Remote monitoring and outcomes in heart failure: 5-year study

Abstract

Aim

Managing heart failure is particularly challenging for patients with cardiac implantable electronic devices (CIEDs), often suffering from chronic conditions. This study aimed to evaluate the clinical outcomes of patients with implantable cardioverter-defibrillators or cardiac resynchronization therapy defibrillators who were remotely monitored over a 5-year period.

Methods

Clinical data were sourced from the Electrophysiology Registry of the Cardiology Unit. Survival rates, cardiovascular events (i.e. hospitalization or death), and the risk factors associated with these events were the main focus of the analysis. Univariate survival analyses were performed based on patient characteristics [ischemic disease, age over 75 years, atrial fibrillation, severe chronic kidney disease (CKD)]. Additionally, multivariate survival analysis was conducted using Cox regression.

Results

A total of N = 402 patients were included in the analysis. Over the 5-year follow-up, there were N = 68 deaths and N = 190 cardiovascular events. Kaplan–Meier analysis indicated that ischemic disease, age at least 75 years, atrial fibrillation and severe CKD were statistically significant risk factors for both death and cardiovascular events (log-rank test, P < 0.001). Cox regression analysis confirmed that ischemic disease, age 75 years and older, atrial fibrillation, implantation with a CRT-D device and severe CKD were all associated with a significantly higher risk of cardiovascular events and mortality.

Conclusion

Despite the proven effectiveness of remote monitoring, the clinical impact of heart failure in patients with CIEDs remains substantial. Continuous enhancements are necessary to facilitate early intervention and improve outcomes for these high-risk groups.

Introduction

Heart failure is one of the leading causes of hospitalization in Western countries, imposing a substantial burden on healthcare systems and significantly impacting patients’ quality of life.^1–4 Cardiac resynchronization therapy (CRT) is a well established and effective therapeutic option for patients with heart failure with more than two decades of clinical use.⁵ Large randomized trials and meta-analyses have confirmed its benefit across a substantial proportion of patients with reduced ejection fraction, demonstrating significant reductions in mortality and heart failure-related hospitalizations.^6–9 Recent advances in implantation tools have further streamlined the procedure and improved its safety profile.⁵ Italy remains among the countries with the highest rate of CRT implantations relative to population size.⁵ In addition, imaging modalities have assumed an increasingly important role, not only for identifying eligible candidates but also for guiding the selection of optimal left ventricular pacing sites.⁵

Managing heart failure is particularly challenging because patients with cardiac implantable electronic devices (CIEDs) often have multiple chronic conditions. These can include diabetes, ischemic disease, atrial fibrillation, chronic obstructive pulmonary disease, osteo-articular diseases, chronic kidney disease (CKD), and cognitive impairment.^10–12

The complexity of managing these comorbidities, combined with the progressive nature of heart failure, necessitates innovative approaches to patient care and monitoring. Traditional follow-up models, based on scheduled in-office visits, often fail to capture early signs of deterioration, leading to delayed interventions and preventable hospitalizations. Moreover, healthcare systems face increasing resource constraints, making it essential to identify strategies that optimize efficiency without compromising patient safety.¹³ Remote monitoring has emerged as a promising solution, allowing early detection of worsening heart failure by regularly collecting and monitoring key indicators like weight, heart rate, blood pressure, thoracic impedance and heart rhythm abnormalities.^14–16 This technology enables healthcare providers to detect subtle changes in a patient's condition before they manifest as severe symptoms requiring urgent medical attention.

The integration of remote monitoring with implantable cardioverter-defibrillators (ICDs) and CRT-Ds represents a significant advancement in heart failure management. The early warning system allows patients and physicians to adjust therapy or intervene promptly in case of anomalies, potentially preventing emergency department visits and hospitalizations.^17,18 However, remote monitoring clinical application is characterized by substantial heterogeneity across available technologies. Current remote monitoring platforms differ in their technical architecture, data acquisition modalities, transmission systems and alert algorithms, resulting in variable depth, frequency and clinical relevance of the information provided.^19,20 In addition to traditional single-parameter alerts, several multiparametric systems have been introduced to enhance the early detection of clinical deterioration.²⁰ These include HeartLogic, which integrates multiple physiological sensors into a composite index; thoracic impedance-based monitoring, designed to identify early signs of fluid accumulation; and device-derived arrhythmia burden alerts, which quantify atrial or ventricular arrhythmias over time.^19,21 Such diversity in remote monitoring technologies and alarm strategies has important implications for workflow integration, interpretation of alerts, and the overall effectiveness of remote monitoring in routine clinical practice.¹⁹ Despite the theoretical benefits of remote monitoring, long-term data on its effectiveness in real-world settings remain limited, particularly regarding its impact on mortality and hospitalization rates across different patient populations. The aim of this study was to assess the clinical consequences of remotely monitored patients treated with ICDs or CRT-Ds over an extended follow-up period of 5 years. The analysis focused specifically on two critical outcomes: survival and cardiovascular events (i.e. hospitalization or death). Furthermore, this investigation sought to evaluate the risk factors associated with these events, which is important to optimize patient selection and management strategies.

This study represents a continuation and expansion of our previous research,²² which initially evaluated the impact of remote monitoring in heart failure patients compared with standard management over a 2-year period. Building upon our earlier findings, we extended our observation period to investigate long-term outcomes and identify additional risk factors influencing outcomes.

Methods

Clinical data for this retrospective analysis were extracted from the Electrophysiology Registry of the Trento Cardiology Unit, which has been systemically collecting patient information from January 2011. The study protocol was approved by the local ethics committee. The investigation conformed to the principles outlined in the Declaration of Helsinki. All patients gave written informed consent, and data were treated confidentially.

The study population consisted of adult patients (aged ≥18 years), who: received either ICD or CRT-D therapy; were discharged alive from the hospital, after implantation; were monitored with remote monitoring; were followed up by the Electrophysiology Unit of the Santa Chiara Hospital (i.e. data were properly tracked in the registry) for maximum period of 5 years or died before the end of the study period. Patient observation started at the date of CIED implantation, which served as the index date.

Demographic and clinical characteristics of CIED implanted patients were collected (Supplementary Table S1). From a clinical standpoint, survival analysis was conducted, and the incidence of cardiovascular-related events (hospitalization or death) was measured.

Remote monitoring

The remote monitoring carried out is aimed at evaluating the operating status of the implanted device and evaluating some predominantly arrhythmic clinical parameters (related to the arrhythmic burden) and in any case is based on the transmission of data directly recovered from the device and through the device itself. Upon discharge from the hospital following implantation of the ICD or CRT-D, the patient is provided with a transmitter associated with the implanted device, which will be placed in the patient's bedroom. The transmitter, connected to the data or telephone network, transfers the data stored in the implanted device to the device manufacturer's platform. An algorithm will evaluate whether the data values are within the normal range or not, sending an alarm to the hospital in charge of the patient if the data are outside the normal range. The doctor or nurse who is in charge of remote monitoring management will access the platform and consult all the available information relating to the patient who generated the alarm, defining which strategy should be adopted to deal with the alarm received. Data transmissions can take place for several different reasons:

• (1)Functional problem (e.g. battery exhausted).
• (2)Clinical problem (recording of an episode of ventricular arrhythmia).
• (3)Programmed by the doctor (equivalent of scheduled interrogations with the only difference that they are carried out remotely).
• (4)Request of the patient (when the patient feels ill and wants to have doctor support).

As reported in the previous analysis, the remote monitoring is carried out daily (weekdays) and provides for the analysis of alarms according to a colour code. It is performed by two appropriately trained dedicated nurses who consult with the electrophysiologist cardiologist as needed. The basic parameters of the devices were always re-checked during the visit in person (sensing, impedances, threshold, programming, battery life).²² The physicians involved in remote monitoring activities are the same as those who also conduct outpatient visits and they are part of the electrophysiology team of the Trento Hospital Cardiology. Patients with a defibrillator undergo a scheduled outpatient visit once a year. Extra visits (unscheduled visits) were based on the severity of the remote monitoring alarm. In the initial phase of the program, remote monitoring activities were characterized by relatively manual processes: alert review was mostly sequential, with prioritization determined case by case and with a higher degree of hands-on involvement from clinical staff. As the program expanded and experience accumulated, more structured workflows were introduced to improve the timeliness and specificity of alert management. In the most recent phase, the remote monitoring program further matured through closer integration between the clinical team, technical personnel, and information systems. Workflows benefited from greater automation, both in alert prioritization and in the generation of clinical reports, allowing quicker evaluation of patients potentially at risk. Moreover, technological advancements in the devices—such as improved transmission reliability, an expanded set of monitored parameters, and enhanced diagnostic capabilities—helped increase the sensitivity and specificity of remote monitoring, making the overall process more responsive and targeted. Data transmission is not always followed by action. If needed, the action choice is dictated by the information found during the interrogation. Actions can vary from a simple observation of the evolution to a verification phone call or summoning the patient to the clinic cardiology, and up to the patient's admission.

Statistical analysis

Descriptive analyses and simple parametric and nonparametric testing (t-test and chi-square test) were used to identify baseline characteristics influencing mortality and/or cardiovascular events.

Continuous variables were examined by independent t-tests, whereas categorical variables were examined by chi-square test or Fisher's exact test. Survival curves and event rates were estimated using the Kaplan–Meier method and log-rank test. Binary outcomes (e.g. proportion of patients with at least one hospitalization) were compared by using nonparametric test of proportions. A P value of less than 0.05 was considered significant.

Two main evaluations were conducted:

• (1)In the initial univariate analysis, each risk factor was examined individually to assess its singular association with cardiovascular events and mortality. Univariate survival analyses were conducted by patient characteristics (presence of ischemic disease, age over 75 years, atrial fibrillation and device type).
• (2)Subsequently, using the Cox regression model, a multivariate analysis was implemented that allowed the simultaneous control of multiple variables, more precisely identifying independent risk predictors.

This methodology enabled the estimation of adjusted hazard ratios, providing a comprehensive evaluation of the impact of different factors on clinical course and prognosis. The adopted statistical approach ensured a comprehensive and scientifically rigorous assessment of potential determinants of cardiovascular outcomes. All statistical data were analysed using the Stata 13 software.

Results

Baseline characteristics

In the enrolment period, N = 402 patients carrying either ICD or CRT-D devices met the inclusion criteria and were included in the analysis (Table 1). The mean age (± standard deviation) was 66.3 ± 12.6 years. The majority of patients were men (N = 311, 77%). Respectively, 65 and 35% of patients were implanted with ICD or CRT-D.

Table 1.

Patients’ characteristics at baseline

Parameter	Overall (N = 402)	Ischemic disease (N = 201)	Age above 75 years (N = 115)	Atrial fibrillation (N = 122)	Severe CKD (N = 35)
Age (years), mean (SD)	66.3 (12.6)	69.1 (0.7)	78.9 (0.3)	71.4 (0.8)	74.1 (6.7)
Male gender (%)	311/402 (77.4)	179/201 (89.1)	87/115 (75.7)	101/122 (82.8)	30/35 (85.7)
Medical history
Diabetes (%)	90/398 (22.6)	57/198 (28.8)	22/112 (19.6)	32/120 (26.7)	11/35 (31.4)
Hypertension (%)	234/397 (58.9)	138/197 (70.1)	84/112 (75.0)	84/120 (70.0)	28/35 (80.0)
Severe CKD (%)	35/398 (8.8)	27/198 (13.7)	15/112 (13.4)	19/120 (15.8)	35/35 (100)
Stroke or TIA (%)	28/398 (7.0)	17/198 (8.6)	14/112 (12.5)	12/120 (10.0)	5/35 (14.3)
Myocardial infarction (%)	158/401 (39.4)	154/201 (76.6)	53/114 (46.5)	44/121 (36.4)	23/35 (65.7)
Coronary disease, without MI (%)	76/400 (19.0)	56/200 (29.5)	20/113 (17.7)	29/121 (24.0)	9/35 (25.7)
Deep vein thrombosis (%)	234/398 (58.8)	114/199 (57.3)	62/122 (55.4)	82/120 (68.3)	28/35 (80.0)
Atrial fibrillation (%)	122/402 (30.4)	63/201 (31.3)	51/115 (44.5)	122/122 (100)	19/35 (54.3)
Pharmacological treatment
Treatment with ACE inhibitors/ARBs (%)	304/392 (77.6)	158/195 (81.0)	85/110 (77.3)	90/118 (76.3)	23/35 (65.7)
Treatment with beta-blockers (%)	368/391 (94.1)	187/195 (95.9)	104/110 (94.6)	110/118 (93.2)	34/35 (97.1)
Treatment with diuretics (%)	293/392 (74.7)	157/195 (80.5)	93/110 (85.6)	104/118 (88.1)	33/35 (94.3)
Treatment with anticoagulation drugs (%)	134/395 (33.9)	69/197 (35.0)	50/112 (44.6)	103/120 (85.8)	19/35 (54.3)
Treatment with antiarrhythmic drugs (%)	86/393 (21.9)	54/195 (27.7)	27/110 (24.5)	39/119 (32.8)	13/35 (37.1)
Cardiac implantable electronic device
Treatment with ICD (%)	261/402 (64.9)	140/201 (69.7)	64/115 (55.6)	64/122 (52.5)	15/35 (42.9)
Treatment with CRT-D (%)	141/402 (35.1)	61/201 (30.4)	51/115 (44.4)	58/122 (47.5)	20/35 (57.1)
Cardiac indicators
CHA2DS2-VASc score ≥2 (%)	319/397 (80.4)	186/198 (93.9)	111/111 (100)	114/119 (95.8)	35/35 (100)
Left ventricular ejection fraction (%)
<35%	229/397 (57.7)	122/199 (61.3)	66/114 (57.9)	77/120 (64.2)	27/35 (77.1)
35–44%	66/397 (16.6)	38/199 (19.1)	22/114 (19.3)	24/120 (20.0)	5/35 (14.3)
45–54%	36/397 (9.1)	20/199 (10.1)	13/114 (11.4)	9/120 (7.5)	3/35 (8.6)
>55%	66/397 (16.6)	19/199 (9.6)	13/114 (11.4)	10/120 (8.3)	0/35 (0.0)
NYHA functional classification (%)
Class I	106/401 (26.4)	39/201 (19.4)	29/114 (25.4)	23/121 (19.0)	1/35 (2.9)
Class II	258/401 (64.3)	114/201 (71.6)	75/114 (65.8)	3/121 (68.6)	29/35 (82.9)
Class III	37/401 (9.2)	18/201 (9.0)	10/114 (8.7)	15/121 (12.4)	5/35 (14.3)

ACE, angiotensin-converting enzyme; ARB, angiotensin II receptor blocker; CHA₂DS₂-VASc score, congestive heart failure, hypertension, age over 75 years, diabetes mellitus, stroke, vascular disease, age 65–74 years, sex category (female); CKD, chronic kidney disease; CRT-D, cardiac resynchronization therapy defibrillators; ICD, implantable cardioverter-defibrillators; MI, myocardial infarction; NYHA, New York Heart Association; SD, standard deviation; TIA, transient ischemic attack; yrs, years.

More than half of the patients had a history of hypertension (59%) and N = 90 had diabetes (23%). Almost all the patients were in treatment with beta-blockers (94%) and a large number with angiotensin-converting enzyme (ACE) inhibitors/angiotensin II receptor blockers (ARBs, 78%) or diuretics (75%). The majority of patients had New York Heart Association (NYHA) class II (64%) and sex category score (CHA₂DS₂-VASc) at least 2 (80%).

The analysis focused on examining the potential prognostic implications of several critical clinical factors, including ischemic heart disease, advanced age (75 years and older), and atrial fibrillation. In particular:

• (1)N = 201 patients had ischemic disease (50%). Baseline characteristics were significantly worse in patients with ischemic disease compared with patients without ischemic disease (N = 201, 50%, Table 1 and Supplementary Table S 2).
• (2)N = 115 patients had age above 75 years (29%). With the exception of a few factors (e.g. use of diuretics, history of hypertension, severe CKD, or stroke/transient ischemic attack), baseline characteristics were balanced between patients above and below 75 years old (N = 287, 71%; Table 1 and Supplementary Table S2).
• (3)N = 122 patients had atrial fibrillation (29%). Similarly to the previous subgroup, except for a few factors (e.g. history of hypertension, severe CKD, or deep vein thrombosis), baseline characteristics were balanced between patients with and without atrial fibrillation (N = 280, 29%; Table 1 and Supplementary Table S2).
• (4)N = 35 patients (9%) had severe CKD, defined as estimated glomerular filtration rate (eGFR) less than 30 ml/min/1.73 m². In alignment with the other subgroups, baseline characteristics were balanced between patients with and without severe CKD (N = 363, 91%; Table 1 and Supplementary Table S2).

Survival and cardiovascular-related event rates

After a comprehensive follow-up of 5 years since CIED implantation, there were N = 68 deaths and N = 190 cardiovascular events among the study participants, corresponding to an annal rate of 3.63 and 14.26%, respectively (Table 2).

Table 2.

Survival and cardiovascular-related events rates

Patients’ group	Mortality			Cardiovascular eventsc
	Events (N)	Total rate (%)	Annual rate (%)a	Events (N)	Total rate (%)	Annual rate (%)a
All patients
Overall	68	16.92	3.63	190	47.26	14.26
Ischemic disease
Yes	51	25.37	5.68	110	54.73	17.72
No	17	8.46	1.75	80	39.80	11.24
P valueb	<0.001			0.005
Patients’ age
Below 75 yrs	30	10.45	2.17	120	41.81	11.93
Above 75 yrs	38	33.04	7.80	70	60.87	21.40
P valueb	<0.001			<0.001
Atrial fibrillation
Yes	31	25.41	5.68	78	63.93	23.49
No	37	13.21	2.79	112	40.00	11.19
P valueb	0.002			0.005
Severe CKD
Yes	15	42.86	10.81	28	80.00	35.26
No	51	14.05	2.97	159	43.80	12.83
P valueb	<0.001			<0.001

CKD, chronic kidney disease; CV, cardiovascular; yrs, years.

^aCalculated considering the N of events and the total analysis time.

^bLog-rank test.

^cHospitalization or death.

Research examining mortality and cardiovascular event outcomes showed distinct patterns across patient subgroups. Ischemic disease emerged as a significant predictor of mortality, with affected patients experiencing a more than three-fold higher annual death rate compared with those without ischemic disease (5.68 vs. 1.75%, P < 0.001; Table 2). Similarly, patients with ischemic disease showed significantly higher rates of cardiovascular events (17.72 vs. 11.24%, P = 0.005; Table 2).

Age stratification revealed pronounced differences, with patients aged over 75 years demonstrating significantly elevated risks for both mortality and cardiovascular events. The annual mortality rate in the older cohort was 7.80% compared with 2.17% in patients below 75 years old (P < 0.001; Table 2). Similarly, the cardiovascular events rate was markedly higher in the elderly group (21.40 vs 11.93%, P < 0.001; Table 2).

Atrial fibrillation was associated with adverse outcomes across both endpoints. Patients with atrial fibrillation exhibited a higher annual mortality rate (5.68 vs 2.79%, P = 0.002) and substantially increased cardiovascular events rate (23.49 vs. 11.19%, P = 0.005; Table 2) compared with those without the condition.

Additionally, patients with CKD exhibited a higher annual mortality rate (10.81 vs. 2.97%, P = 0 < 0.001) and substantially increased cardiovascular events rate (35.26 vs. 12.83%, P < 0.0001; Table 2) compared with those without the condition.

Univariate and multivariate analyses

Comprehensive survival analyses were conducted using Kaplan–Meier methodology, with detailed graphical representations provided in the Graphical abstract.

Statistical evaluation through log-rank testing revealed that all examined variables showed statistically significant associations with increased risk for both mortality and cardiovascular events (P < 0.001 for each variable) (Fig. 1).

Fig. 1.

Kaplan–Meier estimates for survival and cardiovascular events, by subgroup. ^*Hospitalization or death. AF, atrial fibrillation; CKD, chronic kidney disease; CV, cardiovascular; KM, Kaplan–Meier; yrs, years.

To further substantiate these findings, multivariate Cox proportional hazards regression analyses were performed. These advanced statistical models confirmed that several key clinical characteristics were independently associated with significantly elevated risks for both death and cardiovascular events. The detailed outcomes from the Cox regression analysis, presented in terms of hazard ratio and SD, are presented in Table 3. The presence of ischemic heart disease (hazard ratio for mortality: 3.187; hazard ratio for cardiovascular events: 1.48) or severe CKD (hazard ratio for mortality: 2.265; hazard ratio for cardiovascular events: 1.608) and the implantation of a CRT-D (hazard ratio for mortality: 2.689; hazard ratio for cardiovascular events: 1.507) significantly increased the likelihood of both cardiovascular events and mortality (P < 0.05). Advanced age (75 years and older) was a robust predictor of mortality (hazard ratio 2.698), rather than the presence of atrial fibrillation, emerged as a critical risk factor for hospitalization (hazard ratio 1.642).

Table 3.

Results of Cox regression for survival and cardiovascular events, by subgroup

Risk factors	Mortality			CV eventsa
	Hazard ratio (HR)	Std deviation (SD)	P value	Hazard ratio (HR)	Std deviation (SD)	P value
Ischemic disease (yes vs. no)	3.187	0.925	<0.001	1.480	0.227	0.010
Age ≥75 years (yes vs. no)	2.698	0.675	<0.001	1.295	0.204	0.098
Atrial fibrillation (yes vs. no)	1.553	0.391	0.08	1.642	0.254	0.001
Implant type (CRT-D vs. ICD)	2.689	0.685	<0.001	1.507	0.232	0.008
Severe CKD (yes vs. no)	2.265	0.675	0.006	1.608	0.345	0.027

CKD, chronic kidney disease; CRT-D, cardiac resynchronization therapy defibrillator; CV, cardiovascular; ICD, implantable cardioverter defibrillator; Std, standard.

^aHospitalization or death.

Discussion

Optimizing the efficiency of remote monitoring for CIEDs requires a comprehensive management model. This model should clearly define roles and responsibilities, actively involve healthcare professionals, and establish a robust action plan to address the CIED system alarm. This ensures achievement of the desired outcomes, such as improvement of clinical outcomes and reduction of acute care costs. The remote monitoring management model currently used in the Trento Cardiology Unit of the Santa Chiara Hospital has evolved through progressive changes while incorporating ongoing technological advancements in cardiology. Initially, the alert review relied on case-by-case prioritization and clinical judgment, with limited technological support. Over time, standardized triage protocols, harmonized workflows, and improved devices have reduced nonrelevant alerts and enhanced efficiency. In the most recent phase, greater automation, integration with information systems, and advanced device capabilities have increased sensitivity and specificity, enabling earlier identification of clinically relevant conditions and more timely interventions. This transition highlights how remote monitoring maturation likely contributed to improved patient outcomes.

The previous analysis, published in 2023, provided evidence that remote monitoring led to significantly lower short-term (2-year) morbidity and mortality risks in patients carrying CIEDs, compared with standard monitoring based on the traditional in-office visit approach.²² Moreover, a lower proportion of patients in the remote monitoring group (25.1%) were hospitalized for cardiovascular-related reasons, compared with the standard monitoring group (51.3%; P < 0.0001, two-sample test for proportions).²² It is important to note that the number of alarms and actions was not the primary outcome of this analysis. The reduced hospitalization rate leads to a significant decrease in inpatient costs, thereby alleviating the economic burden on the healthcare system. Overall, the implementation of the remote monitoring program proved to be cost-saving from both the payer and hospital perspectives. The investment needed to fund remote monitoring, which includes a service fee from the payer's perspective and staffing costs for hospitals, was more than offset by the reduced rate of hospitalizations for cardiovascular-related disease. This resulted in savings of €4771 per patient for payers and €6752 per patient for hospitals over 2 years.²²

The present analysis provides additional critical insights into the remote monitoring management model, underling a particular unmet need in specific subgroups where there is still considerable room for improvement, such as patients with ischemic disease, age above 75 years or atrial fibrillation. All examined variables showed statistically significant associations with increased risk for both mortality and cardiovascular events. The use of innovative solutions, such as advanced predictive analytics, personalized treatment plans, and enhanced monitoring technologies that can predict adverse events more accurately and reduce the outcome gap between ischemic and nonischaemic patients could improve the prognosis of these patients.

Compared with the previous analysis, the robustness of this study is significantly enhanced by the large sample size and the extended follow-up period to 5 years. These factors contribute to the reliability of the findings and provide a more comprehensive understanding of the long-term benefits of remote monitoring.

Patients with ischemic heart disease, severe chronic kidney disease, advanced age (≥75 years), or atrial fibrillation emerged as having high-risk profiles, consistently associated with increased mortality, cardiovascular events, or hospitalizations. These subgroups may benefit from more intensive remote monitoring follow-up and tailored surveillance algorithms, aimed at anticipating deterioration and enabling timely interventions. In particular, closer monitoring of arrhythmia burden in patients with atrial fibrillation, and proactive management of comorbidities such as CKD, could mitigate the risk of adverse outcomes.

At the organizational level, these results highlight the need to strengthen remote monitoring pathways by integrating multiparametric alerts into structured care models, fostering collaboration between device clinics and heart failure teams, and supporting patient adherence through education and engagement strategies. Such improvements may enhance the translation of remote monitoring data into actionable clinical decisions, thereby maximizing patient benefit in routine practice.

The present findings should be interpreted in the context of remote monitoring evidence. The TRUST trial²³ demonstrated that the remote monitoring of ICDs reduced in-office visits without compromising safety, while the IN-TIME trial¹⁸ confirmed improved clinical outcomes, including reduced mortality, in patients with heart failure. Contemporary reviews and registry data (Pierucci et al., 2025¹³; Pizarro et al. 2025²⁴; Spethmann et al., 2024²⁵) confirm that remote monitoring reduces hospitalizations and improves patient outcomes, while highlighting the growing role of wearable devices and artificial intelligence. Systematic reviews (Masotta et al., 2024²⁶) and multiparametric monitoring studies (Boriani et al., 2024²⁰) further emphasize the variability of adherence and the need for integrated care pathways. These comparisons underscore the added value of our real-world analysis, which complements trial data by reflecting routine clinical practice over long-term follow-up.

In line with these reports, our data confirm that remote monitoring contributes substantially to the early identification of clinically relevant events and to the reduction of visits. At the same time, differences in the proportion of arrhythmia-related events and in the balance between remote monitoring-triggered vs. emergency-initiated visits underscore the added value of real-world evidence, which complements trial data by reflecting routine clinical practice. This contextualization strengthens the interpretation of our results and highlights the importance of integrating remote monitoring into standard care pathways for patients with heart failure and implantable devices.

The presence of ischemic heart disease, CKD, CRT-D implantation, advanced age, and atrial fibrillation as predictors of adverse outcomes is consistent with recent evidence. Recent meta-analyses^27,28 and reviews^13,24 have similarly identified these comorbidities and clinical parameters as major determinants of mortality and hospitalization in patients under remote monitoring. This alignment underscores the robustness of our results and highlights the added value of long-term real-world data.

Beyond the established effects on mortality and heart-failure-related hospitalizations, timely evaluation of remote monitoring-generated alerts and prompt clinical intervention may confer additional, clinically meaningful benefits. Early identification of device-detected abnormalities can reduce the need for urgent, unplanned hospital admissions, thereby limiting exposure to hospital-acquired complications and lowering the risk of device-related infections associated with repeated invasive procedures.^29,30 Similarly, earlier detection and management of arrhythmias or impending heart failure decompensation may help mitigate thromboembolic risk, particularly in patients with atrial fibrillation or fluctuating volume status.^29,31 Moreover, by enabling the proactive management of lead or device issues before they progress to overt malfunction, remote monitoring may decrease the likelihood of extraction procedures, which are known to carry substantial morbidity and mortality.²⁹ These broader advantages highlight the potential of remote monitoring not only as a tool for surveillance but also as a strategy to support safer, more efficient, and more preventive long-term care.

Limitations

As for most observational studies, certain methodological issues could limit the validity of the results of the present study. In particular, one of the primary limitations of the study is the potential for selection bias. The study population may have included patients with more severe conditions, which could have affected the generalizability of the findings. Future studies should aim to include a more diverse patient population to validate these results across different demographics and clinical settings. Moreover, the inclusion of patients with more severe conditions might have influenced the outcomes, making it challenging to generalize the findings to a broader patient population. This limitation should be addressed in future research to ensure the applicability of the results to a wider range of patients.

Adherence to remote monitoring is a key determinant of its effectiveness, as highlighted in the article from Ziacchi et al.³² In the present analysis, adherence data were not systematically collected throughout the entire 5-year follow-up, which limits the ability to provide comprehensive adherence metrics. Nevertheless, this limitation has been acknowledged, and future investigations should aim to systematically capture adherence data, as their variability over long-term follow-up may offer valuable insights into the overall effectiveness of remote monitoring programs. Moreover, clinical information regarding LVEF and NYHA class was collected only at baseline. As a result, longitudinal data on the evolution of heart failure status are not available, which represents a limitation of the present analysis. Although complete longitudinal medication data were not available for this retrospective analysis, it is important to clarify that heart failure management at our centre has been organized through a dedicated heart failure outpatient clinic since before 2010, staffed by physicians specifically trained, with careful attention to guideline-directed medical therapy (GDMT). Throughout the study period, our team systematically implemented the core pillars of heart failure pharmacological treatment as soon as they were endorsed by successive ESC guideline updates. Even if quantitative data on therapy adjustments are not available, remote monitoring notifications frequently prompted remote clinical reassessment, often resulting in diuretic titration when signs of congestion were detected, while optimization of background GDMT was usually deferred until the subsequent visit at the heart failure outpatient clinic. The interaction between GDMT and remote monitoring is worth consideration. Contemporary heart failure management increasingly relies on the synergistic use of optimized pharmacological therapy and continuous physiological surveillance. Emerging evidence indicates that even low-dose angiotensin receptor-neprilysin inhibitor (ARNI) therapy may yield measurable improvements in physical activity levels and functional capacity in patients with HFrEF when assessed through longitudinal remote monitoring data.³³ These findings suggest that remote monitoring may serve not only as a tool for early detection of clinical deterioration but also as a sensitive instrument to quantify the functional response to GDMT titration, particularly in patients who are unable to tolerate full target doses. Integrating remote monitoring-derived metrics with pharmacological optimization may, therefore, enhance individualized treatment strategies, support earlier recognition of suboptimal therapeutic response, and ultimately contribute to improved clinical outcomes.

Conclusion

In conclusion, the analysis in a real-world setting aims to contribute valuable insights to the growing body of evidence regarding the long-term effectiveness of remote monitoring in heart failure patients with cardiac devices. These findings may help inform clinical decision-making and optimize the implementation of remote monitoring programs in similar healthcare settings. While promising evidence for the benefits of remote monitoring are shown, further research is needed to address the identified limitations and to explore innovative solutions that can enhance patient outcomes even further. The development of advanced predictive tools and personalized treatment strategies will be crucial in bridging the gap between ischemic and nonischaemic patients and in improving the overall effectiveness of remote monitoring.

Acknowledgements

This work was financially supported by Boston Scientific.

Conflicts of interest

There are no conflicts of interest.