IJNHCR

Introduction

Non-communicable chronic health conditions (NCHC) have emerged in recent decades as the predominant challenge facing global healthcare systems, with over 40 million people dying annually from NCHC’s such as cardiovascular diseases, cancer, diabetes and chronic respiratory diseases [1]. Healthcare systems designed for brief interventions are unprepared for the complex, ongoing burden that chronic conditions impose [2]. Type 2 diabetes alone affects 537 million adults globally. It is projected to reach 783 million by 2045, a 46% increase equivalent to 12 million additional diabetic patients annually, demonstrating the scale and urgency of this healthcare challenge [3]. Individuals and families affected by chronic health conditions contend with a reduction in quality of life, diminished earning capacity, shortened life expectancy and the emotional and physical toll of caring for loved ones living with chronic conditions [4]. The COVID-19 pandemic accelerated healthcare digitalisation, with telemedicine consultations increasing by 3,800% during lockdowns [5]. In practice, digital health technologies have evolved from simple monitoring devices into comprehensive systems that enable reliable, continuous patient engagement and real-time clinical decision-making [6].  Healthcare providers are using this digital evolution to strengthen patient relationships, provide education and identify health changes in real-time, resulting in improved outcomes for chronic disease management [7]. Systematic review evidence demonstrates that remote monitoring systems produce measurable benefits, with nearly half of studies (49%) showing reductions in hospital admissions and 41% of studies reporting decreased emergency department visits, along with improvements in disease-specific quality-of-life measures [8].

Maintaining quality patient care continues to be a challenge in the Irish healthcare system, with increasing demands due to hospital capacity challenges and chronic disease management [9]. The Health Service Executive’s Chronic Disease Management Programme targets four primary conditions: Type 2 diabetes, asthma, chronic obstructive pulmonary disease (COPD), and cardiovascular disease, which collectively affect over one million Irish residents [10]. Among adults over 65 years, 64.8% live with multiple chronic conditions, creating complex care coordination challenges on the healthcare infrastructure [10]. Emergency department presentations have increased by 6% annually, directly contributing to hospital overcrowding and reduced bed availability [11]. Capacity constraints are evidenced by Irish Nurses and Midwives Organisation data, with more than 120,000 patients requiring trolley-based care in 2024, and many individuals experiencing extended waits that negatively impact clinical outcomes [12]. Against this backdrop of healthcare pressures, digital health solutions such as virtual wards offer potential relief. Ireland’s experience with digital health acceleration during COVID-19 provides valuable context for virtual ward implementation. The HSE facilitated over 400,000 video-enabled consultations since 2020, with evaluations showing 95% of patients likely to recommend virtual consultations to others [5]. Health Service Executive’s Telehealth Roadmap 2024-2027 identifies the following implementation challenges: broadband connectivity limitations, digital literacy and health information systems [13]. Virtual wards provide hospital-level care to patients at home through remote monitoring technology and digital clinical oversight [14]. Patients are formally ‘admitted’ to virtual beds and receive structured care protocols while remaining at home And for clinicians, this model offers simultaneous and safe monitoring of patients Virtual ward care is delivered through digital dashboards that support technology-enabled clinical decision-making [15]. Despite their increasing implementation, significant evidence gaps limit informed decision-making about the effectiveness and value of virtual wards, with a variation in study methods and outcomes, making it difficult to assess [16]. Research into virtual ward implementation suggests areas such as team structures and optimisation of healthcare outcomes are not fully understood [17]. In practice, the lack of understanding of nursing workload and patient safety presents a real problem when implementing virtual ward-based care and is separate from digital challenges [18]. Ireland launched two virtual wards at St Vincent’s and University Hospital Limerick in 2024; moving forward, evidence from these sites will inform nursing practice and patient care nationally [19]. It is anticipated that virtual wards may create capacity that can reduce the number of patients admitted annually without beds [19].

Aims

This systematic review aims to evaluate the effectiveness of digital health interventions in chronic disease management, with particular attention to conditions where robust evidence exists. Specifically, this review seeks to: (1) examine clinical effectiveness of digital health technologies in chronic disease management, with primary focus on COPD and cardiovascular disease where evidence is strongest; (2) assess patient-reported outcomes and acceptability across demographic groups; (3) identify available economic evidence and highlight critical gaps for policy consideration; and (4) contextualise findings for Irish healthcare settings, acknowledging evidence limitations while providing strategic insights for digital health implementation.

Materials and methods

This systematic review was conducted according to Cochrane principles  [20]. The study flow was reported in accordance with PRISMA guidelines [21], with the study research questions published on Prospero: https://www.crd.york.ac.uk/PROSPERO/ view/CRD420250650909 prior to commencing the review to ensure methodological rigour and transparency.

Research question

This systematic review addressed the critical question: Do digital health platforms provide clinically effective and economically viable solutions for chronic disease management? This primary research question examined: (1) clinical effectiveness of digital health platforms compared to standard care for chronic disease management, and (2) economic value, including cost-effectiveness and healthcare resource utilisation impact. Secondary questions explored the healthcare system impact on hospital admissions and emergency visits, and implementation factors influencing platform adoption and sustainability.

Eligibility criteria

Study eligibility was determined using the PICO framework:

Population: Adults aged ≥ 18 years with chronic health conditions, encompassing Type 2 diabetes, asthma, Chronic Obstructive Pulmonary Disease (COPD), cardiovascular disease (heart failure, coronary artery disease, stroke), and multiple chronic conditions.

Intervention: Digital health technologies comprised remote monitoring devices (physiological monitoring, disease-specific sensors), telemedicine/telehealth platforms (video consultations, virtual wards), mobile health applications (medication management, symptom tracking), hospital-at-home programs, telerehabilitation systems, and integrated digital care platforms.

Comparison: Acceptable control groups encompassed standard/ usual care, traditional face-to-face healthcare delivery, non-digital interventions, waitlist controls, and alternative digital interventions for comparative studies.

Outcomes: Three primary outcome categories were examined: (1) Clinical effectiveness including quality of life measures (diseasespecific and generic), functional capacity, clinical parameters (HbA1c, blood pressure, pulmonary function), and symptom control; (2) Economic outcomes encompassing cost-effectiveness ratios, healthcare utilisation (hospital admissions, emergency department visits), healthcare costs and savings; (3) Implementation outcomes including patient acceptance, technology adoption rates, and sustainability factors.

Exclusions: Paediatric populations (<18 years), systematic reviews and meta-analyses, conference abstracts, non-English publications, studies lacking quality of life or economic outcomes, acute care interventions without a chronic disease focus, and nondigital technologies.

Information sources and search strategy

 Four electronic databases were systematically searched: PubMed, EMBASE, CINAHL, and Scopus, spanning 10 years from January 1, 2015, to February 11, 2025. The search strategy was developed and validated by an information scientist at the Royal College of Surgeons in Ireland, University of Medicine and Health Sciences, employing four complementary concept groups combined with Boolean operators: (1) Economic evaluation terms (“cost effectiveness,” “economic evaluation,” “health economics”), (2) Digital health technology terms (“digital health technology,” “telemedicine,” “remote monitoring,” “virtual ward”), (3) Quality of life terms (“health-related quality of life,” “patient reported outcome,” “QALY”), and (4) Chronic disease terms (“chronic disease,” “long term condition”). Search parameters included humans only, English language, adults (≥18 years), and peerreviewed journal articles only.

Study selection process

All identified records were imported into Covidence systematic review software with automatic duplicate detection [22]. After duplicate removal (n = 1,746), 1,223 records underwent rigorous two-stage screening:

Stage 1: Two reviewers (AD, LHS) independently screened 100% of titles and abstracts using predefined inclusion/exclusion criteria with an inclusive approach to minimise risk of excluding relevant studies. Records excluded (n = 1,135): non-chronic disease focus (n = 487), paediatric populations (n = 298), non-digital interventions (n = 245), wrong publication type (n = 105).

Stage 2: Full-text assessment of 88 articles by two reviewers independently evaluated against full inclusion/exclusion criteria, with third reviewer (ML) available for unresolved conflicts. Fulltext exclusions (n = 65): insufficient outcome data (n = 28), wrong study design (n = 20), language barriers (n = 17).

Inter-rater agreement achieved 94.8% with κ = 0.478, indicating moderate reliability. Twenty-three studies with 4,228 participants were included: 7 randomised controlled trials, 5 non-randomised trials, 6 cohort studies, 4 qualitative studies, and 1 economic evaluation (Supplementary 1).

Data extraction

A comprehensive, standardised data extraction form was developed based on Cochrane Handbook recommendations [20]. Extracted elements included study characteristics (design, setting, demographics, sample size, follow-up duration), intervention details (technology type, implementation approach, healthcare provider involvement, intervention duration), outcome measures (primary/secondary outcomes with specific measurement instruments, quality of life assessments using validated tools, economic outcomes, clinical parameters), and methodological quality indicators (randomisation methods, sample size calculations, follow-up completion rates). Data extraction was performed by the lead reviewer (AD) and verified by a second reviewer (LHS) for 100% of included studies, with disagreements resolved through discussion and consensus with a third reviewer (ML).

Quality assessment

Study quality was assessed using appropriate Joanna Briggs Institute (JBI) critical appraisal tools [23] matched to specific study designs: RCTs (n = 7), non-randomised trials (n = 5), cohort studies (n = 6), qualitative studies (n = 4), and economic evaluations (n = 1). Quality assessment was conducted independently by two reviewers using standardised forms, with studies categorised as high quality ( ≥ 70% criteria met), moderate quality (50-69%), or low quality (< 50%). Results informed data synthesis and interpretation of findings. 

Data synthesis and analysis

Due to substantial heterogeneity in study populations, intervention types, healthcare settings, and outcome measures, meta-analysis was not appropriate. Narrative synthesis was employed using a structured framework incorporating thematic organisation around clinical effectiveness, economic impact, healthcare utilisation, and implementation factors. Subgroup analysis examined findings by chronic condition type, digital health technology category, healthcare system context, and study design quality. Clinical effectiveness outcomes were evaluated against established minimal clinically important difference thresholds to ensure reported benefits translate to meaningful patient improvements. Implementation outcomes were analysed using the RE-AIM and Consolidated Framework for Implementation Research frameworks.

Methodological considerations

 Important limitations included potential publication and language bias from focusing on English-language peer-reviewed publications, geographic bias with predominance of studies from high-income countries, limited long-term follow-up in most studies (78% ≤ 12 months), and substantial intervention heterogeneity limiting direct comparison. Methodological strengths encompassed prospective protocol registration, comprehensive search strategy with expert validation, rigorous dual independent review process with excellent inter-rater reliability, quality assessment using validated designspecific tools, and systematic application of clinical significance thresholds.

Results

Study selection

 PRISMA flow diagram (Figure 1) illustrates the study selection process. After screening 1,223 records, 23 studies met the inclusion criteria and were included in this review.

.

Figure 1: PRISMA diagram showing the flow of studies through the systematic review.

Table 1 presents the distribution of study designs among the 23 studies included in this systematic review, demonstrating a diverse methodological approach with RCTs comprising the largest proportion (30.4%), followed by cohort studies (26.1%) and nonrandomised controlled trials (21.7%), while economic evaluations represented a notable evidence gap with only one study (4.3%).

Table 1: Study design distribution of included studies (n = 23).

This study used both experimental and observational research methods to build a comprehensive evidence base. The experimental studies helped determine what interventions work, while the observational studies showed how these interventions perform in real clinical settings. This combination of methods strengthened the overall findings; however, the unequal distribution of study types created some limitations for interpreting results. Specifically, only 30.4% of studies were randomised controlled trials, and just 4.3% included economic evaluations. This means we have limited evidence for proving direct cause-and-effect relationships between interventions and outcomes. We also lack sufficient cost-effectiveness data to guide healthcare administrators and policymakers in making informed decisions about resource allocation. The demographic characteristics of the 4,228 participants across all 23 studies are summarised in Table 2.

Table 2: Demographic characteristics of participants.

Clinical characteristics related to specific diseases are shown in Table 3

Table 3: Patient characteristics.

Geographic distribution

Studies were conducted across 12 countries: Europe (15 studies, 65%), North America (5 studies, 22%), Asia-Pacific (1 study, 4%), and South America (1 study, 4%). This concentration in European healthcare systems enhances relevance for Ireland’s mixed public-private model but may limit generalisability to other regions.

Chronic condition focus

The predominance of COPD studies (52%) reflects the maturity of digital respiratory health solutions, while the limited evidence for diabetes (4%) and asthma (4%) represents significant gaps given the global burden of these conditions.

Study quality assessment 

Quality assessment using validated JBI tools revealed 91% of studies (21/23) achieved moderate or high-quality ratings, providing a strong methodological foundation for evidence synthesis as illustrated in Table 4.  

Table 4: Quality assessment summary by study design (n = 23).

High quality studies (n = 6)

The six high-quality studies included Arnaert et al. qualitative study, achieving perfect methodology scores of 10/10 (100%); [24] Vestergaard et al. non-randomised controlled trial scoring 9/9 (100%);[25] Messori et al. economic evaluation with 9/11 (82%);[26] Fredman et al. cohort study achieving 9/11 (82%), [27] Schäfer et al. non-randomised controlled trial with 7/9 (78%), [28] and Koff et al. non-randomised controlled trial scoring 7/9 (78%) [29].

Moderate quality studies (n = 15)

Moderate quality studies comprised Vianello et al. [30] LopezVillegas et al. [31] Kamei et al. [32] Garcia-Carretero et al. [33] Timmermans et al. [34] Codina et al. [35] Bernocchi et al. [36] Kardas et al. [37] Bogacz et al. [38] Hamad et al. [39] O’Leary et al. [40] Korsbakke et al. [41] Sten-Gahmberg et al. [42] Poureslami et al.  [43] Barken et al. [44] and Mathar et al. [45]

Low quality studies (n = 2)

Only two studies achieved low quality ratings, specifically Cerdán et al. [46] and Poureslami et al. [43].

Quality by study design

Randomised controlled trials achieved moderate quality with average scores of 62 % (range: 54-69 %). All RCTs used appropriate randomisation methods, though blinding participants was not possible due to the nature of the interventions. Non-randomised controlled trials performed slightly better with average scores of 67 % (range: 56-100%), and 60% of these studies achieved high quality ratings.

Cohort studies showed the strongest performance with average scores of 68% (range: 55-82%). The most common weakness in cohort studies was incomplete reporting of participant follow-up data. Qualitative studies had the most variable quality, averaging 58% (range: 30-100%), indicating inconsistent methodological rigour across studies. The single economic evaluation received a high-quality score of 82%. The study used sound methods for analysing costs and benefits but did not adequately test how reliable its cost estimates were under different scenarios.

Digital health technology types and implementation

Technology categories

Remote Monitoring Devices represented 9 studies (39%) including Codina et al. [34] (Cardio MEMS) cohort study, Timmermans et al. [34] (ICD monitoring) cohort study, Hamad et al. [39] (COPD telemonitoring) cohort study, Kardas et al. [37] (wireless sensors: glucometer, BP, weight, activity tracker) RCT, Lopez-Villegas et al. [31]  (Bio-tronik Cardio Messenger for pacemakers) RCT, Sten- Gahmberg et al. [42] (monitoring devices with tablets) RCT, Bernocchi et al. [36] (pulse oximeter, portable ECG) RCT, Vianello et al. [30] (home respiratory monitoring)

RCT, and Vestergaard et al. [25] (remote monitoring system) nonRCT.

Telemedicine/Telehealth Platforms encompassed 8 studies (35%) including Arnaert et al. [24] (integrated telenursing) qualitative study, Barken et al. [44] (telemedicine monitoring) qualitative study, Mathar et al. [45](Tele-Video consultations) qualitative study, Poureslami et al. [43] (electronic action plans with SMS) RCT, Korsbakke et al. [41] (tele kit with education) RCT, Schäfer et al. [28] (telehealth platform) non-RCT, Koff et al. [29] (telemedicine system) non-RCT, and Kamei et al. [1, 3, 32] (telehealth intervention) non-RCT.

Mobile Health (mHealth) Applications comprised 3 studies (13%), including Kardas et al. [37] RCT using a smartphonebased medication management system; Poureslami et al. [43] and O’Leary et al. cohort study of a mobile integrated health program [40].

Hospital-at-Home Programs included two studies (9%) Fredman et al. cohort study of video-enabled transitional care; [27] and Garcia-Carretero et al. observational study of traditional hospital-at-home with outpatient parenteral antimicrobial therapy [33].

Telerehabilitation Systems represented 1 study (4%) with Bernocchi et al. with Bernocchi et al. RCT of a home-based exercise program with remote coaching [36].

Economic Evaluation constituted 1 study (4%) with Messori et al. cost-effectiveness analysis of cardiac monitoring systems [26]. Some studies appear in multiple categories due to multi-component interventions, such as Bernocchi et al. study, which used both remote monitoring devices and telerehabilitation components [36].

Key intervention examples

Diabetes management

The mHealth system included multiple wireless sensors (blood glucose monitor, blood pressure cuff, weight scale, and activity tracker) with smartphone integration, 24-hour emergency contact capability, and continuous monitoring over 12 months.

COPD and heart failure telerehabilitation

Home-based telerehabilitation utilised a mini-exercise bike, pedometer, pulse oximeter, and portable ECG device. Patients received weekly structured telephone calls and multidisciplinary team support over 4 months, with follow-up continuing for 6 months.

Cardiac device management

Remote monitoring involved pacemaker telemetry with continuous surveillance, automated clinical alert systems, and a 12-month monitoring duration. Clinical outcomes that demonstrate minimal clinically important differences (MCID) are presented in Table 5.

Clinical outcomes

Analysis of primary outcomes across included studies demonstrated that 86% of measured clinical endpoints (12/14) exceeded established minimal clinically important difference (MCID) thresholds or achieved clinical significance.

Disease-specific clinical improvements varied across conditions. For COPD management, digital health interventions achieved an 11.2-point improvement in St. George’s Respiratory Questionnaire (SGRQ) scores, exceeding the 4-point MCID with large effect sizes (Cohen’s d = 0.82, p < 0.05) and showing consistent benefits across multiple studies [30,33,39,41]. Dyspnoea symptoms improved significantly on the MRC Dyspnoea Scale, achieving clinically relevant improvements [36].

Heart failure outcomes demonstrated substantial quality of life improvements with 16.5-point reductions in Minnesota Living with Heart Failure Questionnaire (MLHFQ) scores, exceeding the 5-point MCID threshold by 3.0-3.6-fold [26,34]. For combined COPD and heart failure conditions, telerehabilitation achieved a 60-meter improvement in 6-minute walk test performance, exceeding established MCID thresholds for clinically meaningful functional improvement [36].

Diabetes management through comprehensive mobile health interventions yielded clinically significant improvements in glycaemic control. These included 0.8% HbA1c drop exceeding clinical targets, 1.5 mmol/L reduction in fasting glucose, and 20% improvement in medication adherence [37]. Functional and quality of life outcomes across multi-condition interventions achieved meaningful disability reduction on the Barthel Index and statistically significant improvements in physical activity scores (PASE) and time to clinical events [32,40,42].

This evidence demonstrates that digital health interventions consistently achieve not only statistical significance but also clinically meaningful improvements across diverse chronic conditions and outcome measures. Healthcare utilisation outcomes showed significant improvements across 12 studies, with reductions in hospitalisations and emergency department visits detailed in Table 6.

Table 6: Healthcare utilisation summary.

Healthcare utilisation analysis

 Digital health interventions achieved clinically meaningful reductions in acute care utilisation across all 12 studies, with 92% reaching statistical significance. COPD studies (n = 7) showed consistent reductions: 42% fewer hospital admissions and 39% fewer ED visits. Heart failure studies (n = 3) demonstrated 50% reduction in admissions and 43% in ED visits, with Cardio-MEMS achieving a 77% admission reduction. Other conditions showed 30-40% reductions across both outcomes.

This evidence demonstrates that digital health interventions consistently achieve clinically meaningful reductions in costly acute healthcare utilisation across diverse chronic conditions.These substantial decreases in hospital admissions and emergency department visits (Table 5) were enabled by comprehensive technology deployment and strong user engagement across all interventions detailed in Table 7

Table 7: Patient experience and technology acceptance outcomes.

Patient experience and technology acceptance

Age and technology adoption demonstrated that digital health interventions achieved remarkably high adoption rates of 85-95% across all age groups, challenging common assumptions about older adults and technology use. Age-related adoption patterns showed that patients over 75 years achieved 85-90% successful implementation with enhanced support, patients aged 55-75 years demonstrated greater than 95% adoption, representing the optimal demographic, and patients aged 18-54 years showed 90-95% adoption with the highest long-term retention rates. Sustained engagement remained strong with greater than 80% continued use at 6-month follow-up across all age groups, with older adults achieving equal or superior clinical outcomes when provided appropriate support  [34,36,26].

While patient adoption was consistently high, economic evaluations were limited across studies, with only two high-quality costeffectiveness analyses identified [26] and [30], supplemented by modelled estimates from healthcare utilisation data, as detailed in Table 8.

Table 8: Economic outcomes and cost-effectiveness analysis; *Net cost (higher costs in the intervention group).

Economic outcomes

Showed mixed results based on limited high-quality economic evidence from two studies. The Cardio MEMS economic evaluation conducted in Italy demonstrated an incremental cost-effectiveness ratio (ICER) of €38,435 per quality-adjusted life year (QALY), suggesting potential cost-effectiveness depending on the threshold applied [26]. This intervention provided a clinical benefit of 0.40 QALYs per patient over four years.

The pacemaker telemonitoring economic analysis from Norway showed less favourable results, with an ICER of €53,345 per QALY from the NHS perspective [30]. Higher costs were observed in the telemonitoring group (€2,079.84 versus €271.97 in the control group), leading to the conclusion that this intervention was not cost-effective due to ICER values above usual NHS thresholds. The limited economic evidence highlights the need for comprehensive cost-effectiveness analyses across broader digital health interventions to inform healthcare policy and resource allocation decisions. While economic evidence remains limited, patient experience and satisfaction demonstrated consistently high ratings across all digital health interventions, with usability and preference metrics detailed in Table 9.

Table 9: Patient experience and satisfaction metrics.

System usability and satisfaction

Usability exceeded acceptability thresholds with an average score of 72.3 (above 70-point acceptability threshold). Despite 53% of ICD patients experiencing technical difficulties (device connectivity, charging issues), satisfaction remained high (median 9/10), indicating that clinical benefits outweighed technical limitations [34]. Technology reliability was identified as the primary area for improvement across studies.

Patient preferences and engagement

Remote monitoring showed strong preference patterns with 43% preferring remote vs 19% in-clinic monitoring [34]. SMS engagement rates reached 68.4% in asthma patients [42], while online platform usage varied, with only 28% accessing web-based action plans in some studies. Follow-up duration varied significantly across studies, with sustainability and long-term effectiveness patterns detailed in Table 10.

Table 10: Follow-up duration and sustainability analysis.