The aging world population and an increased prevalence of chronic conditions and multimorbidity have placed greater pressure on health care systems and professionals [1]. In Europe, over 50 million people have at least 1 chronic disease [2]; and in the United States, it is estimated that the number of people aged 50 years and older with at least 1 chronic disease will increase by 100%, from 71 million in 2020 to 142 million by 2050 [3]. The incidence of life-threatening falls among older adults (aged ≥70 y) continues to rise globally [4]; together with other age-related medical emergencies, such as stroke [5], these events claim millions of lives each year [5,6]. The notable changes in global demographics, combined with the prevalence of frailty among older adults, underscore the importance of expanding the human resources available in the public health sector. Nevertheless, due to economic and occupational constraints, this may not be a feasible solution [7]. To alleviate the pressure on health care systems, more sustainable initiatives need to be implemented. One potential solution often discussed in the literature is the transition to home-based care (HBC) [8,9]. HBC covers a wide continuum of care and involves delivering increasingly complex health care services to individuals in their own residences, allowing them to maintain their independence as an alternative to relying on residential, long-term, or institutional nursing care [10,11]. This transition is often supported by older adults because the majority choose to remain in their own homes for as long as possible [12,13].

To support the transitioning to HBC, there is a growing demand for new technological innovations and collaborative initiatives [14]. This often involves modifications of peoples’ homes and the use of medical equipment that not only facilitates long-term health monitoring but also enables the management of medical emergencies that demand rapid medical attention [15-17]. This is particularly important in the case of older adults (aged ≥65 y) because their inability to manually initiate an emergency alarm after a sudden deterioration in physical condition (eg, acute stroke or hip dislocation after a fall) may lead to long-term consequences and increased mortality [15]. Studies have shown that Internet of Medical Things (IoMT; Figure 1) platforms have the potential to support the transitioning to HBC by streamlining workflows, reducing costs and time delays, and improving patient well-being [18-20]. While previous research has demonstrated the usefulness of stand-alone platforms for the monitoring of chronic conditions and detection of abnormalities, there has been limited focus on achieving higher levels of interoperability within IoMT [21-24]. Although stand-alone platforms may be feasible in certain situations, the lack of interoperability with external systems such as electronic health records (EHRs), emergency medical services, and public safety answering points (PSAPs) may lead to data becoming trapped within silos, potentially resulting in delayed information transfer, suboptimal decision-making, and patient harm [7,21,25,26]; for example, a study conducted by Magrabi et al [27] showed that approximately 20% of reported patient safety hazards were linked to deficient information transfer. Furthermore, Hyvämäki et al [26] showed that inadequate documentation and use of information in HBC plays a significant role in interorganizational health information exchange–related incidents, resulting in, for example, delayed care and patient harm.

At Chalmers University of Technology in Gothenburg, Sweden, the Care@Distance research group is active in the field of remote and prehospital digital health. The group is dedicated to improving health care delivery through interoperable cutting-edge solutions, encompassing clinical decision support systems, artificial intelligence (AI), modern IT, and innovative user interactions. While interoperability offers numerous advantages, we acknowledge that current medical data still contain nonstandard elements [28,29], and vendors continue to use their own proprietary solutions [29]; for example, most inpatient EHRs include over 5000 variables [28,29], making it difficult to ensure a shared understanding of medical concepts across domains and organizations [29]. One system may code “A fall on and from stairs and steps” as W10 in the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM), while another system uses Systematized Nomenclature of Medicine–Clinical Terms (SNOMED CT) code 900000000000448009 for “Fall on stairs.” Establishing a shared understanding of these 2 terms requires a translation [21]. As the number of unique systems and standards increases, the need for separate translations also increases, especially considering that SNOMED CT includes over 340,000 medical concepts (clinical findings, procedures, substances, etc) [29]. Due to these complex translation processes within health care, involving various file formats (eg, text, video, images, and audio), communication protocols, and semantics, over 80% of all medical data tend to be overlooked or discarded [19]. This not only hampers communication between systems but also limits the use of AI, international cooperation, and research [29]. As the technological landscape expands, navigating among the available technologies and standards is becoming increasingly challenging. Hence, in this study, we summarize existing knowledge and best practices to overcome interoperability issues in IoMT platform development to manage medical emergencies within HBC and prehospital care settings.

This scoping review aims to summarize and map the enabling technologies that can be used to develop interoperable platforms to manage medical emergencies in HBC and prehospital care. We have proceeded from a 6-level interoperability model comprising device, network, syntactical, semantic, cross-platform, and cross-domain interoperability and a 5-level IoMT reference model comprising perception, transport, processing, application, and business layers [11]. These models provide adequate granularity and context to be applied within the context of IoMT [10]. Relevant actors and domains within the IoMT domain include homes, sensor providers, emergency medical services, PSAPs, social security services, and hospitals. The aim is to describe technologies and their use in an accessible way, enabling cross-disciplinary discussion between clinicians and engineers. This scoping review can potentially serve as a guide for software developers, clinicians, and other practitioners aiming to develop interoperable IoMT platforms.

This scoping review follows the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines for scoping reviews (refer to Multimedia Appendix 1 for the PRISMA-ScR checklist) [30] and adheres to the methodology outlined in the research protocol [11]. Given the broad spectrum of research questions addressed in this scoping review, the methodology includes 4 distinct search strategies, referred to as strategies A to D in the research protocol [11]. The relationships between these search strategies are illustrated in Figure 2.

This approach allows each research question to be addressed more systematically by separating the search terms, search periods, and goals within each search strategy. Search strategies A and B were completed with the research protocol [11] (Figure 2) and addressed the following research questions:

These 4 research questions were addressed in the research protocol [11], resulting in the definition of a 6-level interoperability model and an IoMT reference model to be used as reference materials in this scoping review. On the basis of these definitions, this scoping review will proceed with search strategies C and D (Figure 2), focusing on answering the following research questions:

Research questions 1 and 2 are addressed in search strategy C, and research question 3 is addressed in search strategy D in this review.

Six electronic databases were used in this study: IEEE Xplore, PubMed, Scopus, ACM Digital Library, Sage Journals, and ScienceDirect [11]. In the published protocol, we had proposed using the Google Scholar database, but we replaced it with the ACM Digital Library in this review to ensure higher precision in search results [31,32] (refer to Multimedia Appendix 2 for the search terms used for search strategies C and D). The retrieved articles were assessed for inclusion based on predefined eligibility criteria and underwent a 2-step screening process using the web application Rayyan (Rayyan Systems Inc) [33]. The screening was performed by HJ and MS, and any disagreements were resolved through discussion between them [11].

The eligibility criteria were originally defined in the research protocol [11]. However, a minor adjustment was necessary to streamline the screening process and enhance clarity (Textbox 1). The last inclusion criterion was updated to emphasize a stronger focus on technologies that address interoperability issues.

In search strategy C, data were extracted to map interoperability solutions to their respective interoperability models, categorizing them based on the specific level of interoperability addressed. The mapping was based on the 6-level interoperability model and the 5-level IoMT reference model established in the research protocol [11]. Similar mappings that could be found in the literature acted as a reference and, combined with the expertise of our research group, were used to validate the mappings.

In search strategy D, previous efforts in IoMT platform development for prehospital care and HBC were examined. A thematic analysis approach was used to identify and summarize the technologies and standards used in these development processes (Textbox 2). This analysis was performed by 1 reviewer (MS) using Microsoft Excel (version 2403) and the web application Dedoose (version 9.0.107; SocioCultural Research Consultants LLC). The thematic analysis was based on the framework outlined by Braun and Clarke [34]. The framework was adapted to this study and included five steps:

From the thematic analysis, a list of commonly used technologies was compiled. Each technology was accompanied by a descriptive overview of its application area and function, along with a mapping to one of the levels in the interoperability model defined in the research protocol [11]. For each study, the main areas of interest were identified, such as citation details (eg, author, year of publication, and country of origin) and key study characteristics (eg, type of IoMT platform, platform requirements, interoperability challenges, and the technologies and standards used to achieve different levels of interoperability). The results were summarized using narrative descriptions as well as figures and tables. The data extraction process ensured that sufficient data were gathered to answer each research question.

The study selection flowchart is presented in Figure 3. A total of 2949 articles were identified (n=2306, 78.2% retrieved through search strategy C and n=643, 21.8% retrieved through search strategy D). After screening and snowballing, 158 articles were included in this study (n=84, 53.2% from strategy C, n=29, 18.4% from strategy D, and n=45, 28.5% from snowballing). Of the 33 articles excluded after title and abstract screening, 7 (21%) were left out because they were in non-English languages.

This section is divided into 5 subsections (Figure 4). In the first, we provide an overview of the technologies and standards applicable to IoMT, emphasizing their relevance in addressing various levels of interoperability, as introduced in the research protocol [11]. In the second subsection, we present common interoperability requirements identified in the included studies, focusing on their application in prehospital care and HBC scenarios. Next, in the third subsection, we explore the interoperability challenges associated with IoMT platform development. In the fourth subsection, we summarize previous efforts in IoMT platform development and the technologies and standards used in these initiatives. Finally, in the fifth subsection, we draw upon our research findings to present actionable insights and recommendations for addressing interoperability challenges.

In the following subsections, we have categorized and summarized the enabling technologies based on the specific interoperability level they address (refer to Multimedia Appendix 2 for a description of the enabling technologies).

IoMT platforms for prehospital care and HBC have specific requirements. These platforms are required to operate in time-critical environments, where large amounts of sensitive data need to be exchanged in real time across various domains. In this subsection, we have summarized common interoperability requirements identified in the included studies and grouped them based on their potential application in IoMT settings (Table 1).

Achieving interoperability is not a trivial task. The higher the interoperability level, the more complex the endeavor due to the involvement of additional technologies and standards. Previous research has reported recurring interoperability challenges frequently encountered in IoMT platform development for prehospital care and HBC. Some of these challenges are reported in Table 2.

Among the examined studies (n=25), we could identify 7 frequently reported challenges associated with interoperability. Among the 7 reported challenges, 12 (21%) were related to latency problems, 8 (14%) to proprietary protocols, 13 (23%) to the volume and complexity of data, 12 (21%) to privacy and security concerns, 8 (14%) to semantic coding issues and 4 (7%) to poor connectivity.

In this section, we provide an overview of the technologies and standards used frequently to address interoperability challenges in IoMT platform development for prehospital care and HBC (Table 3).

As can be seen in Table 3, IoMT platform development often requires a combination of technologies and standards. Simply specifying “Bluetooth,” “Wi-Fi,” or “SNOMED CT” is not sufficient because no single standard, protocol, or technology can address higher levels of interoperability [100]; for example, SNOMED CT can be complemented by more domain-specific standards such as LOINC [29], while Bluetooth can be used in conjunction with a gateway, MQTT, and FHIR to enhance interoperability within IoMT [19]. Table 2 reveals variations in the frequency of the technology and standards used. Consequently, we have summarized the most used technologies and standards in Figure 8. Technologies and standards that appeared in 2 or fewer articles were excluded from Figure 8. These include Web Real-Time Communication (WebRTC), openEHR, voice over IP, webhooks, and Substitutable Medical Applications, Reusable Technologies (SMART) on FHIR, as well as the interoperability guidelines and frameworks such as FIWARE, Integrating the Healthcare Enterprise (IHE), Continua, and Open Integrated Clinical Environment (OpenICE). The top 5 reported technologies were cloud computing (19/37, 51%), REST APIs (17/37, 46%), Wi-Fi (17/37, 46%), gateways (15/37, 41%), and JSON (14/37, 38%).

Our recommendations (Textbox 3) are based on the findings presented in Table 3 together with discussions within our research group. These recommendations focus on technologies and standards that support cross-domain interoperability across all layers of the IoMT reference model, spanning from data collection by sensors to cloud-based data processing and secure cross-domain data exchange. In broad terms, an IoMT platform for HBC and prehospital care should include, in our opinion, the capabilities presented in Textbox 3.

This scoping review provides insights into the enabling technologies that can address interoperability issues in IoMT settings. Although the primary focus has been on solutions for prehospital care and HBC, the results are also applicable to other settings with similar characteristics and requirements. The results show that higher levels of interoperability in IoMT can be achieved by combining various technologies and standards from multiple interoperability levels.

On the basis of the studies (n=30) presented in Table 3, it seems that contemporary IoMT studies on prehospital care and HBC tend to focus on lower levels of interoperability. This conclusion is in line with previous research conducted by Rubí and Gondim [19]. One possible explanation for the predominant focus on device integration and lower levels of interoperability among the reviewed studies is our use of “IoMT” as a search term. Although IoMT is broad and includes integration with medical devices as well as medical applications, the term is somewhat ambiguous and might introduce some biases toward device-centric approaches. We believe that future research could benefit from using a more refined definition to emphasize a shifted focus toward higher levels of interoperability. Therefore, we propose the introduction of the term Internet of Medical Systems (IoMS) to address this gap. This new terminology could help distinguish studies focusing on device-to-device integration from those focusing on integration with EHRs and other health applications.

We acknowledge that some may argue that technologies and standards such as blockchain, fog computing, GDPR, and the like may primarily focus on aspects other than interoperability (eg, security, communication, networking, data transfer, or architectural design). However, proceeding from the 6-level interoperability model [11], our interpretation of interoperability extends beyond mere data exchange and interpretation. Even if two systems possess the capabilities to exchange and interpret information, we believe that it is not enough to achieve higher levels of interoperability within health care without considering privacy and security aspects. However, our findings indicate varying levels of focus on technologies such as blockchain and SMART on FHIR within the research community, with some of the studies (4/30, 13%) analyzing these technologies in depth, while others (26/30, 87%) tend to overlook them. As an example, of the 30 included studies, 8 (27%) reported the use of authentication and authorization mechanisms [79,80,90,91,95,96,101,105], but only 7% (2/30) reported the use of Open Authorization (OAuth) or SMART on FHIR as mechanisms for authentication and authorization [7,23].

On the basis of our findings, it seems that more comprehensive platforms tend to incorporate a wider range of technologies compared to simpler platforms (Table 3); for example, platforms that reported interoperability capabilities with external systems or the use of AI seem to more frequently use technologies and standards to semantically code the data, including ontologies, FHIR, LOINC, ICD, openEHR, RxNorm, and SNOMED CT. However, this relationship was not confirmed statistically, and we acknowledge that other factors may influence platform complexity, such as regulatory requirements, available technologies, and expertise among developers.

Table 3 reveals that 30% (9/30) of the IoMT platform development studies for prehospital care and HBC reported the use of semantic frameworks or ontologies. As the creation of ontologies is a necessary but time-consuming and labor-intensive task, research has examined the potential to achieve ontology alignment by using AI; for example, Dam et al [106] used augmented intelligence to map data across various domains to facilitate the integration of diverse data sources. According to Tangi et al [107], AI has the potential to not only establish a common language and foster a shared understanding of data but also clean and structure the data. While we acknowledge that using AI for interoperability is an emerging trend, such studies have not been incorporated in this review.

In the reviewed literature, semantic standards and ontologies were described as essential for establishing connections with external systems and enabling the use of AI. Interestingly, standards for image and video formats, such as Digital Imaging and Communications in Medicine (DICOM), have not been mentioned in the examined studies. Interoperability enhancing organizations and initiatives such as FIWARE, IHE, Continua, and OpenICE were reported to guide researchers and developers [88], but these organizations and initiatives were only mentioned in 3 (10%) of the 30 studies.

Fog and edge computing were reported to be used to overcome problems related to limited bandwidth, often associated with high data rate applications, such as fall detection systems or electrocardiography monitoring [97]. Our findings suggest that fog and edge computing are often combined with cloud computing, especially in AI-intensive applications that require more processing power. To improve the performance of AI algorithms and enable machines to correlate data, studies reported the use of ontologies, resource description framework (RDF), and OWL [24,99]. Furthermore, platforms requiring real-time functionalities (alerts or notifications) reported the use of MQTT (7/30, 23%) and WebSocket (7/30, 23%) [82,85].

While we have documented the use of commonly used technologies and standards in this study, it is important to acknowledge that the accuracy of these findings depends on the transparency and thoroughness of the explanations within the included studies; for instance, some of the studies (30/63, 48%) offered detailed insights into their IoMT platform development processes and the technologies and standards used, while others (33/63, 52%) focused on broad aspects of the topic, as in the studies by Rakhman et al [108] and Shim et al [109]. As a result, some of the studies (33/63, 52%) were excluded from Table 2. Furthermore, because technologies and standards can have versatile applications, specifying their exact application area can often be challenging. Therefore, the findings presented in Table 3 should be considered as a guidance rather than absolute truths.

In this study, the mapping was performed based on a 6-level interoperability model defined in the study protocol [11]. To facilitate the mapping process, any inconsistencies between our model and the levels described in the included studies were addressed using a flowchart model [11]. This was done to manage different naming conventions across different models; for instance, one study may label a specific interoperability level as “technical interoperability,” [105] while others may refer to it as “foundational interoperability” [110]. Similarly, the specific functions and components of each level may not be consistent across different models or frameworks, which can create ambiguity and uncertainty about how to define and interpret interoperability. Hence, proceeding from a different interoperability model [71] would most likely result in a somewhat different mapping outcome.

It is important to note that although we mapped the enabling technologies to specific interoperability levels, it is uncommon for a technology to address a single interoperability level; for example, the FHIR standard not only specifies data exchange formats and semantics but also supports REST APIs, enabling it to address syntactic, semantic, and cross-platform interoperability concurrently. Hence, in this study, we mapped FHIR to cross-platform interoperability, which represents the highest level of interoperability addressed by FHIR. However, saying that the FHIR standard solves cross-platform interoperability is not completely true because it needs to be combined with technologies on lower interoperability levels. We argue that the correct phrasing should be that FHIR can solve cross-platform interoperability issues if it is appropriately combined with other relevant technologies and standards.

Another aspect worth mentioning regarding the mapping is the importance of examining the primary focus of the technologies and standards; for example, the IEEE 11073 PHD standard involves data models that specify how a measurement and observations of vital signs should be represented, including descriptions of the different units, data types, and semantic meanings [38,40]. On the basis of this description, it is likely that the IEEE 11073 PHD standard should be categorized at the semantic interoperability level. However, due to its primary focus on device-to-device communication within LANs or PANs, the IEEE 11073 PHD standard does not fully meet the requirements for network interoperability, let alone semantic interoperability. Hence, the IEEE 11073 PHD standard was mapped to the device interoperability level in this study.

In our study, we found that technologies such as BLE and Zigbee, which primarily target device interoperability, are not sufficient for IoMT applications in prehospital or HBC settings due to their restricted physical communication range (ie, 20-100 meters) [41,111]. For physicians to remotely monitor patients and for devices to automatically send alarms during medical emergencies, the first step is to extend the communication range by enabling network interoperability [49]. Otherwise, the physician’s system would need to be in the same room as the patient, which obviates the concept of remote monitoring; for example, by combining BLE with long-range wide area network (LoRaWAN) or 4G or 5G—or 6LoWPAN with CoAP—long-range data transmission can be achieved [112]. This allows physicians to remotely monitor patients residing several kilometers away from the hospital [25]. However, choosing a particular combination of technologies from different interoperability levels does not ensure the achievement of the desired level of interoperability. This is because interoperability initiatives require collaboration among all involved stakeholders; for instance, a standard such as IEEE 11073 can be adopted in several ways, which means that simply specifying the use of IEEE 11073 may not be sufficient to achieve device interoperability. Instead, a standard should be tailored to meet the specific requirements of a particular use case or application. This can be achieved using profiles to ensure that the standard is implemented in a way that suits the needs.

Although close collaboration among stakeholders is essential to achieve higher levels of interoperability, studies have shown this endeavor to be challenging; for example, a study conducted in 2021 by Everson et al [113] revealed that more than half (55%) of health information exchange organizations in the United States reported varying degrees of intentional information blocking. The authors showed that the most prevalent form of information blocking involved the refusal to share information, which 14% of the health information exchange organizations routinely observed among EHR vendors. The reason for blocking information according to Everson et al [113] had to do with regional competition among vendors. By limiting the sharing of information with other actors, vendors can ensure that health care providers stick to their platform [113,114]. While this approach benefits vendors, it creates disadvantages for patients and the health care industry. The findings of Everson et al [113] suggest that while technologies and standards can streamline data exchange in health care, the reluctance of organizations to engage in such efforts impedes progress toward higher levels of interoperability.

The recommendations provided in this study draw on insights from previous IoMT platform development projects and our research group’s expertise. They focus on technologies and standards that together can support cross-platform or cross-domain interoperability. We acknowledge that multiple technologies and standards can be used to develop IoMT platforms. We are also aware that there is no “one right way” of adopting a specific technology or standard. Rather, we argue that interoperability initiatives and development projects should be facilitated by close collaboration among the stakeholders involved.

With this study and our recommendations, we hope to bring together the research, developer, and health care communities by highlighting relevant technologies and standards to foster collaboration and improve interoperability initiatives within the health care domain. While we recognize the importance of independence in driving innovation, we hope to see more IoMT platforms using common standards and best practices already available in the market. We further hope to see an increased engagement in interoperability initiatives such as FIWARE, IHE, and OpenICE.

We acknowledge the advantages of IoMT platforms capable of delivering personalized patient care and enhancing patient safety. However, we do not believe that technologies and standards are the only solutions. To achieve the highest level of interoperability, we argue that it is necessary to have a profound understanding of organizations’ working processes, routines, and policies. Hence, we look forward to delving into these challenges in close collaboration with relevant stakeholders and developing an IoMS platform that supports cross-domain interoperability.

This scoping review adheres to the published protocol [11] and focuses on technologies addressing interoperability issues within the context of IoMT platform development in HBC and prehospital care settings. Hence, studies focusing on hospital systems [115], platforms for intensive care units [81], integration platforms [93], or platforms for research or clinical trials [92,116] were omitted. Furthermore, most of the studies (33/63, 52%) displayed a deficiency in information or a lack of transparency regarding the development process and the technologies used; therefore, they were excluded from Table 1 and Table 3. As the focus of this study was to identify and summarize best practices and frequently used technologies, we did not assess the effectiveness of technology implementations in detail.

To manage the scope of this review process effectively, the reviewers made a deliberate decision to limit the included studies to those specifically addressing interoperability in an IoMT context. This was necessary to avoid overwhelming screening efforts with an unmanageable number of studies. We are aware that we might have missed relevant articles with this strategy. Furthermore, we are aware that technologies used in other sectors, for example, Industry 4.0, manufacturing, and transportation, could also be applicable to IoMT, especially considering that these domains share some common characteristics (eg, requirements for real-time, cross-domain information exchange). Another limitation of this study might be the use of the term “IoMT” in our literature search. Although IoMT is often defined as the interconnection of medical devices and applications, there seems to be a predominant focus on devices and lower levels of interoperability among these studies. This limitation could potentially be eliminated by the introduction and use of the new term IoMS.

Although recent advancements in AI have demonstrated its potential to enhance interoperability, studies reporting the use of AI for this purpose were excluded from this review. Another limitation of this study is that, to expedite the process, the coding scheme was tested by only 1 reviewer (MS).

In this study, we have demonstrated that the highest level of interoperability can be theoretically achieved through a strategic combination of various technologies and standards. Furthermore, we have provided a summary of the relevant technologies and standards that can be used in IoMT platforms to overcome interoperability issues in HBC and prehospital care settings.

Despite the availability of innovative and suitable technologies, the IoMT research community has reported limited interest in adopting technologies and standards such as Docker, blockchain, SMART on FHIR, HIPAA, and GDPR to achieve cross-domain interoperability. Most of the studies (17/30, 57%) have primarily focused on lower levels of interoperability (up to the semantic interoperability level). This observation highlights a significant research gap, particularly in achieving cross-domain interoperability within IoMT for HBC and prehospital care. To emphasize the need for higher levels of interoperability and to support future research, we advocate for the introduction of the term IoMS. In addition, the reluctance of organizations and vendors to share information and participate in interoperability initiatives highlights the importance of considering these aspects when addressing interoperability challenges, especially in procurement processes [117].