A scoping review was chosen for this study to systematically chart and explore the existing evidence on economic evaluations within the context of RPM for chronic conditions. This review follows the Joanna Briggs Institute methodology for scoping reviews, which aligns with the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) guidelines [31]. The detailed protocol for this review has been registered in the Open Science Framework [32]. To investigate the quality of the reporting in the economic evaluation studies, we adhered to the CHEERS (Consolidated Health Economic Evaluation Reporting Standards) checklist [33].

The inclusion criteria include (1) studies targeting patients with CVD, CRD, DM, psychiatric disorders, and cancer as the focus of RPM; (2) studies conducting economic evaluation of RPM interventions, including CEA, CUA, CMA, CCA, CA, and BIA; (3) studies conducted in various health care settings, encompassing different levels of care, such as primary care, specialist care, and other relevant health care contexts. No geographical restrictions, beyond requiring English-language publications; and (4) empirical studies using randomized controlled trial design, pre-post studies, or observational designs.

The exclusion criteria include (1) studies not directly related to chronic conditions or RPM; (2) studies not involving economic evaluations, BIA, or cost aspects in the context of RPM for chronic conditions; (3) studies conducted exclusively in nonhealth care settings or unrelated to care delivery; and (4) exclude editorials, commentaries, opinion pieces, and studies lacking sufficient detail or relevance to the research question. We excluded modeling studies, even if they used empirical inputs, to focus directly on observed data.

This approach allows for a richer exploration of cost identification, measurement, and valuation in real-world RPM settings.

A research librarian assisted in developing and executing the search strategy across four electronic databases: Embase (Ovid), MEDLINE (Ovid), Web of Science (Core Collection), and CINAHL (Ebsco) in week 40 of 2023. The strategy combined relevant subject headings (eg, MeSH and Emtree terms) and free-text terms related to digital health (eg, telemedicine, telehealth, remote care, eHealth, mHealth), chronic conditions (eg, multimorbidity, CVD, and diabetes), and economic evaluations (eg, cost-effectiveness, cost-utility, and health economics). Reviews were screened for additional relevant references not identified through database search. The full search strategy for each database is provided in Multimedia Appendix 1.

The bibliographic references were imported into Rayyan, a web-based tool designed to facilitate the selection of studies for systematic review screening. The Rayyan tool is also useful for other literature review methods, like this scoping review. Before beginning the full screening process, the review team conducted a calibration exercise on a sample of 20 papers to ensure consistent interpretation of the inclusion and exclusion criteria. Titles and abstracts screening was then independently performed by two independent reviewers (SB and EB), applying the predefined eligibility criteria. The papers were included if they met the eligibility criteria. After the initial screening, selected papers underwent a full-text review, which was also conducted independently by both reviewers (SB and EB). Any discrepancies at either stage were resolved through discussion until a consensus was reached.

The data extraction process involved designing and creating an Excel (Microsoft Corp) form for consistency across the reviewers. The form included fields for title, authors, publication year, country, study design, methods, diagnosis, comparator, cost estimation details, outcome measures, intervention descriptions, and main findings. Extracting data was categorized according to four key themes: study characteristics (design, population, and setting), economic evaluation methods (types of analysis, comparator, perspectives, and outcome measures), cost estimation (identification, measurement, and valuation), and adherence to CHEERS. This thematic structure facilitated synthesis and comparison across studies.

Study quality was assessed using the CHEERS guidelines and checklist by SB and EB independently, with disagreements resolved by discussion.

Data synthesis was conducted using a thematic approach, systematically organizing the collected data. This process focused on categorizing and analyzing the findings of the included studies, identifying common themes or patterns across them. The synthesis comprehensively explored the various cost identification, measurement, and valuation methods used in economic evaluations of RPM for chronic conditions.

A total of 3049 records were identified in the search. Two of the papers refer to the same study but one as a subgroup analysis, so both were retained. The screening began with a title and abstract review, which excluded 2971 records not meeting the inclusion criteria, leaving 78 papers for full-text assessment. Seven records were inaccessible or unavailable and could not be retrieved, resulting in 71 full texts assessed against the inclusion criteria. Based on the full-text assessment, an additional 30 were excluded for various reasons (Figure 1). After exclusions, 41 papers, representing 40 studies, were included in the final review.

The included studies reveal a diverse range of medical conditions, perspectives, study designs, types of analysis, and outcome measures, summarized in Table 1. All the studies comparator, besides the BIA, are usual care.

The distribution of chronic conditions examined in the included studies shows that CVD was the most common focus in single-diagnosis studies, with 17 studies [36,37,39,41,46,47,49,50,54,55,64-66,68,69,74,75], followed by CRD, represented in 12 studies [35,40,44,51,53,56,61,62,67,71-73]. Within CRD, chronic obstructive pulmonary disease (COPD) was specifically addressed in 11 of these studies, and asthma in 1 study [71]. DM was the focus of 3 studies [43,60,70]. Additionally, 9 studies involved multiple chronic conditions, mostly CVD, COPD, or DM in the same study [38,42,45,48,52,57-59,63]. These studies span a broad range of countries, with the majority originating from high-income countries in Europe, North America, and Australia.

In this review, a variety of study designs and types of economic analyses are used across the included studies. Randomized controlled trials emerged as the predominant study design, accounting for 19 of the total studies included [41,45-48,53,57-60,62-73,75]. Moreover, matched control group designs were used in 12 studies [34,42-44,49,50,54,61,64-66,74], while 8 studies used a pre-post single-group design [36,38-40,52,55,56,63].

In terms of economic analyses, we classified studies based on the method of economic evaluation reported by the authors themselves, with adjustments made when methods were not explicitly reported. CEA emerged as the predominant approach, with 20 studies using this method [34,37,39,40,42,44,47,48,50,53,54,57,58,67-73,75]. Within CEA, 13 studies specifically conducted CUA [44,47,48,53,57,58,67-73,75]. The remaining studies showed diverse methodological approaches: 7 studies conducted CCA [36,38,41,45,46,55,59], 2 studies carried out cost-benefit analysis [43,66] and 3 studies conducted CMA [61-63], aimed at identifying interventions with equivalent effectiveness but differing costs. A total of 8 studies used CA methods [49,50,52,56,60,64,65,74], with 1 study focused exclusively on BIA [51].

Various economic outcome measures were reported across the studies, including mortality, QALYs, hospital readmission rates, and cost reductions. QALYs were a prominent measure in studies using CUA, with interventions for conditions like COPD and heart failure showing mixed results in cost-effectiveness. For instance, Stoddart et al [67] found a 15% probability of being cost-effective for COPD at £30,000 (US $40,685.41) per QALY, while Sydow et al [68] found heart failure interventions to be cost-effective when measured by QALY improvements. Other outcome measures included health care costs (eg, Herold et al [49] and Riley et al [65].

The second key aspect outlined in this scoping review is to describe cost identification, measurement, and valuation methods summarized in Table 2. More details about the cost categories can be found in Multimedia Appendix 2 [34-75].

The selection of cost components to include in each study depended on the specific perspective adopted and the focus of the economic evaluation conducted. The included studies considered a broad range of cost components. In the Danish studies, the perspective was that of the public payer, specifically the Danish public health care system, which included costs for prehospital services, in-patient and out-patient care, primary health care, prescription medicines, and intervention costs [69,72,72]. A total of 2 studies adopted a societal perspective, incorporating patient and family costs, productivity losses, and health care utilization and intervention costs [47,71]. Studies focusing on health insurance companies included costs relevant to the insurance plan, such as hospital services, ambulatory care, prescription medicines, and medical devices, while excluding costs like telemonitoring if not directly borne by the insurer [34,37,51,66,75]. German statutory health insurance studies, costs associated with hospital services, prescription medicines, and other covered interventions were included [34,49,66,68]. Studies from the perspective of hospitals or health services have primarily included costs related to hospital services, medical treatments, consultations, outpatient care, and interventions provided by the respective institutions [36,38,40,41,44,45,48,50,52,53,57,58,60,62,64,74]. Studies involving home health care agencies included costs associated with providing home-based health care services, such as home visits, medical treatments, and other related expenses [40,61,62,67]. In total, 10 of the studies did not report inclusion of the intervention costs [39,42,43,46,49,52,54,56,59,65].

Our synthesis identified variations in the data sources used for cost estimation and how they were used to derive cost values. We also observed a diverse array of terminology used in the publications to denote comparable sources. For example, many studies relied on electronic health records, patient administration systems, and hospital electronic databases as sources for cost measurement. Some studies reported health care utilization based on patient-recorded visits (eg, Apantaku et al [36] and Henderson et al [48]), while another study combined patient-recorded data with electronic patient records [47].

For cost valuation and unit cost estimation, studies used diverse approaches, including tariffs and official price lists. Diagnosis-related group tariffs were commonly used [58,61,69,72,73]. National or regional costing guidelines were applied, such as the Dutch manual for cost research [71], Official French national tariffs 2019 [64], and the Catalan Institute of Health official pricing [42]. Health insurance claims served as a cost source in multiple studies [37,49,52,60,68,75] and standard price weights were used in the UK context [67]. One study was identified as a BIA, using unit cost data and medication use from published studies and adjusted to align with the payer of the study population [51].

In addition to the diverse methods of costing observed, the reviewed studies applied both microcosting and gross-costing methodologies. Some studies provided detailed cost breakdowns, while others relied on aggregated estimates.

The studies varied in terms of intervention types, populations, and settings, but all were assessed based on their adherence to the CHEERS checklist items, as illustrated in Figure 2. The BIA study was not evaluated against the CHEERS checklist, as it is not applicable to this type of analysis. For more details, see the CHEERS checklist items (Checklist 1) and the study-specific CHEERS 2022 adherence table available in Multimedia Appendix 3.

Assessment of reporting quality using the CHEERS checklist revealed varying levels of adherence across different criteria. All studies (40/40, 100%) adequately described their study population and inclusion or exclusion criteria. Most studies (35/40, 88%) clearly identified relevant comparators and justified their chosen time horizons (100%). While all studies, except the BIA study, had usual care as a comparator, the CHEERS checklist applies a stricter definition of a comparator, which may account for differences in reporting.

However, several key reporting elements showed notable gaps. Only 23% (9/40) of papers reported their discount rate, while 60% (24/40) addressed the selection, measurement, and valuation of health outcomes. Critical methodological elements were often underreported; 50% (20/40) of studies detailed the dates of estimated resource quantities and unit costs, including currency and conversion information. Similarly, 55% (22/40) described methods to characterize uncertainty in their analyses. Less than half (18/40, 45%) of the studies reported how uncertainty about analytic judgments, inputs, or projections affected their findings, including choices regarding discount rates and time horizons.

The newer elements introduced in the CHEERS 2022 checklist received minimal attention across the reviewed studies. Few papers addressed items such as health economic analysis plans, characterization of heterogeneity and distributional effects, or stakeholder engagement during study design and implementation. This limited reporting likely reflects the timing of these studies, many of which were conducted before or shortly after the updated guidelines’ release.

This review highlights the variability in economic evaluation methodologies used in RPM studies and the inconsistent application of cost identification, measurement, and valuation methods. The included studies applied different approaches to defining and measuring costs, contributing to challenges in comparability.

Reported economic outcomes for RPM varied, with some studies identifying cost savings and improvements in health-related measures. This variation aligns with findings in previous reviews on DHIs [9-12,25,27,76-78].

Traditional health economic evaluation methods primarily focus on measuring health outcomes, often through metrics like QALYs, which may not fully capture the nonhealth and process outcomes relevant to DHIs, including RPM [24,26-28]. For instance, equity impacts, patient satisfaction, and process improvements are crucial, but often overlooked. Gomes et al [24]and Benedetto et al [26] have highlighted these limitations and proposed alternative approaches, such as impact matrices and CCA, which offer a broader view of the outcomes. The National Institute for Health and Care Excellence (NICE) similarly recommends considering both process and health outcomes in economic evaluations, pointing to examples such as reduced outpatient consultations as relevant process improvements [79].

A key challenge identified in the review was the inconsistent application of resource costing methods. The three main components, cost identification, measurement, and valuation, were often reported with limited transparency. This might bias results, make cross-study comparison difficult, and thereby, limit generalizability. In the field of health economics, accurately estimating resource costs is crucial for conducting robust CA and cost-effectiveness studies [14,23]. Resource costing methods are the primary costing approach for assessing the economic implications of health care interventions, diseases, and utilization patterns. The specific context of the economic evaluation further influences how resource quantities are determined. Cost identification involves assessing all relevant resources consumed, from direct medical costs to indirect costs like patient travel and productivity losses [14]. However, study perspectives varied, influencing which costs were considered.

Cost measurement approaches also differed, with some studies using the microcosting approach, a detailed assessment, and others relying on macrocosting, using aggregated cost data. The choice between these approaches depends on data availability and study objectives, though inconsistent application adds variability. The valuation step assigns unit prices to each resource, ideally reflecting opportunity cost, though practical factors often lead to using prevailing market prices and tariffs, further complicating consistency. Factors like regional price differences, choice of price weights, and whether charges or actual costs are used can affect generalizability and comparability across studies [14].

Despite these established principles, the practical application of resource costing in our review remains inconsistent. Many studies lacked transparency in cost identification, measurement, and valuation details, making comparisons difficult. Challenges also include adjustments for time (eg, discount rates) and inflation, as well as the transparency of data sources. Moreover, the absence of detailed tracking systems, particularly in settings with block funding, limits the precision of cost analyses. In such cases, reliance on broad estimates instead of granular data can affect the accuracy of cost evaluations and bias the results [14]. Additionally, task-shifting, delegating certain responsibilities to less specialized health workers, can influence cost outcomes but was underexplored in the reviewed studies.

The limited use of BIA across studies, seen in only one, points to gaps in assessing affordability alongside cost-effectiveness. While the International Society for Pharmacoeconomics and Outcomes Research acknowledges BIA as essential for a comprehensive economic evaluation [80,81], few studies implemented it, leaving a need for more comprehensive financial assessments.

Another limitation of the current evidence base is its exclusive focus on high-income countries. The absence of studies from low- and middle-income countries creates a knowledge gap, particularly given that these regions might benefit from remote monitoring solutions to address health care access challenges.

The updated CHEERS 2022 standards call for thorough reporting in economic evaluations, including cost-analysis plans and subgroup analyses. However, most of the studies in this review predated or only partly conformed to these updated guidelines. The inconsistent application of sensitivity analyses, time discounting, inflation adjustments, and discount rates further complicated comparisons across studies. Standardizing these practices would improve both transparency and comparability.

Finally, the practical aspects of conducting cost analyses in DHIs also require attention. For instance, accurately measuring health care professionals’ time spent on RPM activities can be challenging in the absence of standardized unit prices. Additionally, studies often failed to adjust for inflation or to use appropriate discount rates when presenting cost data, making comparisons between studies difficult. Future studies should aim for more transparent and detailed reporting on these practical aspects to enhance the reliability and generalizability of their findings.

This review’s strengths include its comprehensive search strategy, systematic approach to data extraction, and detailed quality assessment using the CHEERS checklist. However, several limitations should be noted. First, the focus on specific chronic conditions aligned with Norway’s implementation program may have excluded relevant studies in other conditions. Second, the heterogeneity in reporting styles and methodologies made direct comparisons challenging. Another limitation is the search date of this review, week 40 of 2023, leaving possible relevant studies published in 2024 out. Finally, the rapid evolution of digital health technologies means some newer implementation models may be underrepresented in the literature.

Future research should prioritize (1) the development of standardized reporting frameworks specifically adapted for DHIs, (2) investigation of RPM implementation in low- and middle-income countries, (3) integration of broader outcome measures beyond traditional health economic metrics, (4) increased use of budget impact analyses alongside cost-effectiveness studies, and (5) examination of long-term economic impacts and sustainability of RPM programs.

Policy makers play a crucial role in fostering the advancement and adoption of evidence-based RPM strategies and they should actively support research initiatives addressing the identified knowledge gap, particularly in areas such as standardized reporting, BIA, and implementation in diverse health care settings. Established funding mechanisms for rigorous economic evaluation will also incentivize researchers to conduct high-quality studies that inform decision-making. Furthermore, initiatives should be put in place to foster collaboration among researchers, health care providers, and relevant stakeholders to facilitate the translation of research findings into real-world practice.

This review underscores the variability in economic evaluations of RPM and the inconsistent reporting of cost identification, measurement, and valuation. The limited use of BIA and the lack of studies from low- and middle-income countries highlight gaps in the literature.

There is an increasing acknowledgment of the need for more comprehensive methodological approaches, including the integration of process outcomes and BIA. Furthermore, adherence to updated reporting standards like CHEERS 2022 is still limited, with most studies not including key elements like sensitivity analyses or detailed cost estimation practices.

Inconsistent application of time discounting, inflation adjustments, and resource costing principles, along with a lack of transparency in cost data, further complicates the interpretation of findings. Establishing standardized, transparent reporting protocols will advance this field and significantly enhance the comparability and generalizability of economic impact assessments for DHIs.