Background

Documentation in healthcare around the world is fragmented. The will and voice of patients are absent. They lack ownership and control of their health data. Previous work has shown the need for interoperability, but the focus had been on centralised architectures [1, 2]. The structure remains centralised and security breakages have caused great harm. Simultaneously, new technologies have matured, enabling more secure solutions for globally distributed health care platforms. Blockchain-based Personal Health Records (PHR, ISO/TR 14292:2012) have emerged as a predominant solution in the healthcare landscape, offering enhanced security and patient control [3].

Here, the idea of a Global Patient co-Owned Cloud (GPOC) encompasses a globally distributed and securely blockchain protected and patient co-owned platform of PHRs.

This is presented in the GPOC-series [4,5,6,7]. Its systematic review and meta-analysis expose the core facets of a GPOC [4]. The GPOC Survey shows a global consensus for its necessity [5]. A summit echoes this [6].An additional review and interview series explored the ethics and policies relevant to a GPOC [7]. However, the entire scientific focus of the series is to eventually deliver a functional online platform.

Here, we demonstrate the technical GPOC Sandbox. It is based on the series’ respective scientific methods and conclusions on all aspects of an ideal solution given the current technical possibilities. The gathered insights range from the optimal security, privacy, blockchain, platform architecture and encryption types to regulatory adaptations, e.g., GPDR-compliance, ethical considerations and feedback from key option leaders from all UN member states and 18 of the largest international health organisations. Our meta-analysis gave insights on cross-border data transfers and the GPOC ethics article on international data protection regulations.

The purpose is to allow all interested parties to explore and contribute to this project. This is because the concept requires a global effort. The Sandbox contains a platform of several protocols. The Sandbox’s technical design is based on insights from the GPOC Series. The structure is modular and explores several new technologies. It is consensus-based and patient-centric. Co-ownership is its core.

The Sandbox investigates biometrical authorisation and hashing protocols [8, 9]. It investigates patients’ management and movability of PHRs. Furthermore, it presents distributed ledger infrastructure. This permits global healthcare communication [10, 11]. The challenges for a GPOC include ensuring a unified underlying protocol is maintained for cross-border delivery of care based on data management regulations. This currently does not exist, and it is what the GPOC seeks to address. Open-source operating systems visualise the sandbox. It works with various systems without requiring any particular adaptations. Hence, it is an agnostic platform.

Technical sandboxes in domains outside healthcare, such as software development and finance, provide isolated environments for testing new code and innovative products, respectively. These sandboxes facilitate innovation while ensuring compliance and security before broader implementation. Similarly, the GPOC Sandbox aims to test and validate the integration, security, and functionality of a co-owned global cloud for personal health records. After launch the sandbox will provide performance metrics and testing benchmarks to validate security, efficiency, and usability.

The GPOC Sandbox explores several concepts, including the integration of Systematised Nomenclature of Medicine (SNOMED) and the International Classification of Diseases (ICD-11). This is done to ease communications and medical research [12]. These standardised terminologies ensure consistent, precise, and interoperable data exchange across different healthcare systems, facilitating accurate diagnosis, treatment, and global health analytics [12, 13].

Here, we will present the GPOC Sandbox’s technical components and implications. The results elaborate on the sandbox's twelve modules, its blockchain technologies, and its accessibility. The discussion explores the role of blockchain in the GPOC framework, alongside decentralisation, security considerations, and future developments. The methods outline the construction process of the sandbox, highlighting the integration of blockchain protocols and data availability. Finally, we conclude the possible global impact and that the ongoing and future development of the GPOC concept will require international partnerships.

Methods

We used open-source tools to create the GPOC Sandbox [14]. The goal was to incorporate the conclusions from the GPOC Series. We use an open-source cloud service to demonstrate the platform. Data preprocessing was dependent on the data-type to be processed, which included text data, numerical, tabular data, imaging data etc. For all forms of data standard pre-processing libraries for natural language processing and machine learning were utilised, but we also utilise in-house and various globally accessible libraries to pre-process the data. The challenge for some countries is firewall restricting access to some of these libraries and a good reason why a sandbox is necessary. It is also desirable because it allows a standardised and agreed protocol and libraries to be assembled and assessable to all interested developers.

Mina protocol

The Mina protocol was used as the underlying smart contract blockchain protocol. Thus, it implements zero knowledge proof through succinct non interactive arguments or knowledge (ZK-Snarks). The aim is to prove information without additional information leakage [15, 16].

Mina boasts a 22 KB blockchain size compared to more than 250 GB size for other blockchains [16]. These protocols may be optimal for a GPOC. Zero knowledge proof implementation enables security and sustainability [17]. It has a lightweight carbon footprint. The described technology stack may have two sections. One frontend working on-chain and one backend off-chain allowing verified data management on an additional private blockchain network.

Figure 1 illustrates one approach. Ganache, a local blockchain development tool to tests smart contracts. The implementation of the smart contract Mina implementation is achieved with TypeScript, in contrast to Solidity for Ethereum. Currently, it is in development and available in a public blockchain format. However, it has the potential to be developed into a permissioned use case for GPOC [16].

Fig. 1
figure 1

Example technology stack for gpoc

Full size image

Illustration of an example use of Mina for on-chain data processing, employing zero-knowledge proofs. The figure shows step 1 (left) with the frontend zero-knowledge application and step 2 (right) with the offchain blockchain application. A verification proof is stored locally on the private blockchain for network participants, including clinicians, patients, and their families. The queried data undergo conversion into a homomorphic encrypted form, are processed through a prover function, and are verified using ZK-Snarks (zero-knowledge-succinct non-interactive argument of knowledge). When queried by a public network participant, such as a company or researcher, and with owner permission, the data query's proof is verified by a verifier function with a cryptographic key stored on-chain. The state of the blockchain during interaction can be stored off-chain to expedite subsequent queries. The off-chain stack can also be accessed offline [16]. The Mina protocol was selected and is sufficient as it offers an environmentally friendly alternative to traditional blockchain architectures falling in line with the UN sustainable development goals and carries a smaller data storage footprint. The image is free to use and was created using Keynote 11 and Adobe Photoshop 2021 for the study by the last author.

Ethereum and hyperledger protocol

We employed the Ethereum smart contract protocol, implemented using Solidity within an individual permissioned use-case named Hyperledger Besu [18]. This configuration facilitates enterprise-grade platforms specifically tailored for sandbox development. Notably, Hyperledger Besu provides flexibility in supporting various networking protocols, liberating sandbox users from infrastructure limitations. This ensures a familiar working environment, enabling users to create their relevant GPOC.

Data availability

All the data and code generated in this study are provided in the Supplementary Information. Source data are provided with this paper. There are two repositories associated with this study:

  1. 1)

    The generated code and source data are available in the GPOC Sandbox, DOI: 10.5281/zenodo.10547507

  2. 2)

    Supplemental materials and UX/UI wireframes are available in the article repository on Figshare, DOI: 10.6084/m9.figshare.c.7067762

Results

The GPOC Sandbox comprises twelve modules. Its backend design emphasises portability and module scalability, leveraging blockchain technology. Users have the flexibility to choose and research the type of GPOC they wish to create. The GPOC Sandbox is openly available on a repository on Zenodo, https://doi.org/10.5281/zenodo.10547507.

The roadmap for constructing the sandbox, outlining the flow of steps, research gaps, limitations, and assumptions, is presented in the GPOC Series, summarised in Fig. 2. It also illustrates the sandbox architecture. The GPOC Sandbox will be officially launched upon the publication of this research article. Hence, future research will be able to report how it has been utilized.

Fig. 2
figure 2

Overview of the GPOC series and roadmap to the GPOC sandbox and its architecture

Full size image

The technical requirements of decentralised blockchains, clouds, adaptable UX/UI and homomorphic encryption have been used [8].

Blockchain technologies for the GPOC sandbox

Blockchain solutions, particularly those emphasising zero-knowledge proof and decentralisation, have been strategically chosen for the GPOC Sandbox. With its emphasis on patient co-ownership and secure global healthcare communication, the GPOC concept demands robust and trust-less transactions facilitated by blockchain technology. The unique requirements of GPOC, such as patient co-ownership and participation in global medical research, have directly influenced the technical design of the sandbox, aligning the chosen blockchain technologies with the GPOC vision of democratising healthcare.

Scalability and robustness

The GPOC Sandbox incorporates Blockchain-based Zero-Knowledge Proof (BZKP), aimed at achieving scalability and robustness in protecting sensitive PHR data. The implementation strives to align with the patient-centric goals of GPOC, facilitating secure and efficient healthcare data management.

Data storage reduction

Through the use of BZKP, the GPOC framework anticipates a reduction in data storage requirements, addressing challenges commonly associated with blockchain technology in healthcare applications. This approach supports the sandbox's goal of enhancing data security and accessibility.

Table 1 shows several blockchains relevant to GPOC. Moreover, the Internet of Things (IoT) increases the number of PHR sources. These are often owned by patients. Thus, patients become co-contributors to their own PHRs.

Table 1 Blockchain technologies relevant to Global Patient Co-Owned Cloud (GPOC) in healthcare: overview and potential applications
Full size table

Blockchains that may be relevant to GPOC. Such healthcare networks can share and procure sensitive patient data. It can be exchanged between laboratories, clinics, hospitals, and caregivers. Applications of these decentralised blockchains can be used to accurately identify mistakes. Hence, we provide an overview of common blockchains relevant to healthcare and their potential use for GPOC. Blockchains may be the cusp of a new healthcare era [19,20,21,22,23,24,25,26,27,28,29,30].

The GPOC Sandbox is downloadable with minimal installation requirements. Illustrative examples are included. However, users have the freedom to adapt and research their GPOC version and user interface (UI/UX). Moreover, a collection of ergonomic and minimalist UX/UI wireframes for GPOC is available on the article repository on Figshare, https://doi.org/10.6084/m9.figshare.c.7067762. A forthcoming article will deliver user feedback on the overall usability.

In 2020 the structure of the GPOC Series’ five parts was planned. The initial GPOC systematic review and meta-analysis was recently published [1]. The following GPOC Survey received answers from 100% of the 193 UN member states and three other states and 18 organisations participating [2]. Hereafter, the review and inquiries on ethics, policies and regulations were assembled [3]. Finally, we organised a Delphi-style GPOC Summit [6]. All these four publications in the GPOC Series emanated in the eventual GPOC Sandbox construction. Thus, it is based on all the insights from the entire GPOC Series. This roadmap is illustrated in the upper left quadrant of the figure. At the centre the start of the programming of the sandbox is visible. It results in a tailored structure of 12 modules and quick installation module, that is adapted for ‘emergency’ study situations in the field. The modules include security features such as firewall, blockchains and homomorphic encryption. It also provides economic and predictive models for public health usage. Finally, it is equipped with modules for standardised evidence based healthcare, open PHR and a Quantum Toolbox for research, international development and dissemination of AI in medicine. In the event of the creation of GPOC Foundation for global collaboration a dedicated module for this purpose has been added. The image is free to use and was created using Keynote 11 and Adobe Photoshop 2021 for the study by the last author. For details and Source Data, see the respective publications.

The study resulted in displaying a range of applications of blockchains relevant for the GPOC, which is illustrated in Fig. 3.

Fig. 3
figure 3

Applications of blockchains for GPOC

Full size image

Illustration of some applications of blockchains relevant for the GPOC technical solution. Note that tokens have both virtual and real-world values, i.e., there are also disadvantages with blockchains, which are elaborated below. The image is free to use and was created using Keynote 11 and Adobe Photoshop 2021 for the study.

Homomorphic encryption

Fully Homomorphic Encryption (FHE) is currently the most relevant method to GPOC [40, 52,53,54]. It supports the use of analytics on encrypted data [8]. This technology ensures that sensitive patient information remains secure while allowing for meaningful analysis and insights to be derived, thus supporting the sandbox's objectives of secure data management and patient co-ownership.

The study resulted in relevant GPOC features that are displayed in Fig. 4.

Fig. 4
figure 4

Optimal UX/UI for GPOC

Full size image

Illustration of the science of optimal UX/UI, relevant for a global platform such as GPOC. The central mission is to make it as accessible as possible and prevent discrimination against those with a disability, etc. It should present a solution that is simple, inclusive, adaptive, efficient, and truly global. The image is free to use and was created using Keynote 11 and Adobe Photoshop 2021 for the study by the last author. A suggested collection of ergonomic and minimalist UX/UI wireframes for GPOC, also free to use and created for the study by the last author, is available on the article repository on Figshare, https://doi.org/10.6084/m9.figshare.c.7067762.

Discussion

Introduction

A global world with frequent travel requires a patient-centric and movable PHR. The GPOC concept suggested here can be further investigated in the Sandbox. The chosen solutions for the GPOC Sandbox are discussed below.

Blockchains in general

Blockchain technology serves as a decentralised and transparent ledger system that records transactions across a network of computers. Its widespread adoption extends beyond healthcare, finding utility in domains like finance, supply chain management, and voting systems. Blockchain's immutable and tamper-resistant nature ensures data integrity and trust in various applications, revolutionising processes through disintermediation and enhanced security measures.

Role of blockchain in GPOC framework

Blockchains play a crucial role in the GPOC framework by allowing the permanent recording of encrypted data, rendering access nearly impossible without the requisite encryption codes. Within a peer-to-peer network-driven system, users collaboratively solve complex cryptographic nonce-based hashes, creating fingerprints that serve to prove the authenticity of transactions. The trust-less nature of this interaction is key, as it certifies the origin of transactions without the need for a central party. This security is further reinforced by consensus algorithms operating on game theory, ensuring that the addition of blocks is a rigorous and secure process [31].

Decentralisation and security

A blockchain is a linear transaction ledger that is duplicated and distributed across an entire network of peer-to-peer (P2P) computers. Each user stores one ledger copy, and all user computers are nodes. Validation of the encrypted data creates durability and transparency, resulting in traceability from the genesis block.

Regulations may require keeping information no longer than necessary. Blockchain solutions for healthcare try to address this issue with off-chain interaction processing [32].

For healthcare, the decentralised and transparent blockchain technology is strategic for solving issues and providing complications. PHRs require both privacy protection and accessibility in the event of appropriate healthcare actions. This is accentuated in a GPOC.

Zero-knowledge proof and IoT model

Blockchain-based Zero-Knowledge Proof (BZKP) is an Internet-of-Things (IoT) model. It is patient-centric and aims to protect sensitive PHR data [15]. Its scalability, robustness and immutability are suitable to GPOC [15]. Blockchains accumulate large amounts of data and BZKP reduces storage.

PHR interoperability and blockchain-based solutions

As discussed earlier, the prominence of blockchain-based PHRs in healthcare reflects their widespread adoption. Their popularity is attributed to the heightened security and patient empowerment they afford, aligning seamlessly with the goals of GPOC. For instance, MyHealthData permits downloads from multiple institutions via mobile devices and a blockchain relay server. It is designed for PHR interoperability [33]. The recently published Blockchain-Based Deep Learning as-a-Service (BinDaaS) is a combination of blockchain and a deep-learning platform with inbuilt clinical predictions. This provides superior performance, accuracy, end-to-end latency and mining time compared to other models [34].

Outsourced PHR clouds and security features

For the usage of outsourced PHR clouds, key features of a secure health cloud have been presented in a case study of blockchain-assisted PHRs [35]. A hybrid-blockchain solution addresses some security issues with sharing. Analysis with the blockchain benchmark tool Hyperledger Calliper revealed high performance [36]. For GPOC Hyperledger Besu was used [18].

Most blockchain-based PHR solutions have focused on single chains. The latest leakage mitigations require multi-chains. Hence, the Relay-Chain as a Service (RaaS) and a cross-blockchain PHR solution may be suitable for patients visiting many hospitals [35]. This was deemed relevant to GPOC and can be further explored in the sandbox.

Moreover, the unique requirements of GPOC, such as patient participation in global medical research, have been considered in the technical design of the sandbox. The chosen blockchain technologies align with the GPOC vision of democratising healthcare and contributing to the dissemination of artificial intelligence within the medical domain.

Centralised vs decentralised clouds

In the GPOC framework, understanding the nuances of cloud infrastructure becomes pivotal. Clouds, whether decentralised with globally distributed storage or centralised under singular control, directly impact the co-ownership and security aspects of GPOC. As we navigate through the intricacies of PHR data encryption, a crucial facet in GPOC's commitment to secure health data management, we encounter challenges such as time consumption and escalating costs, particularly with an increasing number of access policy attributes. Recognising the need for enhanced performance, GPOC introduces Fine-Grained Access Control with User Revocation (FGUR). This not only addresses performance concerns but also aligns with GPOC's overarching goal of empowering patients in managing their health data. A strategic combination of Broadcast Ciphertext-Policy Attribute-Based Encryption (BCP-ABE) and attribute hierarchies of Comparison-Based Encryption (CBE) further reinforces the GPOC commitment to robust security measures [37].

Challenges with centralised clouds

Centralised clouds mean storage and transfer by trusted third parties (such as Amazon, Google, and Microsoft). Here there are weaknesses that can harm the data. Currently, most PHR solutions are centralised. However, the Diagonal Digital Signature Algorithm (DDSA) using Merkle Patricia Hash Trie (MPHT) algorithm is a PHR sharing solution with blockchain [38].

Decentralised blockchain solutions

In the context of centralised clouds, considerations align closely with GPOC. The challenges associated with centralised clouds directly impact GPOC's mission of co-ownership and secure health data management. The GPOC Sandbox addresses these challenges by adopting a decentralized approach, ensuring trust-less transactions and empowering users in co-managing their health data securely.

A main issue with centralised clouds is the loss of privacy and security of sensitive PHRs [39]. Therefore, we argue that outsourcing solutions for PHRs have critical such issues [40].

Ethereum and smart contracts

To solve this issue, decentralised blockchains ensure trust-less transactions. Each network member possesses an identical copy of data in a distributed ledger; any alteration is rejected by the other users. For instance, Ethereum, a decentralised and open-source blockchain, incorporates smart contract functionality. Serving as the native cryptocurrency of the platform, Ethereum empowers the development of applications on its blockchain [31, 41, 42]. Hence, this is the chosen solution for the GPOC Sandbox.

Hyperledger for secure health data

Hyperledger, a platform for collaborative, permissioned private blockchains, aligns with GPOC's focus on secure and co-owned health data. Its support for emerging architecture design, including hybrid infrastructures that unify permissioned and public networks, underscores its suitability for GPOC [43].

Diverse blockchain ecosystems

Diverse ecosystems, like Directed Acyclic Graph (DAG)-based (e.g., Hedera Hashgraph, Holochain) and blockchain-like systems (e.g., Nano, IOTA, Obyte), demonstrate unique designs for efficiency and privacy [44,45,46,47].

Scalability and layer 2 protocols

Layer 2 protocols (e.g., Cellar, Loom, Ark, Cosmos, and Tesseract) facilitate scalability and privacy through state transfer channels [46]. In healthcare, GPOC should support state change propagation and reversibility [48, 49]. Proposals for scalability, sich as sharding and block-size modifications, contrast with the limitations of slow and expensive layer 1 networks [30]. Emerging healthcare chains, such as HealthChains, are also under consideration [48].

Future developments in blockchain for healthcare

Even though, blockchain implementation may be expensive, user costs may be lower and energy consumption may be higher. Moreover, lost key generation may be impossible, and storages may exceed hard disc capacities. The security issue with social engineering remains. However, there is software capable of relying on decentralised or token-based distributed ledgers with effective cryptographics. Recent solutions have been suggested, such as Healthfetch, an influence-based context-aware prefetch scheme in citizen-centred health storage clouds and a cloud-dew architecture based PHR framework [50, 51]. These were considered, but they were not relevant for GPOC’s decentralised protocol.

Integration with global medical research

A GPOC should support global medical research on its valuable content. However, the co-owning patients should be able to opt-in for participation. Hence, a microflow of payments to patients needed to be modelled in the sandbox. Moreover, the possibility of contributing to global research and the dissemination of AI needs to be considered. Additionally, bias mitigation and the promotion of equal healthcare access are important. The development of AI for GPOC may lead to a global increase in Evidence-Based Medicine (EBM).

UX/UI design for patient-centric care

The most effective and ergonomic UX/UI is a science in itself [9]. Its adaptability to local or personal preferences is relevant in patient-centric care. Large swathes of the world may access PHRs via smartphones. It is pivotal to adapt the UX/UI for elderly or impaired individuals [10,11,12]. The UX/UI of GPOC should lead to an efficient workflow. In contrast, social media designers often wish to prolong logged in sessions and increase the advertising value. The PHR content is already valuable per se. Hence, there is less value in digital addiction.

Future developments may include large natural language processing, multi-chains, quantum AI and security for GPOC [55, 56].

In summary, every technical decision made in the development of the GPOC Sandbox has been intentionally aligned with the core principles of the GPOC concept. Thus, reinforcing its potential impact on global health and medical research.

Future work and partnerships

Partnerships with organisations such as the United Nations, International Committee for Red Cross, World Health Organisation and the World Economic Forum would allow us to build a more robust GPOC architecture for alpha and beta testing. Specifically, for PHR data management this will be in a form of an application incorporating the ratified protocol from the underlying unified consensus in current and future GPOC Summits. Ongoing multi-institutional research organisational partnerships will further enable us to develop the sandbox to cater for individual global patient data management. Additionally, we intend to build on top of the Mina protocol because of its environmentally friendly footprint.

Global collaborations include several universities, institutes and advanced technological frameworks. These include Karolinska Insitutet, MIT, Harvard, Oxford University and Imperial College London with its Global Health Innovation Network (GHIN). With these partners a detailed roadmap for implementing the GPOC sandbox, including timelines, milestones, and resource requirements will be included in the forthcoming article.

Limitations of sandbox implementation

The sandbox implementation is limited by the current lack of international collaboration on PHR design and data protection regulation. This limitation is further affected by various design preferences and constraints based on specific local regulations that may affect the use of a sandbox.

Conclusions

In conclusion, a global multi-institutional collaboration for a sandbox allows standardised PHR data access, streamlining the delivery of care for various patients across various locations. It is freely available online for all interested parties to research and explore. Here, we incorporate the GPOC concept. It encompasses a PHR co-ownership, trisected between the patient, clinicians and clinic. It is a distributed platform based on blockchains. We aimed to include the insights from the articles in the GPOC-series. Thus, the presented cloud-based ledger-like sandbox is the result. Its modules lie open for global research and adaptation. Hence, it contributes to the democratisation of healthcare. It facilitates the research and spread of AI within medicine. The GPOC Sandbox may have an impact on global health.