Clinical and Economic Impact of Remote Monitoring on the Follow-Up of Patients with Implantable Electronic Cardiovascular Devices: An Observational Study

Abstract

Traditional follow-up of patients with cardiovascular devices is still an activity that, in addition to serving an increasing population, requires a considerable amount of time and specialized human and technical resources. Our aim was to evaluate the applicability of the CareLink^® (Medtronic, Minneapolis, MN) remote monitoring system as a complementary option to the follow-up of patients with implanted devices, between in-office visits. Evaluated outcomes included both clinical (event detection and time to diagnosis) and nonclinical (patient's satisfaction and economic costs) aspects. An observational, longitudinal, prospective study was conducted with patients from a Portuguese central hospital sampled by convenience during 1 week (43 patients). Data were collected in four moments: two in-office visits and two remote evaluations, reproducing 1 year of clinical follow-up. Data sources included health records, implant reports, initial demographic data collection, follow-up printouts, and a questionnaire. After selection criteria were verified, 15 patients (11 men [73%]) were included, 63.4±10.8 years old, representing 14.0±6.3 implant months. Clinically, 15 events were detected (9 by remote monitoring and 6 by patient-initiated activation), of which only 9 were symptomatic. We verified that remote monitoring could detect both symptomatic and asymptomatic events, whereas patient-initiated activation only detected symptomatic ones (p=0.028). Moreover, the mean diagnosis anticipation in patients with events was approximately 58 days (p<0.001). In nonclinical terms, we observed high or very high satisfaction (67% and 33%, respectively) with using remote monitoring technology, but still 8 patients (53%) stated they preferred in-office visits. Finally, the introduction of remote monitoring technology has the ability to reduce total follow-up costs for patients by 25%. We conclude that the use of this system constitutes a viable complementary option to the follow-up of patients with implantable devices, between in-office visits.

Introduction

Anti-bradycardia electric therapy was introduced in clinical practice more than 50 years ago, with proven effectiveness in terms of morbidity and mortality reduction and in quality of life improvement.¹ More recently, other electronic devices, particularly those delivering cardiac resynchronization therapy (CRT) and implantable cardioverter-defibrillators (ICDs), were introduced in clinical practice. These devices have the ability of restoring the normal ventricular activation sequence and/or detect and treat malignant tachyarrhythmias using electric therapy.²

In the last few years the development of the technology used in these devices along with emergence of new techniques and technologies has drastically increased the density of analysis and interpretation algorithms and therefore the complexity and duration of routine follow-up. Additionally, the development of the technology used in the batteries allowed an extension in the life of generators, thus increasing the mean number of follow-up years.³ If we add to that the increase in patients' mean life expectation, we can easily understand that follow-up assumes gigantic proportions in direct relation with the enormous subset of the population that we are managing.⁴

Traditional follow-up is still an activity that requires a considerable amount of time and specialized technical and human resources. It also has the inconvenience of long intervals between follow-ups, in which the physician has no access to the patient's information or the generator.⁵ Therefore, the adoption of new methodologies that allow simultaneously maintaining the same security and reliability levels, follow-up efficiency, and a tighter surveillance was deemed necessary. In this context, the possibility of remote monitoring (RM) of different devices poses a viable and safe complementary option in monitoring implantable device carriers between in-office visits.⁶

Fundamentals of Implantable Electronic Cardiovascular Devices

The term “implantable electronic cardiovascular device” (IECD) is a relatively recent classification that intends to group all electronic devices that are able to record, identify, analyze, and/or to treat cardiac rhythm and/or conduction disturbances.² IECDs include pacemakers, ICDs, and CRT devices as well as implantable loop recorders and hemodynamic monitors.^7,8 ICDs are devices designed to deliver high-energy shocks after detecting a potentially malignant ventricular arrhythmia.⁹ In turn, CRT has drastically transformed the treatment of patients with advanced congestive heart failure, refractory to drug therapy,¹⁰ with its success being responsible for the exponential growth in implant rate.¹¹

Fundamentals of Follow-Up

In general, guidelines recommend a methodology and follow-up model similar to pacemakers, ICDs, and CRT devices, with the major difference being the frequency of follow-up. They stress the need to revise parameters obtained during the implant prior to clinical discharge and later reprogramming in subsequent follow-ups according to results of performed tests and the patient's clinical needs.^4,8,12

Follow-up frequency and method are dependent on multiple factors, including the presence of other diseases, generator age, and patient's geographic accessibility to the center.⁴

Fundamentals of RM

Historically, healthcare involves traveling, whether the provider goes to the encounter with the patient or, more recently, where the patient goes to meet the provider. These displacements involve direct costs—fuel or transport tickets—and indirect costs in terms of time spent in displacement, delay in treatment, and loss of productivity, among others. All the participants involved in this process rapidly recognized that the development of electronic communications would have the potential to improve delivered care through cost reduction and delays inherent in the in-office visits. Hence, RM appears in the dependence of telemedicine and centers on clinically relevant data collection in the patient's own house and places other than the conventional hospitals, clinics, and private practices and subsequent data transmission to a central location for review. The conceptual model underlying almost all forms of RM is the understanding that the patient may suffer clinically relevant changes between in-office scheduled follow-ups and that these changes might be detected by obtaining and measuring determined physiological parameters. This treatment model assumes that if these changes are detected and treated earlier, they might positively influence the patient's clinical condition.¹³

Telemedicine as a concept concerns a multidisciplinary activity that involves medicine, informatics, and telecommunications with the objective of delivering healthcare using telematic solutions. It involves the use of modern information technology, especially interactive two-way audio/video communications, computers, and telemetry, in order to deliver healthcare services to patients and to facilitate information exchange between primary care providers and specialists distant from each other.¹³ Formally we can establish the concept of telemedicine as the use of informatics and telecommunications applied to medical practice, performed by physicians and other healthcare providers, including clinical assistance, teaching/research, and medical data transfer.¹⁴

Cardiology, because of the nature of physiological and biological signals used in the diagnosis and therapeutic control, benefits tremendously from using this type of application.¹⁵

Carelink^® System Fundamentals

The CareLink (Medtronic, Minneapolis, MN) system allows the reception, storage, transmission, and secure presentation of data acquired by the implantable device through the Internet. This system “accepts” encrypted data from the patient's implanted device through a remote monitor made available to the patient and “converts” it for “presentation” in a secure Web site—the CareLink Clinician Web site. The system includes three key elements: the implanted device, which gathers and stores physiological and device-related data, transmitting them afterward to the remote monitor through a normalized radio telemetry device; the remote monitor, which receives and temporarily stores gathered data and transmits them to the CareLink network through a serial connection over an analog phone line; and the CareLink network, which uses a series of secure servers that collect, process, and store data sent by the remote monitor. This information is then made available in a secure way to authorized healthcare professionals through the CareLink Clinician Web site.¹⁶ Implanted devices managed in this platform may be programmed to notify healthcare professionals on alert or sound events and notifying methods. Notification of implanted devices using the Conexus™ (Medtronic) wireless telemetry is available to the physician through the CareLink system by three methods: Web site, e-mail, and text message.¹⁷

Subjects and Methods

This is an observational, prospective, longitudinal study, initiated in September 2009 and concluded in October 2010. The study protocol was submitted to an expert panel for validation of the questionnaire to apply during the investigation as well as of the methodology. The protocol was approved by the Health Ethics Committee of the Centro Hospitalar do Porto, Porto, Portugal. All participants gave informed consent prior to participation in this study.

The aim of this study was to evaluate the applicability of the CareLink RM system as a complementary option to the follow-up of patients with cardioverson-defibrillation and/or resynchronization devices, between in-office visits.

Study Population

The target population was composed of all patients with an implanted cardioversion-defibrillation and/or resynchronization device for more than 3 months before enrollment in the Arrhythmology, Pacing, and Electrophysiology Unit of the Centro Hospitalar of Porto and with the capability of remote transmission via the CareLink system.

Selection Criteria

Inclusion criteria were defined as (a) having a cardioversion-defibrillation and/or resynchronization device implanted for more than 3 months prior to enrollment, (b) having access to a phone line, (c) having the capability to use the communicator in the patient's residence or that of a close relative, physically near, and (d) having signed the informed consent. Exclusion criteria were defined as (a) no access to a phone line compatible with the communicator in use, (b) having clinical conditions that may limit the participation (speech and/or hearing impairment, with no family support), (c) refusal to participate in the study, (d) being already included in another study/clinical trial, and (e) age under 18 years old.

Patients who did not fulfill the above inclusion criteria or had some exclusion criterion were not enrolled. The population was sampled by convenience among the patients who presented in the in-office visit between October 13, 2009 and October 20, 2009. The initial sample included 43 patients, with 11 being excluded for not having a wireless remote transmission-compatible device, 7 for not having a phone line, 2 for refusal, and 2 for being already enrolled in other studies. In addition, we had four deaths in this population. During the study, 2 patients were lost to follow-up.

System Costs

The equipment for RM was provided by the manufacturer, with no associated costs for the hospital or the patient. Transmission over the phone line was also provided by the manufacturer at no cost. The costs described for in-office visits entail device investigation as well as clinic visit.

Study Variables

Beyond sociodemographic data, clinical data were considered and included the following: (a) type of implanted device, (b) generator model, (c) implant date, (d) etiology, (e) New York Heart Association (NYHA) functional class, (f) number of events, (g) type of events (major and minor), (h) number of transmitted events, (i) admissions, (j) cause of admissions, and (k) time to diagnosis (time between event identification and the next in-office visit). Moreover, to assess technical applicability and cost control we used (a) cost of in-office and remote visit, (b) cost of displacements, (c) time spent in displacements, (d) waiting time for the in-office visit, (e) cost by accompanying person, (f) type of transmission (remote follow-up, RM, or patient-initiated activation), and (g) cause of transmission. Finally, to assess patient adherence to the equipment we used the following variables: (a) monitor set-up, (b) programming head and antenna positioning for questioning and transmission, (c) overall ease of use, (d) time spent in questioning and transmission, (e) RM influence in calmness/anxiety, (f) user satisfaction with remote equipment, (g) preferred visit method, and (h) method preference reason.

Additionally, we performed a comparison of the study subset with a hypothetical scenario that included in-office visits only.

Data Collection

Data collection was performed in the Arrhythmology, Pacing and Electrophysiology Unit of the Centro Hospitalar of Porto and divided in four moments: two in-office visits (at the beginning, when sociodemographic data were collected, and at the end of the study, when the questionnaire was administered) and two remote evaluations between in-office visits, at 6 and 9 months of follow-up (scheduled every 3 months). This methodology reproduced 1 year of clinical follow-up, assuming that all patients had a predischarge evaluation as reference. Data were obtained by consulting the patient's health record, implant report, initial sociodemographic data collection, follow-up evaluation summary, and the questionnaire administered at the final in-office evaluation.

All data processing was performed by a hospital RM specialist and reported to the responsible physician for action if needed.

Statistical Analysis

The statistical analysis of data was performed using SPSS version 16.0 (SPSS Inc., Chicago, IL) and R Statistical Computing version 2.12.0 statistical software.

Results

The final sample included 15 patients, 11 of whom were males (73.3%). Mean (±standard deviation) age of participants was 63.4±10.8 years (range, 48–83 years).

Regarding implant months, at the time of the first in-office visit, the mean (±standard deviation) number of implant months was 14.0±6.3 months, ranging between 5 and 24 months.

Regarding patient geographic distribution, 12 (80%) were from the Porto district, 2 (13.3%) from the Aveiro district, and 1 (6.7%) from the Bragança district. In education terms, 1 patient (6.7%) had had no formal instruction, 9 (60%) had finished grammar school, 3 (20%) had finished junior high school, and 2 (13.3%) had finished high school.

Regarding the type of implanted device, 10 patients (66.7%) were carriers of a single-chamber ICD, and 5 were carriers of CRT devices with defibrillation capability. As for etiology, 9 patients (60%) had received implants because of ischemic cause, and the remaining patients had received them for a nonischemic cause. Twelve patients (80%) had had devices implanted for primary prevention, with the remaining having had implants for secondary sudden death prevention. Ten patients (66.7%) were in NYHA class I, and the remaining 5 (33.7%) were in NYHA class II. General sample characteristics are summarized in Table 1.

Table 1.

General Sample Characteristics

CHARACTERISTIC	VALUE	P
Age (years)
Mean (SD)	63.4 (10.8)	0.669^a
Minimum; maximum	47; 80
Implant interval (months)
Mean (SD)	14.0 (6.3)	0.765^a
Minimum; maximum	5; 24
Gender [n (%)]
Male	11 (73.3)	0.121^b
Female	4 (26.7)
Municipality [n (%)]
Porto	7 (46.7)	1.000^b
Gondomar	3 (20.0)	0.039^b
Santa Maria da Feira	2 (13.3)	0.010^b
Bragança	1 (6.7)	0.002^b
Vila Nova de Gaia	1 (6.7)	0.002^b
Valongo	1 (6.7)	0.002^b
District [n (%)]
Porto	12 (80.0)	0.039^b
Aveiro	2 (13.3)	0.010^b
Bragança	1 (6.7)	0.002^b
Education [n (%)]
No formal instruction	1 (6.7)	0.002^b
Grammar school	9 (60.0)	0.606^b
Junior high school	3 (20.0)	0.039^b
High school	2 (13.3)	0.010^b
College	—	<0.001^b
Device [n (%)]
ICD-VR	10 (66.7)	0.302^b
CRT-D	5 (33.3)
Etiology [n (%)]
Ischemic	9 (60.0)	0.606^b
Nonischemic	6 (40.0)
Prevention [n (%)]
Primary	12 (80.0)	0.039^b
Secondary	3 (20.0)
NYHA Class [n (%)]
Class I	10 (66.7)	0.302^b
Class II	5 (33.3)	0.302^b
Class III	—	<0.001^b
Class IV	—	<0.001^b

There was a total of 15 subjects surveyed.

By Kolmogorov–Smirnov test (H0: normal distribution) with histogram visual analysis.

By Wilson proportion test with continuity correction (H0: proportion=0.5).

CRT-D, cardiac resynchronization therapy-defibrillation; ICD-VR, implantable cardioverter-defibrillator-single chamber; NYHA, New York Heart Association; SD, standard deviation.

Visit Description

We performed 60 visits with 30 in-office and 30 remote visits. Table 2 summarizes the general visit characteristics.

Table 2.

General Appointment Characteristics

	IOA 1		IOA 2
	N (%)	P ^a	N (%)	P ^a
Patients' professional status
Still working	1 (6.7)	0.002	1 (6.7)	0.002
Retired	14 (93.3)		14 (93.3)
Accompanied patients	7 (46.7)	1.000	6 (40.0)	0.606
Accompanying person's professional status
Still working	5 (71.4)	0.450	4 (66.6)	0.683
Retired	1 (14.3)	0.131	1 (16.7)	0.221
Unemployed	1 (14.3)	0.131	1 (16.7)	0.221
Work absence
None	13 (86.7)	0.010	14 (93.3)	0.002
Patient only	—	<0.001	—	<0.001
Accompanying person only	2 (13.3)	0.010	1 (6.7)	0.002
Both	—	<0.001	—	<0.001
Mode of transport to appointment
Public transportation	12 (80.0)	0.039	11 (73.3)	0.121
Own means	2 (13.3)	0.010	3 (20.0)	0.039
Ambulance/taxi	1 (6.7)	0.002	1 (6.7)	0.002

There was a total of 15 subjects surveyed.

By Wilson proportion test with continuity correction (H0: proportion=0.5).

IOA, in-office appointment.

As for professional status, 14 (93.3%) patients were retired, and 1 (6.7%) was still working, with no changes in status during visits.

In the first in-office visit, there was accompaniment in 7 (46.70%) visits: 5 (71.4%) accompanying persons were still working, 1 was retired (14.3%), and 1 was unemployed. The visit accompaniment resulted in no employment absences for either the patient or the accompanying individual in 13 visits (86.7%), and in 2 visits (13.3%) only the accompanying person missed work. Twelve visits (80.0%) required public transportations, 2 (13.3%) were by private means, and 1 (6.7%) was by taxi.

In the second in-office visit, there was accompaniment in 6 (40.0%) visits: 4 (66.6%) accompanying persons were still working, 1 was retired (16.7%), and 1 was unemployed. The visit accompaniment resulted in no employment absences for either the patient or the accompanying individual in 14 visits (93.3%), and in 1 visit only the accompanying person missed work. Traveling to the visit was by public transportation for 11 (73.3%) of the visits, 3 (20%) were by own means, and 1 (6.7%) was by taxi.

Use and Satisfaction

The opinion questionnaire regarding the monitoring system was administered in the last in-office visit, and the results are summarized in Table 3. From the analysis of Table 3 we observed that regarding monitor set-up, 11 (73.3%) patients considered it “easy” to use, whereas 2 (13.3%) found it “very easy” to use. Similarly, 9 (60%) patients declared the head and antenna positioning was “easy,” and 3 (20%) found it “very easy.” Regarding general use, answers were identical to monitor set-up. For Question 4, 3 (20%) patients considered the time needed for questioning and transmission as of “intermediate duration,” 10 (66.7%) classified it as “brief,” and 1 (6.7%) noted it as “very brief.” However, 1 patient considered this time as “long.” As for the influence of having a RM device in the state of calm/anxiety, 4 (26.7%) patients stated this had no influence, 7 (46.7%) patients stated RM was influencing “positively,” and 4 (26.7%) stated it was “very positively.” Finally, regarding satisfaction in the use of RM, 10 (66.7%) patients were “satisfied,” and 5 (33.3%) were “very satisfied.”

Table 3.

User Experience and Satisfaction with the Remote Monitoring System

	VERY DIFFICULT	DIFFICULT	INTERMEDIATE DIFFICULTY	EASY	VERY EASY
1. Monitor set-up	—	—	2 (13.3)	11 (73.3)	2 (13.3)
2. Head and antenna positioning	—	—	3 (20.0)	9 (60.0)	3 (20.0)
3. General use	—	—	2 (13.3)	11 (73.3)	2 (13.3)
	Very long	Long	Intermediate duration	Brief	Very brief
4. Time needed for questioning and transmission	—	1 (6.7)	3 (20.0)	10 (66.7)	1 (6.7)
	Very negatively	Negatively	No influence	Positively	Very positively
5. Influence in calm/anxiety state	—	—	4 (26.70)	7 (46.70)	4 (26.70)
	Very unsatisfied	Unsatisfied	Indifferent	Satisfied	Very satisfied
6. Satisfaction in the use of remote monitoring	—	—	—	10 (66.7)	5 (33.3)

There was a total of 15 subjects surveyed. Data are number (%).

As for the preferred method (Table 4), 8 (53.3%) patients preferred the in-office visit, and 4 (26.7%) preferred RM. The remaining 3 (20%) were “indifferent” to the method.

Table 4.

Preferred Method and Reasons for Method Preference

	N (%)	P ^a
Preferred method
In-office	8 (53.3)	1.000
Indifferent	3 (20.0)	0.039
Remote	4 (26.7)	0.121
Reasons for preferring in-office appointment
Direct physician observation	7 (46.7)	1.000
Direct contact with health professionals	5 (33.3)	0.302
Medical prescription possibility	1 (6.7)	0.002
“Trust” in diagnosis	1 (6.7)	0.002
Possibility of immediate admission	—	<0.001
Other(s)	—	<0.001
Reasons for preferring remote appointment
Reduction in the number of hospital displacements	4 (26.7)	0.121
Less time spent, in general	3 (20.0)	0.039
“Confidence” in the remote diagnosis of the professionals	3 (20.0)	0.039
More frequent monitoring	2 (13.3)	0.010
Less associated costs	1 (6.7)	0.002
Other(s)	—	<0.001

There was a total of 15 subjects surveyed.

By Wilson proportion test with continuity correction (H0: proportion=0.5).

When patients were questioned on the reasons for method preference, the options “direct physician observation” (46.7%) and “direct contact with health professionals” (33.3%) were the most common reasons for preferring the in-office method. Regarding preference for the remote method, the most chosen options were “reduction in the number of hospital displacements” (26.7%), “confidence in the remote diagnosis of the professionals” (20%), and “less time spent, in general” (20%).

Costs and Times Description

Costs and times spent in in-office visits are presented in Table 5. We should highlight that the median of the mean cost of visits for each patient was 2.80€, with a minimum of 1.50€ and a maximum of 100€. Similarly, the median of the time spent in displacements was 137.5 min, with a mean time of 140.0 min and ranging between a minimum of 40 min and a maximum of 420 min. The median of waiting time in our sample was 20 min, with a minimum of 10 min and a maximum of 90 min. The median of the mean waiting time for each patient was 22.5 min, ranging between 10 and 67.5 min.

Table 5.

Costs and Times Associated with In-Office Visits

	MEDIAN (P25; P75)	[MINIMUM; MAXIMUM]
Cost (€)
Patient
Displacements	2.80 (2.00; 4.25)	[1.50; 100.00]
Work	0.00 (0.00; 0.00)	[0.00; 00.00]
Accompanying person
Displacements	0.00 (0.00; 4.00)	[0.00; 100.00]
Work	0.00 (0.00; 0.00)	[0.00; 25.00]
Displacements mean for patient	2.80 (2.00; 7.50)	[1.50; 100.00]
Time (min)
Time spent in displacements	137.5 (120.0; 240.0)	[40; 420]
Visit waiting time	20.0 (15.0; 48,8)	[10; 90]
Mean waiting time by patient	22.5 (15.0; 45.0)	[10; 67.5]
Displacements mean time by patient	140.0 (120.0; 240.0)	[40; 420]

There was a total of 30 in-office visits.

P25; P75, 25th percentile; 75th percentile.

Comparison with an Hypothetical Scenario

Analyzing Table 6, we are able to compare two scenarios: the real scenario (scenario 1), with two in-office visits and two remote visits, and a hypothetical scenario (scenario 2), with four in-office visits, doubling the value of the two in-office visits of the real scenario. From the analysis of these two scenarios, we verified that in the real scenario the total cost of follow-up, including costs with displacements and accompaniment, was 2,276.40€ and that in scenario 2 this would be 3,052.80€. The difference in total cost of follow-up between the two scenarios would be 776.40€. The median of the cost per patient was 118€ in scenario 1 and would be 136€ in scenario 2, with a minimum of 115€ and 130€ and a maximum of 512€ and 924€, respectively, in the two scenarios. The median cost difference between the two scenarios would be 18€, with a minimum of 15€ and a maximum of 412€. The cost reduction percentage, when comparing the two groups, would be 13.24%, with a minimum of 11.54% and a maximum of 44.59%.

Table 6.

Total Costs and Cost Differences Between Two Scenarios

	SCENARIO 1 (2RA + 2IOA)	SCENARIO 2 (4IOA)
Total follow-up cost (€)	2,276.40	3,052.80
Difference in total follow-up cost (€)	776.40
Cost per patient (€)
Median	118.00	136.00
P25; P75	117.60; 132.00	135.20; 164.00
Minimum; maximum	115.00; 512.00	130.00; 924.00
Costs difference (€)
Median	18.00
P25; P75	17.60; 32.00
Minimum; maximum	15.00; 412.00
Costs reduction (%)
Median	13.24
P25; P75	13.02; 19.51
Minimum; maximum	11.54; 44.59

There was a total of 15 subjects surveyed.

IOA, in-office appointment; P25; P75, 25th percentile; 75th percentile; RA, remote appointment.

Events Description

During the study period, 15 events were identified and/or transmitted, and Table 7 gives the description of their particular characteristics.

Table 7.

Events and Way of Event Identification

	N (%)	P
Way of identification
Remote monitoring	9 (60.0)	0.606^a
Patient-initiated activation	6 (40.0)
Remote appointment	—	<0.001^a
Event symptoms
Symptomatic	9 (60.0)	0.606^a
Asymptomatic	6 (40.0)
Type of event
Clinical	5 (33.3)	0.302^a
Device related	5 (33.3)
Transmission/equipment related	5 (33.3)
Other (s)	—	<0.001^a
Event severity
Major	—
Minor	15 (100.0)	1.000^a
Type of event (minor)
OptiVol alert	5 (33.3)	0.302^a
Inappropriate shock	4 (26.7)	0.121^a
HF decompensation	3 (20.0)	0.039^a
Diaphragmatic stimulation	1 (6.7)	0.002^a
Pocket changes	1 (6.7)	0.002^a
Arrhythmias with no hemodynamic compromise	1 (6.7)	0.002^a
Transmitted events
Yes	7 (46.7)	1.000^a
No^b	8 (53.3)
Result
No need for urgent/emergency treatment/care	5 (33.3)	0.302^a
Changes in drug therapy/reprogramming deferred	5 (33.3)	0.302^a
Admission to ward	2 (13.3)	0.010^a
Visit scheduling	2 (13.3)	0.010^a
Admission to emergency department	1 (6.7)	0.002^a
Diagnosis anticipation time (days)
Mean (SD)	57.9 (30.3)	0.872^c
Minimum; maximum	7; 124	<0.001^d

There was a total of 15 subjects surveyed.

By Wilson proportion test with continuity correction (H0: proportion=0.5).

Note that even when events are not transmitted by the patient the events are identified and transmitted by the monitoring system.

By Kolmogorov–Smirnov test (H0: normal distribution) with histogram visual analysis.

By one-way Student's t test (H0: mean=0).

HF, heart failure.

Nine (60%) events were identified by RM and six (40%) by patient-initiated activation. Nine (60%) events were symptomatic, and six (40%) were asymptomatic.

Regarding the type of event, five (33.3%) were classified as clinical events, five were device related, and five were related to the transmission/equipment.

Regarding the severity of events, all events were classified as “minor,” and the most frequent events were OptiVol^® (Medtronic) system alerts (33.3%), followed by inappropriate shock delivery (26.7%) and decompensation of heart failure (20.0%).

Seven (46.7%) events were transmitted by patient initiative.

There was no need of urgent/emergency treatment or care in five (33.3%) events, and in five other events the need for drug therapy changes or reprogramming was deferred. Of the remaining events, two (13.3%) resulted in admission, two (13.3%) resulted in visit scheduling, and one resulted in admission to the emergency department of another hospital.

Finally, it is important to note that the mean (±standard deviation) time of diagnosis anticipation was 57.9±30.3 days, with a minimum of 7 days and a maximum of 124 days.

Clinical Associations

We found an association (p=0.028) between events symptoms and the manner of event identification (Table 8). Analyzing the data in Table 8 verifies that the events identified by patient-initiated activation are dependent on existence of symptoms and that RM has the ability to identify asymptomatic events that only later could be identified and symptomatic events that even were not transmitted by the patient.

Table 8.

Identification of Events Versus Symptoms

	WAY OF IDENTIFICATION
SYMPTOMS	REMOTE MONITORING	PATIENT-INITIATED ACTIVATION	P ^a
Symptomatic	3	6	0.028
Asymptomatic	6	0

There was a total of 15 subjects surveyed.

By Fisher's exact test.

Nonclinical Associations

Also, we were able to find statistically significant differences in the distribution of the variable mean waiting time between the group that preferred the in-office visit and the group that did not express that preference (p=0.017). The median of the group that preferred the in-office visit was 17.5 min, and the group that did not express that preference was 45.0 min (Table 9).

Table 9.

Mean Waiting Time and Preference for the In-Office Visit

	MEAN WAITING TIME (MIN)	P ^a
Group that prefers IOV
Median	17.5
P25; P75	10.6; 28.1
Minimum; maximum	10.0; 30.0	0.017
Group that does not prefer IOV
Median	45.0
P25; P75	22.5; 60.0
Minimum; maximum	15.0; 67.5

There was a total of 15 patients.

By Mann–Whitney test.

IOV, in-office visit; P25; P75, 25th percentile; 75th percentile.

Assuming that the mean waiting time is a factor that influences method preference, we tried to determine the mean waiting time that better distinguishes the two groups. The point that better discriminates the two groups corresponds to a mean waiting time of 37.5 min, with a sensitivity of 71.4% and a 100% specificity. It is important that we found no differences between the variables of time and mean cost of displacements by patient between the two preference groups.

Discussion

The results of this study are applicable to a general discussion complemented with a detailed specific analysis in several points.

We found statistically significant differences in the distribution of the variable mean waiting time between the group that preferred the in-office visit and the group that did not express that preference. This means that patients with longer waiting times do not prefer an in-office visit. Furthermore, it was possible to determine that the point that better distinguishes the two groups is to consider that the preference for an in-office visit disappears when the mean waiting time is greater than or equal to 37.5 min (corresponding to real data of waiting times greater than 45 min). This means that if the waiting time indicates that the patient does not prefer the remote appointment, then the patient really does not prefer that modality. In clinical terms this value could be of some importance, namely as the maximum waiting time for in-office visits.

An important data point that might influence future analysis is the geographic location of patients. It was possible to observe an increase of the cost reduction percentage in opposition to the increase of the geographic distance to the hospital (data not shown). Similarly, we tried to establish an association between geographic location and preference for the remote method but with no statistically significant results. Although it was not possible to establish an association, data analysis seems to suggest the existence of such an association.

Analyzing the data from the in-office visits is important to highlight that currently almost half of the patients were accompanied to the visit. If we consider the mean age of the population and the mean life expectancy, the dependency relation on accompaniment and the need for medically equipped transportations might lead to an increase of the costs associated with this type of visit relative to work absences by the accompanying person and the way of displacement to the visit. From our clinical experience, we verified that as patients grow older and are further in geographic origin from the medical office, the greater is the need for medically equipped transportation or a taxi. Although we were not able to establish a statistically significant association, we believe it is important to rethink in the future the associated costs of transportation, mainly if associated with distant geographic locations.

As for the analysis of the questionnaire, in the questions related to the use of the RM device, it is important to highlight that approximately 86% of patients considered the use of RM as “easy” or “very easy,” similar to reports in the literature.^18,19 Also, 72% of patients considered that the presence of a RM device influenced “positively” or “very positively” their state of calm/anxiety. Finally, and as described by Res et al.,²⁰ it is important to stress that all the individuals of the sample were “satisfied” or “very satisfied” with the use of RM. These data are nevertheless different considering the answer for the preferred method. Despite all patients being satisfied or very satisfied without difficulties in the use of the RM device, 53% of patients still preferred the in-office visit. The results obtained by Masella et al.¹⁸ significantly differ from ours with a preference for the remote method by 73% of patients. The fact that the patients who preferred the in-office method selected the direct physician observation or the direct contact with health professionals as the chief element in that preference leads us to think that those are possibly the main reasons. In our analysis it was not possible to associate the preference for the in-office visit with the occurrence of clinical events or costs, so that the challenge remains to study additional reasons for that preference.

Assuming the clinical importance of detection of events and the therapeutic orientation of these patients reinforced by the existence of a statistically significant association between asymptomatic patients and identification of events by RM, a detailed clinical analysis of these events and resulting orientation is necessary.

Therefore, we have to stress that the mean diagnosis anticipation in patients who noted events, symptomatic or not, is approximately 58 days (p<0.001). Clinically, these data seem important as they allow us to identify and manage/treat patients almost 2 months prior to the next scheduled physical visit. For example, in terms of identifying patients with aggravated heart failure, we might prevent a decompensation, and possibly an admission for acute pulmonary edema, with all inherent costs.

As for the importance of RM as a way to identify events it is important to note that of the 15 identified events, 10 (75%) needed some type of intervention. One should note that for an asymptomatic event with the need of intervention, if RM was not available it would only be identified in the next in-office visit. Similarly, four of the events identified by RM resulted in admission (n=2) or changes in drug therapy/reprogramming (n=2). Heidbüchel et al.²¹ verified that 45.7% of patients with nonscheduled activations needed some type of intervention (reprogramming or admission). In clinical terms it is important that of the five events classified as “clinical events,” all needed some type of intervention, and of those all were correctly and concomitantly identified by RM with one of the events being asymptomatic.

In the study by Marzegalli et al.²² patients declared RM as responsible for time savings (displacements and visits) and cost reduction. The introduction of two remote visits, as opposed to only in-office follow-up, allowed savings of 776€/year of follow-up. This saving allowed a reduction of 25.42% in the total costs. Analyzing the cost per patient the introduction of RM in a follow-up scenario allowed a saving rate of 18€/patient per year or a reduction of 13.24%/patient per year. Raatikainen et al.²³ obtained in their population a reduction of 22€/patient, allowing a 38% reduction in total costs. Also, Fauchier et al.²⁴ observed cost reductions for the patient and reduced displacements with a break-even point at 26 months. Additionally, Raatikainen et al.²³ were able to determine that cost reduction was significant when the distance to the follow-up center was greater than 40 km. In general, the work from Stoepel et al.²⁵ corroborates our findings.

Limitations

It is essential to recognize some limitations in our study. The main limitation is sample size. Although the initial sample was approximately 50 patients, for several reasons the sample size was gradually reduced. The final sample size made extremely difficult the analysis of data, and if the sample size had been slightly bigger, other results could have been added or even verified. Another limitation regards not taking into account the costs associated with admissions. It would be important to know if the cost of identifying the events, and later admission, would not have an effect in increasing costs for patients who otherwise would not be identified and therefore not treated. In this situation we have also to highlight the potential impact in the patient's quality of life. It would be important for cost-benefit, cost-effectiveness, and cost-utility analysis in this particular area.

Conclusions

This article draws conclusions in two major areas—clinical and nonclinical, supported by different types of analysis.

In clinical terms, it is possible to verify that RM has the ability to identify asymptomatic events that otherwise only later would be identified and symptomatic events although not transmitted by the patient. Clinically there is also consideration of the importance of RM in anticipating the diagnosis, with clear benefits to the patient.

Considering nonclinical terms, the mean waiting time for the visit might influence the type of preferred follow-up. Moreover, we were able to establish a mean waiting time at which interval that preference disappears. As for costs it is important to recognize that the use of the remote visit reduces total follow-up costs in this population.

Our results were similar to those described in the literature, and, to the best of our knowledge, our work was the first to verify similar conclusions for a Portuguese population. Additionally, our findings regarding event identification by RM and diagnosis anticipation may prove useful in the management of patients, with clear benefits for both patients and healthcare professionals as well as healthcare systems.

Overall, we can conclude that the use of this RM system constitutes a viable complementary option to the follow-up of patients with implantable devices, between in-office visits.

Footnotes

Disclosure Statement

No competing financial interests exist. P.D.C. was responsible for the conception and design of the article and for drafting the article. P.P.R. was responsible for drafting and critically revising the article. A.H.R. was responsible for critically revising the article.

References

Jeffrey

, Parsonnet

. Cardiac pacing, 1960–1985: A quarter century of medical and industrial innovation. Circulation, 1998; 97:1978–1991.

Luderitz

. We have come a long way with device therapy: Historical perspectives on antiarrhythmic electrotherapy. J Cardiovasc Electrophysiol, 2002; 13,Suppl 1:S2–S8.

Trohman

, Kim

, Pinski

. Cardiac pacing: The state of the heart. Lancet, 2004; 364:1701–1719.

Epstein

, DiMarco

, Ellenbogen

, Estes

NAM

III , Freedman

, Gettes

et al.

ACC/AHA/HRS 2008 guidelines for device-based therapy of cardiac rhythm abnormalities: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the ACC/AHA/NASPE 2002 Guideline Update for Implantation of Cardiac Pacemakers and Antiarrhythmia Devices)

J Am Coll Cardiol, 2008; 51:e1–e62.

Sweesy

, Holland

. Heart rhythm device follow-up. Greenville, SC: Arrhythmia Technologies, 2005.

Costa

, Rodrigues

, Reis

, Costa-Pereira

. A review on remote monitoring technology applied to implantable electronic cardiovascular devices. Telemed J E Health, 2010; 16:1042–1050.

Abraham

. New approaches to monitoring heart failure before symptoms appear. Rev Cardiovasc Med, 2006; 7,Suppl 1:S33–S41.

Wilkoff

, Auricchio

, Brugada

, Cowie

, Ellenbogen

, Gillis

et al. HRS/EHRA expert consensus on the monitoring of cardiovascular implantable electronic devices (CIEDs): Description of techniques, indications, personnel, frequency and ethical considerations. Heart Rhythm, 2008; 5:907–925.

Cuoco

, Gold

. The implantable cardioverter defibrillator. Ellenbogen

, Wood

. Cardiac pacing and ICD's. Singapore: Blackwell Publishing, 2008; 382–416.

10.

Moss

, Hall

, Cannom

, Klein

, Brown

, Daubert

et al. Cardiac-resynchronization therapy for the prevention of heart-failure events. N Engl J Med, 2009; 361:1329–1338.

11.

Bax

, Abraham

, Barold

, Breithardt

, Fung

JWH

, Garrigue

et al. Cardiac resynchronization therapy: Part 2—Issues during and after device implantation and unresolved questions. J Am Coll Cardiol, 2005; 46:2168–2182.

12.

Vardas

, Auricchio

, Blanc

, Daubert

, Drexler

, Ector

et al. Guidelines for cardiac pacing and cardiac resynchronization therapy. Eur Heart J, 2007; 28:2256–2295.

13.

Brennan

, Starren

. Consumer health informatics and telehealth. Shortliffe

, Cimino

. Biomedical informatics: Computer applications in health care and biomedicine, 3rd. New York: Springer, 2006; 511–536.

14.

Costa

, Covas

, Veiga

, Lopes

, Lemos

, Ribeiro

. Grupo de trabalho para o estudo da telemedicina. Lisbon: Secretaria de Estado Adjunta do Ministro da Saúde, Ministério da Saúde, 2002.

15.

Fonseca

. A telemonitorização em Ccrdiologia. Monteiro

, Cunha

, Carrasqueiro

. Telemedicina: Onde estamos e para onde vamos. Lisbon: Associação para a Promoção e Desenvolvimento da Sociedade de Informação, 2007; 117–128.

16.

Jung

, Rillig

, Birkemeyer

, Miljak

, Meyerfeldt

. Advances in remote monitoring of implantable pacemakers, cardioverter defibrillators and cardiac resynchronization therapy systems. J Cardiovasc Electrophysiol, 2008; 23:73–85.

17.

Burri

, Senouf

. Remote monitoring and follow-up of pacemakers and implantable cardioverter defibrillators. Europace, 2009; 11:701–709.

18.

Masella

, Zanaboni

, Di Stasi

, Gilardi

, Ponzi

, Valsecchi

. Assessment of a remote monitoring system for implantable cardioverter defibrillators. J Telemed Telecare, 2008; 14:290–294.

19.

Schoenfeld

, Compton

, Mead

, Weiss

, Sherfesee

, Englund

et al. Remote monitoring of implantable cardioverter defibrillators: A prospective analysis. Pacing Clin Electrophysiol, 2004; 27:757–763.

20.

Res

, Theuns

, Jordaens

. The role of remote monitoring in the reduction of inappropriate implantable cardioverter defibrillator therapies. Clin Res Cardiol, 2006; 95,Suppl 3:17–21.

21.

Heidbüchel

, Lioen

, Foulon

, Huybrechts

, Ector

, Willems

et al. Potential role of remote monitoring for scheduled and unscheduled evaluations of patients with an implantable defibrillator. Europace, 2008; 10:351–357.

22.

Marzegalli

, Lunati

, Landolina

, Perego

, Ricci

, Guenzati

et al. Remote monitoring of CRT-ICD: The multicenter Italian CareLink evaluation—Ease of use, acceptance, and organizational implications. Pacing Clin Electrophysiol, 2008; 31:1259–1264.

23.

Raatikainen

, Uusimaa

, van Ginneken

, Janssen

, Linnaluoto

. Remote monitoring of implantable cardioverter defibrillator patients: A safe, time-saving, and cost-effective means for follow-up. Europace, 2008; 10:1145–1151.

24.

Fauchier

, Sadoul

, Kouakam

, Briand

, Chauvin

, Babuty

et al. Potential cost savings by telemedicine-assisted long-term care of implantable cardioverter defibrillator recipients. Pacing Clin Electrophysiol, 2005; 28,Suppl 1:S255–S259.

25.

Stoepel

, Boland

, Busca

, Saal

, Oliveira

. Usefulness of remote monitoring in cardiac implantable device follow-up. Telemed J E Health, 2009; 15:1026–1030.

Clinical and Economic Impact of Remote Monitoring on the Follow-Up of Patients with Implantable Electronic Cardiovascular Devices: An Observational Study

Abstract

Introduction

Fundamentals of Implantable Electronic Cardiovascular Devices

Fundamentals of Follow-Up

Fundamentals of RM

Carelink® System Fundamentals

Subjects and Methods

Study Population

Selection Criteria

System Costs

Study Variables

Data Collection

Statistical Analysis

Results

Visit Description

Use and Satisfaction

Costs and Times Description

Comparison with an Hypothetical Scenario

Events Description

Clinical Associations

Nonclinical Associations

Discussion

Limitations

Conclusions

Footnotes

Disclosure Statement

References

Carelink^® System Fundamentals