Telemonitoring patients with cardiac implantable electronic devices (CIEDs) can improve their care management. However, the results of cost-effectiveness studies are heterogeneous. Therefore, it is still a matter of debate whether telemonitoring is worth the investment.

OBJECTIVE

This systematic review aims to investigate the cost-effectiveness of telemonitoring patients with CIEDs, focusing on its key drivers, and the impact of the varying perspectives.

METHODS

A systematic review was performed in PubMed, Web of Science, Embase, and EconLit. The search was completed on July 7, 2022. Studies were included if they fulfilled the following criteria: patients had a CIED, comparison with standard care, and inclusion of health economic evaluations (eg, cost-effectiveness analyses and cost-utility analyses). Only complete and peer-reviewed studies were included, and no year limits were applied. The exclusion criteria included studies with partial economic evaluations, systematic reviews or reports, and studies without standard care as a control group. Besides general study characteristics, the following outcome measures were extracted: impact on total cost or income, cost or income drivers, cost or income drivers per patient, cost or income drivers as a percentage of the total cost impact, incremental cost-effectiveness ratios, or cost-utility ratios. Quality was assessed using the Consensus Health Economic Criteria checklist.

RESULTS

Overall, 15 cost-effectiveness analyses were included. All studies were performed in Western countries, mainly Europe, and had primarily a male participant population. Of the 15 studies, 3 (20%) calculated the incremental cost-effectiveness ratio, 1 (7%) the cost-utility ratio, and 11 (73%) the health and cost impact of telemonitoring. In total, 73% (11/15) of the studies indicated that telemonitoring of patients with implantable cardioverter-defibrillators (ICDs) and cardiac resynchronization therapy ICDs was cost-effective and cost-saving, both from a health care and patient perspective. Cost-effectiveness results for telemonitoring of patients with pacemakers were inconclusive. The key drivers for cost reduction from a health care perspective were hospitalizations and scheduled in-office visits. Hospitalization costs were reduced by up to US $912 per patient per year. Scheduled in-office visits included up to 61% of the total cost reduction. Key drivers for cost reduction from a patient perspective were loss of income, cost for scheduled in-office visits and transport. Finally, of the 15 studies, 8 (52%) reported improved quality of life, with statistically significance in only 1 (13%) study (P=.03).

CONCLUSIONS

From a health care and patient perspective, telemonitoring of patients with an ICD or a cardiac resynchronization therapy ICD is a cost-effective and cost-saving alternative to standard care. Inconclusive results were found for patients with pacemakers. However, telemonitoring can lead to a decrease in providers' income, mainly due to a lack of reimbursement. Introducing appropriate reimbursement could make telemonitoring sustainable for providers while still being cost-effective from a health care payer perspective.

TRIAL REGISTRATION

PROSPERO CRD42022322334; https://tinyurl.com/puunapdr.

Real-life data on the workload of cardiac implantable electronic device remote monitoring in a large tertiary center

BACKGROUND

Cardiac implantable electronic devices (CIED) and implantable loop recorders (ILR) are increasingly monitored by systems allowing remote transmission of data from the patient to the hospital. Remote monitoring (RM) has been shown to increase patient satisfaction and safety. However, real-life data on the number and causes of the RM transmissions, and actions initiated by them are scarce.

METHODS

A total of 3446 patients with CIED and 92 patients with ILR were included in the study. Data on the number of alerts, scheduled and patient-initiated transmissions as well as the causes and actions initiated by the transmissions were systematically collected from March 1 to December 30, 2022. The data was subdivided by the device type.

RESULTS

During the study period 7087 remote CIED and 1212 ILR transmissions were generated, (0.2 and 1.3 per patient per month), respectively. Of these transmissions 49% (4084) were automatic alerts, and 29% (2434) and 22% (1781) were scheduled and patient initiated, respectively. Most of the CIED alerts (73%) and the scheduled transmissions (90%) were nonactionable, and only 7% and 5% led to in-office follow-up, respectively. Off all ILR alerts (1011) PM implantation was scheduled to 11 patients.

CONCLUSIONS

RM transmissions were common, but most of them were nonactionable. These real-life findings indicate that detailed analysis of the causes of the RM transmissions is important for optimization of the remote follow-up workload.

Clinic Time Required for Remote and In-person Management of Cardiac Device Patients: Time and Motion Workflow Evaluation

BACKGROUND

The number of patients with cardiac implantable electronic devices (CIEDs) is growing, creating substantial workload for device clinics.

OBJECTIVE

This study aimed to characterize the workflow and quantify clinic staff time requirements to manage CIED patients.

METHODS

A time and motion workflow evaluation was performed in 11 US and European CIED clinics. Workflow tasks were repeatedly timed during one business week of observation at each clinic. Observations were inclusive of all device models/manufacturers present. Mean cumulative staff time required to review a Remote device transmission and for an In-person clinic visit were calculated, including all necessary clinical and administrative tasks. Annual staff time for follow-up of 1 CIED patient was modeled using CIED transmission volumes, clinical guidelines, and published literature.

RESULTS

A total of 276 in-person clinic visits and 2,173 remote monitoring activities were observed. Mean staff time required per remote transmission ranged from 9.4-13.5 minutes for therapeutic devices (pacemaker, ICD, CRT) and 11.3-12.9 mins for diagnostic devices (insertable cardiac monitors (ICMs)). Mean staff time per in-person visit ranged from 37.8-51.0 mins and 39.9-45.8 mins, for therapeutic devices and ICMs respectively. Including all remote and in-person follow-ups, the estimated annual time to manage one CIED patient ranged from 1.6-2.4 hours for therapeutic devices and 7.7-9.3 hours for ICMs.

CONCLUSIONS

CIED patient management workflow is complex and requires significant staff time. Understanding process steps and time requirements informs implementation of efficiency improvements, including remote solutions. Future research should examine the heterogeneity in patient management processes to identify the most efficient workflows.

CLINICALTRIAL

Safety and efficiency of a common and simplified protocol for pacemaker and defibrillator surveillance based on remote monitoring only: a long-term randomized trial (RM-ALONE)

Aims

This trial aimed to evaluate the safety and efficiency of a common and simplified protocol for the surveillance of cardiac implantable electronic devices based on remote monitoring (RM) in patients with pacemakers (PMs) and implantable cardiac defibrillators (ICDs) for at least 24 months.

Methods and results

The RM-ALONE is a multicentre prospective trial that randomly assigned 445 patients in two groups, both followed by RM: the home monitoring-only (HMo) based on RM + remote interrogations (RIs) every 6 months and the HM + IO that adds in-office evaluations every 6 months to RM. Four hundred and forty-five patients were enrolled in the study, 294 PMs and 151 ICDs recipients. In the HMo group, 20% of patients experienced ≥1 major adverse cardiac event (MACE) vs. 19.5% in HM + IO group (P = 0.006 for non-inferiority). The proportion of patients with a PM/ICD who experienced ≥1 MACE was 15.2/29.3% in HMo group and 16.1/26.3% in HM + IO group (hazard ratio 0.95/1.15, 95% confidence interval 0.53–1.70/0.62–2.10). There were 789 in-office evaluations (136 in the HMo and 653 in the HM + IO; P < 0.001). There was a 79.2% reduction of in-office evaluations with no significant differences in unscheduled visits between groups: 122 (54.5%) in HMo and 101 (45.3%) in HM + IO; P = 0.15. The time a physician/nurse spent per patient/follow-up was significantly reduced in the HMo group: 4/5 min (0–30)/(1–30) vs. 10/10 min (0–40)/(1–40) in HM + IO (P < 0.0001).

Conclusion

The RM-ALONE protocol common for ICD and PM surveillance, consisting of RM + RI every 6 months has proven safe and efficient in reducing hospital visits and staff workload.

Cost-utility analysis of telemonitoring versus conventional hospital-based follow-up of patients with pacemakers. The NORDLAND randomized clinical trial

INTRODUCTION

The aim of our study was to perform an economic assessment in order to check whether or not telemonitoring of users with pacemakers offers a cost-effective alternative to traditional follow-up in outpatient clinics.

METHODS

We used effectiveness and cost data from the NORDLAND trial, which is a controlled, randomized, non-masked clinical trial. Fifty patients were assigned to receive either telemonitoring (TM; n = 25) or conventional monitoring (CM; n = 25) and were followed up for 12 months after the implantation. A cost-utility analysis was performed in terms of additional costs per additional Quality-Adjusted Life Year (QALY) attained from the perspectives of the Norwegian National Healthcare System and patients and their caregivers.

RESULTS

Effectiveness was similar between alternatives (TM: 0.7804 [CI: 0.6864 to 0.8745] vs. CM: 0.7465 [CI: 0.6543 to 0.8387]), while cost per patient was higher in the RM group, both from the Norwegian NHS perspective (TM: €2,079.84 [CI: 0.00 to 4,610.58] vs. €271.97 [CI: 158.18 to 385.76]; p = 0.147) and including the patient/family perspective (TM: €2,295.91 [CI: 0.00 to 4,843.28] vs. CM: €430.39 [CI: 0.00 to 4,841.48]), although these large differences-mainly due to a few patients being hospitalized in the TM group, as opposed to none in the CM group-did not reach statistical significance. The Incremental Cost-Effectiveness Ratio (ICER) from the Norwegian NHS perspective (€53,345.27/QALY) and including the patient/caregiver perspective (€55,046.40/QALY), as well as the Incremental Net Benefit (INB), favors the CM alternative, albeit with very broad 95%CIs. The probabilistic analysis confirmed inconclusive results due to the wide CIs even suggesting that TM was not cost-effective in this study. Supplemental analysis excluding the hospitalization costs shows positive INBs, whereby suggesting a discrete superiority of the RM alternative if hospitalization costs were not considered, albeit also with broad CIs.

CONCLUSIONS

Cost-utility analysis of TM vs. CM shows inconclusive results because of broad confidence intervals with ICER and INB figures ranging from potential savings to high costs for an additional QALY, with the majority of ICERs being above the usual NHS thresholds for coverage decisions.

TRIAL REGISTRATION

ClinicalTrials.gov NCT02237404.

Outsourcing the Remote Management of Cardiac Implantable Electronic Devices: Medical Care Quality Improvement Project

BACKGROUND

Remote management is partially replacing routine follow-up in patients implanted with cardiac implantable electronic devices (CIEDs). Although it reduces clinical staff time compared with standard in-office follow-up, a new definition of roles and responsibilities may be needed to review remote transmissions in an effective, efficient, and timely manner. Whether remote triage may be outsourced to an external remote monitoring center (ERMC) is still unclear.

OBJECTIVE

The aim of this health care quality improvement project was to evaluate the feasibility of outsourcing remote triage to an ERMC to improve patient care and health care resource utilization.

METHODS

Patients (N=153) with implanted CIEDs were followed up for 8 months. An ERMC composed of nurses and physicians reviewed remote transmissions daily following a specific remote monitoring (RM) protocol. A 6-month benchmarking phase where patients' transmissions were managed directly by hospital staff was evaluated as a term of comparison.

RESULTS

A total of 654 transmissions were recorded in the RM system and managed by the ERMC team within 2 working days, showing a significant time reduction compared with standard RM management (100% vs 11%, respectively, within 2 days; P<.001). A total of 84.3% (551/654) of the transmissions did not include a prioritized event and did not require escalation to the hospital clinician. High priority was assigned to 2.3% (15/654) of transmissions, which were communicated to the hospital team by email within 1 working day. Nonurgent device status events occurred in 88 cases and were communicated to the hospital within 2 working days. Of these, 11% (10/88) were followed by a hospitalization.

CONCLUSIONS

The outsourcing of RM management to an ERMC safely provides efficacy and efficiency gains in patients' care compared with a standard in-hospital management. Moreover, the externalization of RM management could be a key tool for saving dedicated staff and facility time with possible positive economic impact.

TRIAL REGISTRATION

ClinicalTrials.gov NCT01007474; http://clinicaltrials.gov/ct2/show/NCT01007474.

A Low Critical Event Rate Despite a High Abnormal Event Rate in Patients with Cardiac Implantable Electric Devices Followed Up by Remote Monitoring

Objective Remote monitoring (RM) of cardiac implantable electric devices (CIEDs) has been advocated as a healthcare standard. However, expert consensus statements suggest that all patients require annual face-to-face follow-up consultations at outpatient clinics even if RM reveals no episodes. The objective of this study was to determine the critical event rate after CIED implantation through RM. Methods This multicenter, retrospective, cohort study evaluated patients with pacemakers (PMs), implantable cardioverter defibrillators (ICDs), or cardiac resynchronization therapy defibrillator (CRT-Ds) and analyzed whether or not the data drawn from RM included abnormal or critical events. Patients A total of 1,849 CIED patients in 12 hospitals who were followed up by the RM center in Okayama University Hospital were included in this study. Results During the mean follow-up period of 774.9 days, 16,560 transmissions were analyzed, of which 11,040 (66.7%) were abnormal events and only 676 (4.1%) were critical events. The critical event rate in the PM group was significantly lower than that in the ICD or CRT-D groups (0.9% vs. 5.0% or 5.9%, p<0.001). A multivariate analysis revealed that ICD, CRT-D, and a low ejection fraction were independently associated with critical events. In patients with ICD, the independent risk factors for a critical event were old age, low ejection fraction, Brugada syndrome, dilated phase hypertrophic cardiomyopathy and arrhythmogenic right ventricular cardiomyopathy. Conclusion Although abnormal events were observed in two-thirds of the transmitted RM data, the critical event rate was <1% in patients with a PM, which was lower in comparison to the rates in patients with ICDs or CRT-Ds. A low ejection fraction was an independent predictor of critical events.

Prognostic role of non-sustained ventricular tachycardia detected with remote interrogation in a pacemaker population

Background

Non-sustained ventricular tachycardia (NSVT) can occur asymptomatically and can be incidentally detected in the internal records of pacemakers (PM). The clinical value of NSVT in the population of PM patients is still uncertain.

Our aim was to assess the prevalence of NSVT detected by remote PM control, to describe the clinical and demographic characteristics of patients with NSVT, and to assess the prognostic significance of NSVT in terms of both overall and cardiovascular mortality.

Methods

Consecutive patients followed with PM remote interrogations from September 2010 to December 2015 were included. The transmissions pertaining to the first 12 months of remote control were analysed and the patients were divided by those presenting NSVT and those without NSVT. The two groups were compared in terms of total mortality and cardiovascular mortality based on the administrative data provided by the regional administration of the Italian National Health System.

Results

The prevalence of NSVT in 408 patients (62% males, mean age 75.6; SD 10.6 years old) was 21% in a year. During a mean follow-up duration of 44 months, NSVT did not emerge as independently associated with overall mortality, but was associated with cardiovascular mortality in a competing risk regression model with older age, male gender, diabetes, chronic renal insufficiency, ischemic cardiomyopathy and chronic obstructive pulmonary disease.

Conclusions

We show that NSVT episodes recorded by remote control in a PM population are independently associated with cardiovascular mortality with possible implications for risk stratification and therapeutic options.

Organizational model and reactions to alerts in remote monitoring of cardiac implantable electronic devices: A survey from the Home Monitoring Expert Alliance project

BACKGROUND

This survey aimed to describe the organizational workflow of cardiac implantable electronic devices (CIEDs) remote monitoring (RM) service in ordinary practice.

METHODS

A questionnaire was designed for our purpose and completed by 49 sites participating to the Italian Home Monitoring Expert Alliance.

RESULTS

A dedicated organizational model for RM was set up for 86% of centers. The median RM team consisted of 2 (Interquartile range [IQR]: 1-3) physicians and 1 (IQR: 0-2) nurse. RM service was available in working hours and the median percentage of patients included was 100% (IQR: 10%-100%) for implantable cardioverter-defibrillator (ICD) and cardiac resynchronization therapy (CRT) recipients and 5% (IQR:0%-30%) for pacemakers. In-office follow-up was performed every 12 and 6 months for pacemaker and ICD/CRT recipients, respectively. More than 90% of sites used to activate all technical alerts, with a prompt reaction in case of an out-of-range parameter. The threshold for atrial fibrillation (AF) daily burden notification in most cases ranged from 2.4 to 7.2 hours. All ventricular arrhythmias alerts were usually switched on: an inappropriate therapy or more than one appropriate episode triggered an urgent in-hospital visit. Concerning heart failure, low CRT percentage pacing alert was always used, while the other available notifications were less frequently switched on.

CONCLUSIONS

This survey showed that RM service was usually set up with a primary nursing model including on average two responsible physicians and one nurse and mainly offered to ICD/CRT patients. Technical, AF and ventricular arrhythmia alerts triggered prompt reactions, while heart failure related indexes were generally less applied.