A search protocol was designed to search articles on topics from popular databases. A search string was proposed by using the appropriate search keywords and terms. A combination of search keywords and terms was used in the ACM digital library, IEEE Xplore, SCOPUS, Springer, PubMed, and Google Scholar databases. The search for literature on the topic was performed between January 2015 and March 2023. The year 2015 was chosen as the HF framework was first introduced as a permissioned blockchain in the year 2015 by the Linux Foundation (Inc., 2023). To promote the research results, search keywords were used that would help in answering the research questions. We used Boolean operators (AND and OR) for search strings as follows:

The next step of the process was to perform the screening of the relevant studies. The screening process began with the examination of all studies collected from databases in the previous step. We developed a reference list with the help of Endnote X8.0 that was employed to eliminate duplicate studies. The rest of the duplicated documents were removed manually. Titles and abstracts of documents were assessed to determine their relevancy for the current systematic review. Two authors (M.H and I.G) performed the screening of studies using the inclusion and exclusion criteria (Figure 1). Thus, studies that were not clearly focused on the HF framework and its applications in healthcare regarding EHRs were excluded.

This systematic review followed the studies’ inclusion and exclusion criteria as follows:

The quality assessment of the studies was performed according to the guidelines from Kitchenham and Charters (20). This allowed us to assess the relevance of studies to the research topics in this systematic review. Six randomly selected studies were subjected to the process of quality assessment and checked for effectiveness.

Table 1 shows the checklist of quality assessment criteria (QAC) used to qualitatively analyze the collected studies. This checklist was applied to all the studies identified in this systematic review. There were some studies that did not meet one or more of the above-given checklist items, and hence those studies were removed.

All studies that had passed the QAC had their data extracted to examine the completeness and accuracy of the information contained in each study. Data from studies were extracted, categorized, and stored in the Excel sheets. The categories of extracted data were as follows:

Information about the purpose of study, problem addressed, and proposed solution is vital for a comprehensive understanding of the research

Pros and cons: The information about the advantages and disadvantages of proposed solutions.

A total of 1,048 records were identified by using the initial keyword searches on the databases. Of them, 23 duplicate records were removed because they were either conference papers that did not comprehensively explain the research themes, results, and interpretations or short papers that had insufficient information about the topic. Next, 1,021 records were screened based on the title/abstract, following the studies’ inclusion and exclusion criteria. After assessing the title/abstract of the remaining documents, they were reduced to 110 papers that were found eligible for the full-length text analysis. After reading the full-length text documents, 57 papers remained, providing the final set of articles included in this SLR.

Figure 2 presents the year-wise distribution of publications for the chosen period. It can be noticed that, as a result of the search process, including the paper’s inclusion and exclusion criteria, no publications were found in 2015. From 2016 to 2020, this area of research did not receive much attention from scholars. However, Figure 2 shows an interesting point about the rapid increase in publications from January 2021 to February 2023. The increasing trend in published research indicates a significant surge during these years, suggesting that this area of research may become highly sought-after due to its widespread applications across various domains.

The privacy preservation mechanism is based on four aspects. The first aspect is the inclusion of symmetric cryptography and the separation of transaction information from on-chain records. The second aspect involved the digital certificate that ensures the legitimization of organizations on the blockchain. The third aspect is the design of separate multichannels for information distribution. The final aspect isolates the information privacy between the various organizations on the same channel. The HF framework performs two primary functions for data processing. Information processing is based on two criteria: the first criteria involve keeping confidential information and ledgers on a channel for people outside the channel; the second criteria set a scenario where information and ledgers are shared among organizations. Some of them will be able to see the transaction, while others can know about the occurrence of transactions and verify the authentication of transactions (28). However, the information security mechanism becomes weak when several medical institutes lack coordination in healthcare and consensus to determine how data should be utilized or shared when needed (29, 30). Hence, fragmentation problems exist that can be tackled by proposing and implementing a consensus mechanism. To some extent, the implementation of the HF framework has addressed the issue of low transaction efficiency when compared to conventional blockchain architectures such as Bitcoin and Ethereum, which rely on high-power or very complex algorithms to reach consensus (31). Furthermore, the sharing of prevalent health data becomes more challenging when it is shared among stakeholders in its various formats and standards (32). Thus, it becomes hard to examine, aggregate, and share health records in emergency situations.

IoMT is the core application of IoT, which includes remote monitoring and diagnosing health records. Blockchain-enabled IoMT has been deployed to store data at hospital buildings for the provision of real-time monitoring of temperature, air quality, and environmental hygiene (33). Both passive (e.g., tags sensors) and active devices (robots) that can clean hospitals and disinfect can be used. Data collected from these devices can also be used for prescriptive and predictive analysis.

In the context of healthcare, several sensors, such as wearable and off-body sensors, perform sensing of the patient’s body and periodically send health records to a personal digital assistant (PDA). This data passes through the cloud and medical servers, where it is assessed and a prescription is suggested using the patients’ health records (21). Below, Table 2 shows the features of the studied literature on IoMT in healthcare.

We identified a few benchmark studies that employed IoMT with blockchain frameworks to make health records more secure and confidential. Innovative research on the integration of IoMT with the BCT is in its infancy stage. As listed in Table 2, studies contributed toward the proposal of approaches that employed only HF (21, 24, 25, 27), or in a combination with other frameworks (22–26). Most recent literature shows that the HF framework is mostly used in this emerging area of research. To assess the quality features of medical records, this SLR identified several quality features and relevant issues studied in the literature.

Figure 3 shows the results on the quality features of medical records. This SLR identified 11 quality features and relevant concerns about them. Security as a quality feature has been widely studied in the literature, followed by availability, integrity, privacy, interoperability, and scalability. The rest of the quality features and their issues have been poorly addressed in the literature. Trend analysis, as shown in Figure 3, indicates that security remained among the priority concerns of scholars. Although several issues regarding quality features were mainly focused on the literature, but proposed approaches suffer from certain challenges in their implementation in real-word scenarios. Implementation cost (21), effectiveness (23), performance evaluation (24, 26), and storage cost (25) are open issues in the emerging area of IoMT integrated with the blockchain frameworks.

EHRs contain patients’ sensitive information, such as name, address, medical history, social security, and insurance numbers. Patients’ data have worth for stakeholders, and data exposure to the public has adverse consequences. To reduce the impact of patients’ data exposure, the proposed approach used the HF framework to create and test various scenarios for data security (34). The HF framework complies with the General Data Protection Regulation (GDPR) and covers several areas. It has an edge over the other frameworks in preventing cyber-attacks and is best suited for healthcare applications compared to other frameworks. In the following Table 3, we present a summary of the proposed approaches that utilized the HF framework in the healthcare sector.

Most recent approaches using the HF framework have been proposed to secure the exchange of health records and ensure the confidentiality and integrity of records (40–42). The HF framework is focused on monitoring and tracking EHRs on a cloud server, as revealed in a study (43). Health records stored in HF are acquired by the federated server, where they are analyzed and pre-processed. The analyzed and pre-processed data are fed into a module for tailored recommendations, aiding physicians, nurses, and patients in countering health issues. Thus, the survival rate of patients can be increased using the proposed approach. Since the proposed approach integrates blockchain and federated learning, data from emerging COVID-19-like infections require an update to the proposed framework. Before this study, a tamper-resistant mobile health system was developed and evaluated using the HF framework that showed auditability and trusted computing (44). A mobile application using insomnia therapy data was referred to by HF as a blockchain network. This novel implementation led to improved accessibility and data transparency.

This section presents an overview of the existing literature on the BCT-based HF framework and applications to resolve data privacy and security problems. Audio data stored in the cloud can be exposed if a decryption method is not implemented. When audio conversations between physicians and patients are stored on clouds, security concerns remain and data leakage can occur. To overcome this issue, a scheme using the homomorphic encryption library (HElib) was proposed and implemented on the “Contabo” cloud platform. The proposed scheme has the potential to increase the computation speed and performance if other open-source libraries such as PALASIDE, SEAL, and HEAAN are used in future studies (45). When the trust level of a cloud storing sensitive data decreases due to a single node failure, the cloud data provider becomes vulnerable to attacks and data theft (46). Table 4 below shows the summary of the studies relevant to the research topic.

We performed a comprehensive analysis of the pros and cons of the studied literature on the topic. The problems outlined in Table 4 are based on the choices of authors to use blockchain frameworks. Notably, the HF framework was frequently used in research. Out of 14 studies examined, 10 (or 72%) articles employed the HF framework, while 2 (or 14%) articles used the Ethereum framework, and the remaining 2 (or 14%) articles used a combination of two frameworks. Several advantages are brought by the HF framework compared to the Ethereum framework.

The majority of studies reported the performance aspects of their proposed frameworks. Metrics such as latency, throughput, and execution times were commonly used to highlight the better performance of the HF framework. As listed in the table above, there are core issues related to data privacy and security of health records, effectively prevent potential attacks. Critical systems in healthcare can be protected from external as well as internal attacks by using the applications of the HF framework in healthcare. Figure 4 below gives compelling evidence of the increasing role of the HF framework in tackling the privacy and security challenges of EMRs in the healthcare sector. Comparatively, the Ethereum framework indicates that only a few studies involved Ethereum either alone or in combination with the HF framework to address data security and privacy issues.

Handling EMR operations at hospitals using a decentralized platform is one of the primary requirements. This ensures that patients can receive safe and meaningful assistance from healthcare providers and services. The HF framework ensures low resource utilization and high transaction throughput (59). However, computer specifications and blockchain network size are main concerns that may be addressed by expanding the size of networks and deploying the proposed solutions on cloud architecture. However, migrating patients’ data from hospitals to the cloud have its own threat vectors. Sometimes, devices securing medical records become targets of distributed denial-of-service (DDoS) and ransomware attacks (60). To overcome this issue, an interplanetary file system (IPFS) and blockchain can secure and improve data storage (61). Before the emergence of such modern technologies, medical records were either paper-based or stored in conventional databases, which faced issues of security and data duplication. Duplicity of medical records can be prevented by deploying IPFS on networks (Figure 5).

The key aspect of the blockchain framework is the performance of the proposed solution based on the BCT. Various scholars employed the Hyperledger Caliper as a benchmark method while working on the implementation of BCT. Hyperledger Caliper can be used to generate reports on different performance metrics, including latency, transaction per second (TPS), and execution time (62). The aim of using these performance metrics was to highlight the importance of the HF framework for health records.

Table 5 demonstrates the key metrics and their values to analyze the proposed approaches. It has been pointed out in several studies that the latency of the proposed systems using the HF framework is always lower than the Proof of Work (PoW) protocol, while they exhibit higher throughput values (54). The HF framework-based privacy and security-preserving techniques show a higher level of confidentiality, traceability, and anonymity as compared to the existing approaches. Even these techniques (22) can handle thousands of concurrent transactions in emergency scenarios during the COVID-19 pandemic.