FAIR Equivalency with Regulatory Framework for Digital Health in Uganda

This study explores the possibility of opening a policy window for the adoption of the FAIR Guidelines— that data be Findable, Accessible, Interoperable, and Reusable (FAIR)—in Uganda’s eHealth sector. Although the FAIR Guidelines were not mentioned in any of the policy documents relevant to Uganda’s eHealth sector, the study found that 83% of the documents mentioned FAIR Equivalent efforts, such as the adoption of the National Identification Number (NIN) as a unique identifier in Uganda’s national Electronic Health Management Information System (eHMIS) (findability), the planned/ongoing integration of various information systems (interoperability), and the alignment of various projects with international best practices/standards (reusability). A FAIR Equivalency Score (FE-Score), devised in this study as an aggregate score of the mention of the equivalent of FAIR facets in the policy documents, showed that the documents at the core of Uganda’s digital health/eHealth policy have the highest score of all the documents analysed, indicating that there is a degree of alignment between Uganda’s National eHealth Vision and the FAIR Guidelines. Therefore, it can be concluded that favourable conditions exist for the adoption and


Background
Since the dawn of the Internet, rapid technological advances have made it possible for anyone to generate data. This 'Big Data' deluge has resulted in a multitude of separate data sets that exist as siloes. This data, which is stored in diverse data repositories, is, to a large extent, neither discoverable nor accessible by humans or automated means, greatly hindering knowledge, discovery and innovation. Hence, in its current state, the contemporary digital ecosystem hampers data discovery and reuse, preventing the maximum benefit from being derived from data resources [1]. In order to address this problem, science funders, publishers and government agencies are beginning to require data management and stewardship plans for data generated by publicly-funded projects, including in relation to data collection, annotation, archiving and long-term care [1].
In 2014, a group of academics, industry leaders, funding agencies, and scholarly publishers came together and championed a set of guidelines called the FAIR Guidelines, which were designed to support good data management by making data and metadata easily discoverable by both humans and machines [1]. The four components of these guidelines are that data be 'Findable', 'Accessible', 'Interoperable', and 'Reusable' (FAIR) [2]. When data is not FAIR, it becomes difficult to carry out the deep and broadly integrative analysis that is usually required to answer a research question. It can often take weeks (or months) to gather the necessary data. Considering the rapidly growing and evolving data environment, with new technologies, and new and more complex data types and formats constantly being developed, it is more important than ever that data is machine-actionable/readable. The harmonisation and integration of the data deposited in general-purpose and special-purpose repositories (making such data FAIR) will go a long way towards extracting maximum benefit from data, saving time and significantly reducing the cost of doing research [1].

2 Main Study Objective
This study was conducted under the GO FAIR Implementation Network Africa [7] and the Virus Outbreak Data Network (VODAN) [8], as part of a project that aims to develop a FAIR Data Point for Uganda [9]. It looked at the extent to which conditions (health data policies and documents) in the digital health/eHealth ecosystem in Uganda allow for the opening of a policy window for the adoption and implementation of the FAIR Guidelines and how these guidelines solve the challenges faced by the eHealth sector in Uganda.

3 Specific Study Research Question and Objectives
In order to assess the governance framework for digital health/eHealth policy in Uganda and explore the possibility of a policy window opening for the FAIR Guidelines to be adopted in Uganda's eHealth sector, the main research question was: To what extent and in what way is the lack of policy on data management, data sharing and data analytics contributing to the lack of sustainable digital health solutions in Uganda?

FAIR Equivalency with Regulatory Framework for Digital Health in Uganda
The objectives of this study were to: • Understand the governance framework for digital health/eHealth in Uganda as well as standards for digital health data • Investigate the conditions in the digital health/eHealth ecosystem in Uganda and evaluate the possibility of a policy window opening for the adoption and implementation of the FAIR Guidelines This first objective was achieved by studying the governance framework for digital health/eHealth policy in Uganda from both a health policy perspective and an information and communication technology (ICT) data management perspective, with particular emphasis on digital health data. To achieve the second objective, emphasis was placed on reviewing key policy documents regarding digital health/eHealth governance in Uganda and assessing the context within which FAIR would have to be introduced to maximise the chances of its adoption.

Theoretical Considerations
King don identified that new ideas can enter the policy agenda when the problem, potential policy solutions and political relevance pertaining to the problem are aligned. At such a point of alignment a policy window opens, which is a specific moment in time, when a new idea can enter the policy agenda [10]. The dynamic process that Kingdon describes, allows for the investigation of how policy changes and the role of policymakers and other policy entrepreneurs, who are 'members' of policy networks in a certain domain [10,11]. Van Reisen and others explain that the discrepancy between the FAIR Guidelines and the opinions and beliefs of researchers and practitioners who are expected to use these Guidelines can be considered cultural entropy [4]. The authors go on to argue that if cultural entropy is large, it will hinder the acceptance of the FAIR Guidelines, as it is likely to result in a discrepancy in terms of what is perceived as the problem and what the solution could be [4].

Study Design
This study was a qualitative cross-sectional study that involved analysing both ICT-related and healthrelated policy documents published by the Government of Uganda between 2009 and 2019.

Evolution of ICT and Health-Related Policies in Uganda-Historical Perspective
In order to identify which policy documents and legislation are the basis for digital health/eHealth policy in Uganda, the history of ICT and health-related policies in Uganda was investigated. The year 2000 marked the beginning of implementation of key ICT and health sector reforms in Uganda [12]. In order to understand the prevailing conditions in the digital health/eHealth sector in Uganda, an exploration of the health policies and ICT policies that have been established since then was undertaken, with a focus on the steps that resulted in the final promulgation of eHealth reforms in Uganda.

FAIR Equivalency with Regulatory Framework for Digital Health in Uganda
In the early 2000s, the Government of Uganda realised that there were new issues emerging due to the Internet and the rapid changes in technology. To address these new trends, the government approved the National ICT Policy Framework in 2003 to guide the development of the ICT sector in Uganda. In order to harmonise the different aspects of ICT scattered in different ministries, the Ministry of ICT was created in 2006 in order to bring all aspects of ICT under one roof [13]. To provide further guidance to the ICT sector, a number of policies have been put in place (Figure 1), including the:  The National ICT Policy, which was launched in 2014, states that due to the dynamic nature of the ICT sector, a number of areas have emerged that necessitate the government's policy pronouncement. Among these areas is eHealth [13]. Subsequently, the Ministry of Health developed policies regarding eHealth. These included a draft National eHealth Policy (2013) and a draft National eHealth Strategy (2013), which subsequently gave birth to the National eHealth Policy (2016) [14] and the National eHealth Strategy (2017) [15]. These were meant to guide the use of ICT in supporting health sector transformation, because the Ministry recognised the potential of ICT to transform healthcare delivery by enabling information access and supporting healthcare operations, management, and decision making [14].

Uganda's Na tional eHealth Vision
An investigation into the extent to which the conditions in the digital health/eHealth ecosystem in Uganda allow for the opening of a policy window for the adoption and implementation of the FAIR Guidelines was undertaken to understand the governance framework of the digital health/eHealth sector in Uganda. As mentioned above, the eHealth sector in Uganda was launched by the publication of the National eHealth Policy (2016) [14] and, subsequently, the National eHealth Strategy (2017) [15]. The Government of Uganda, through the Ministry of Health, identified 13 priority areas, which it called pillars, on which emphasis must be placed in order to realise the National eHealth Vision (Figure 2). These pillars were promulgated in the National eHealth Policy and the National eHealth Strategy. A leadership and governance framework has been assigned for the implementation of each pillar, along with an action plan and five-year timeframe in which it should be done [15].

FAIR Equivalency with Regulatory Framework for Digital Health in Uganda
Each of the 13 pillars was assigned a budget and an action plan for the five-year period (2016/17-2020/21). For each pillar, the strategic objectives and associated strategic initiatives were designed to be implemented in three phases: Phase I (2016/17-2017/18); Phase II (2018/19-2019/20) and Phase III (2020/21).

Identificati on of Relevant Documents
A total of 14 documents, including plans, policies/policy frameworks, acts (of parliament), strategies and programmes (here and thereafter referred to as 'policy documents'), were identified as relevant to the research objectives in this study (see Table 1). These policy documents are ICT/health-related and published by the Government of Uganda. Some of the documents were obtained from ICT personnel in the Ministry of Health and others from the websites of relevant government institutions, as shown in Table 1

.6 Mention of the FAIR Guidelines in Policy Documents
To find out whether or not the FAIR Guidelines are mentioned in the policy documents selected as relevant to the eHealth sector in Uganda (Table 1), the documents were analysed using the coding and labelling method to see if they explicitly (or implicitly) mention the components of the FAIR Guidelines (Findable, Accessible, Interoperable and Reusable). In order to support a wide range of special circumstances, the FAIR Guidelines were intentionally designed to be minimally defined, concise, high-level principles [1]. In the 2016 paper that first described the principles, the authors were very brief in their explanations [1]. The GO FAIR website was used to obtain a more detailed explanation of the principles in order to accurately interpret them during the analysis of the policy documents [2,17].
After a comprehensive review of the policy documents, it was discovered that the FAIR Guidelines, as published in the FAIR foundational article by Wilkinson and colleagues in 2016 [1], are not mentioned in any of the policy documents analysed. However, certain terms (single words) or phrases (statements) that are semantically similar to the FAIR Guidelines have been used in the policy documents and, therefore, show a similar policy direction. For example, keywords such as integrate, harmonise, coordinate, incorporate, collaborate, one stop centre, common systems, common infrastructure, shared services, information exchange, alignment, open-source, and common language/vocabulary all relate to 'Interoperability'. The term 'FAIR Equivalent' was, therefore, coined to describe such terms and statements.

FAIR Equivalency with Regulatory Framework for Digital Health in Uganda
Whether or not a statement contained in the policy documents is equivalent to the FAIR Guidelines was ascertained based on whether or not it carries the same intent/purpose or meaning as the one provided by the original FAIR article [1]. For example, where the documents advocate for the use of standards with the intent/purpose of enabling data exchange, the specific statement was labelled as equivalent to the FAIR Guideline of 'Reusability', which requires that data and metadata meet domain-relevant community standards. Statements in the policy documents were labelled by appending the initials of the FAIR Guidelines that they represent, i.e., 'F', 'A', 'I', or 'R' (Supplementary Material Basajja, 2021). The policy documents were arranged in rows. 'FAIR Mention' and 'FAIR Equivalent' were listed in two columns and annotations were tabulated using a binary code system: '1' was assigned to the FAIR Mention or FAIR Equivalent column in cases where the policy documents mentioned either the FAIR Guidelines or their equivalent (FAIR Equivalent), whereas '0' was assigned in cases where neither of the two were mentioned (see Table 2 in Results). This spreadsheet approach is similar to the framework method, an increasingly popular approach to the management and analysis of qualitative data in health research [18][19][20].

Level of FAIR Equivalency in Policy Documents
In order to understand the level of FAIR Equivalency in the policy documents analysed, the analysis was extended to indicate which of the equivalent of the 15 sub-criteria of the FAIR Guidelines (hereafter, referred to as 'FAIR facets') are mentioned in the policy documents. These FAIR facets comprise the following:

FAIR Equivalency with Regulatory Framework for Digital Health in Uganda
The facets of the FAIR Guidelines are: • To be 'Findable': -F1: (meta)data are assigned a globally unique and persistent identifier -F2: data are described with rich metadata (defined by R1 below) -F3: metadata clearly and explicitly include the identifier of the data it describes -F4: (meta)data are registered or indexed in a searchable resource • To be 'Accessible': -A1: (meta)data are retrievable by their identifier using a standardised communications protocol -A1.1: the protocol is open, free, and universally implementable -A1.2: the protocol allows for an authentication and authorisation procedure, where necessary -A2: metadata are accessible, even when the data are no longer available • To be 'Interoperable': -I1: (meta)data use a formal, accessible, shared, and broadly applicable language for knowledge representation. -I2: (meta)data use vocabularies that follow FAIR Guidelines -I3: (meta)data include qualified references to other (meta)data • To be 'Reusable': -R1: meta(data) are richly described with a plurality of accurate and relevant attributes -R1.1: (meta)data are released with a clear and accessible data usage licence -R1.2: (meta)data are associated with detailed provenance -R1.3: (meta)data meet domain-relevant community standards A detailed and comprehensive analysis of the mention of the equivalent of the 15 FAIR facets in each of the policy documents was carried out and organised by codes (the text) and labels (the 15 FAIR facets) [18]. In this analysis, the mention of the equivalent of the FAIR facets in the policy documents was labelled by appending the appropriate FAIR facet (i.e., F1, F2, F3, F4; A1, A1.1, A1.2, A2; I1, I2, I3; R1, R1.1, R1.2, R1.3) to the corresponding statements (Supplementary Material 2 Basajja, 2021). The policy documents were arranged in rows while the FAIR facets were arranged in columns. For each policy document, the mention of the equivalent of the FAIR facet was scored '1', while the lack of mention of the equivalent of the FAIR facet was scored '0' in the corresponding data cell of the Microsoft Excel spreadsheet (see Table 3 in Results). From this data, the FAIR Equivalency Score (FE-Score) was calculated by aggregating the scores across all the 15 facets per policy document. Hence, the highest possible FE-Score for each policy document is 15 (Supplementary Material 3, Basajja, 2021).

Mention of FAIR Guidelines in Policy Documents
The analysis of whether or n ot the policy documents mention the FAIR Guidelines found that although none of the documents specifically mention the FAIR Guidelines, 10 out of the 12 documents (83%) mention the equivalent of the FAIR Guidelines (see Table 2). Only two documents do not mention any

FAIR Equivalency with Regulatory Framework for Digital Health in Uganda
aspects that relate to the FAIR Guidelines. This means that 80% of the documents relate to FAIR Guidelines to some extent. The documents that do not pertain in any way to FAIR Guidelines are from 2009 and 2012, so these are relatively early policy documents.

FAIR Equivalency with Regulatory Framework for Digital Health in Uganda
In order to determine if the policy documents become more FAIR Equivalent over time, a score called the FAIR Equivalent Score (FE-Score) was calculated. The FE-Score is an aggregate score of the mention of the equivalent of the FAIR facets/sub-criteria in all the policy documents. Each mention of any of the 15 FAIR facets/sub-criteria was scored '1', while no mention of the sub-criteria was scored '0' across all documents. The FE-Score was obtained by aggregating the scores for each of the 15 sub-criteria per policy document, with the highest possible score being 15 (Table 3 and Figure 4).

DIS CUSSION
Even though none of the documents specifically mention the FAIR Guidelines, the use of FAIR Equivalent terminology shows that there is not a large gap between the position of the stakeholders that need to be convinced to accept the FAIR Guidelines and the foundation on which these principles are based. That is to say, cultural entropy is small, so it is not likely to be a major hindrance to the acceptance of the FAIR Guidelines.

FA IR Equivalency
The results of the FAIR Equivalency analysis show that, across all of the policy documents, the facets for 'Accessibility' and 'Reusability' are mentioned most frequently (with a sum 22 and 18, respectively). 'Findability' and 'Interoperability' facets feature less often (with a sum of 12 and 13, respectively). The most mentioned FAIR Equivalent facets are A1 '(meta)data are retrievable by their identifier using a standardised

FAIR Equivalency with Regulatory Framework for Digital Health in Uganda
communications protocol' and I1 '(meta)data use a formal, accessible, shared, and broadly applicable language for knowledge representation'. While the least-mentioned FAIR Equivalent facets are F3, A2 and I2.
Regarding the least-mentioned FAIR Equivalent facets, Wilkinson and colleagues [1] described these facets as follows: F3-'metadata clearly and explicitly include the identifier of the data it describes'; A2-'metadata are accessible, even when the data are no longer available'; I2-'meta(data) use vocabularies that follow FAIR Guidelines'. The reason why these FAIR Equivalent facets (F3, A2, and I2) are the least mentioned in the policy documents is because they are data-centric, i.e., they specifically deal with details about data and metadata, which is beyond the scope of the documents analysed, as they are policy-level documents and, therefore, do not explicitly mention such things. Therefore, it is only after the policies or guidelines in the documents have been implemented in a particular context that one can ascertain whether or not the particular implementation conforms with these facets.

Fa cet A1
Facet A1 requires a standardised communications protocol, such as https or ftp, from which data and metadata can be retrieved by their identifier. Therefore, both human and automated users should be able to access the data/metadata by simply following a link. Facet A1 acknowledges that it may not always be possible to allow fully automated access to data, for example, in the case of highly-sensitive data. In such cases, providing contact details (such as an email address or telephone number) for the person who can discuss access to the data still renders the data FAIR [17].

FAIR Equivalency with Regulatory Framework for Digital Health in Uganda
from the Ministry of Health and beyond. This provides a single point of access to valuable information that facilitates evidence-based decision making [21]. A government web portal and local government websites were also identified among the key e-government services as facilities for displaying and disseminating government information and services on the Internet via the World Wide Web [22]. The National e-Government Policy Framework and the National ICT Policy Framework mention implementation of a government open data system to enable citizens to access government information, as one of the priority e-government applications identified in a survey [13,22]. The National Development Plan (NDP) II Public Investment Plan also includes "Development of a Monitoring and Evaluation Web Portal with Database functionality to act as a one stop centre for information on monitoring and evaluation of Government Programmes", as one of the NDP II pipeline projects [23]. In order to achieve NDP II objectives towards

FAIR Equivalency with Regulatory Framework for Digital Health in Uganda
ICT sector targets, one of the interventions is to establish regional information access centres and promote the use of ICT tools for trade, service delivery, and the exchange of information [23].
The Health Sector Development Plan (HSDP) II states that, during the plan period (2015/16-2019/20), special attention will be given to establishing a functional community information system by focusing on the following programme areas: a routine health management information system (HMIS), surveillance, vital statistics, research, health surveys and innovative eHealth solutions. Some of the key interventions in the routine HMIS area include: develop a National Health Information System Framework (NHISF) to create a reliable and accessible environment for managing health data [24].

Fa cet I1
Facet I1 is described as: 'meta(data) use a formal, accessible, shared and broadly applicable language for knowledge representation' [1]. Facet I1 requires use of commonly-used controlled vocabularies and ontologies, and a good data model with a well-defined framework to describe and structure data and metadata. The National eHealth Policy [14], National eHealth Strategy [15] and the National e-Government Policy Framework [22] mention the adoption and implementation of the Enterprise Architecture and Interoperability Framework for both e-government and eHealth implementations, which fulfils these requirements.
The National eHealth Strategy advocates for the development of an eHealth Enterprise Architecture and Interoperability Framework [25], with which all eHealth investments will have to conform and be certified. The National eHealth Strategy states that: Enterprise Architecture defines the structure for design and implementation of eHealth systems, linking the systems to be interoperable and using defined standards, so that health information and data from various sources can be linked and integrated to provide a better understanding of how the health sector is delivering overall health services. In so doing, eHealth as a whole experience reduced risk of fragmentation, less duplication of effort, and greater interoperability. [15] In addition, it calls for the implementation of a health information exchange and a shared electronic health record to allow for the sharing of information among health providers. This is the equivalent of the FAIR 'Interoperability' facet (I).
The Ministry of Health has adopted the District Health Information Software version 2 (DHIS2) as part of the national eHMIS and rolled it out up to the district level [26]. However, data from health facilities below the district level including community-based health data, data from vertical programmes, such as HIV/AIDS, tuberculosis and malaria programmes (which are run by non-governmental organisation or private health providers), and data from some referral hospitals (which are semi-autonomous) are not yet integrated into the national eHMIS/DHIS2 [15]. One of the strategic objectives of the National eHealth Strategy is to develop a community-based health information system and integrate information from related systems and vertical programmes into the national eHMIS [15]. This is in line with the 'Interoperability' principle of FAIR.

FAIR Equivalency with Regulatory Framework for Digital Health in Uganda
Various databases and information systems are mentioned in the National eHealth Policy and the National eHealth Strategy. These information systems have been developed and adopted to varying degrees, but are not yet integrated. Examples of these systems include: Human Resources for Health Information System (HRHIS), Clinical Information System (CIS), the national electronic Logistics and Supplies Management System (LSMIS), health facility systems, the provider and patient/client registry system, and so forth [14,15]. Details about the planned integration of these systems can be found in Supplementary Material 4 Basajja, 2021.

Fa cet F1
Perhaps the most important of the FAIR Guidelines is 'Findability', especially facet F1, which requires that data and metadata be assigned a globally unique and persistent identifier in order to be findable [1]. This is because without an identity that is unique, it is not easy for humans or machines to find a digital object, let alone determine if that object is reusable in their context. Therefore, the remaining three principles of FAIR (Accessibility, Interoperability and Reusability) are largely, or even entirely, hinged on 'Findability'.
Both the National eHealth Policy and National eHealth Strategy acknowledge that a unique health identifier is the cornerstone of most eHealth systems [14,15]. The National eHealth Policy states that there is no unique identifier for health records and that health records are duplicated [14]. The National Identification and Population Databank Project established by the Government of Uganda contains citizens' bio-data including biometric features stored in a central national database [22]. The National Identification Project provides a unique identification number (the National Identification Number, or NIN) for every individual.
Both the National eHealth Policy and the National eHealth Strategy advocate for the reuse of existing initiatives, such as the NIN, as the unique health identifier for electronic medical records, electronic health records, and personal health records [14,15]. Adoption of the NIN as the unique identifier for every patient/ client and enforcing its use in the electronic medical records system is one of the strategic objectives of the National eHealth Strategy [15]. In addition to the patient/client unique identifier, there is also the health facility/health provider unique identifier that the Ministry of Health plans to adopt and enforce [15]. The NIN is the equivalent of the FAIR facet F1.
The Ministry of Health has adopted the Open Medical Records Systems (OpenMRS) as part of an effort to establish an electronic medical records system. OpenMRS is an open-source, electronic health record system supported by a large global network of developers and implementers and used in over 40 countries [27,28]. The system will be used to electronically share an individual's medical record from one medical department to another. It will also help in tracking patients to address loss-to-follow-up of HIV patients, pregnant mothers who miss the four antenatal visits and tuberculosis patients as they receive their treatment. The Ministry planned to roll-out OpenMRS in 1,000 health facilities by the end of 2017 and adopt OpenMRS in more than 425 health facilities [29].

FA IR as a Solution to the Challenges in Uganda's eHealth Sector
The FAIR Guidelines are firmly rooted in the premise that data from publicly-funded projects is a public good. These principles provide the necessary practices to make data accessible to the public so that the maximum benefit can be derived from them. However, the adoption of the idea of the FAIR Guidelines, as a precursor to the establishment of the Internet of FAIR Data and Services, depends on the opening of a policy window [4,10].
But, how can a policy window be opened for the adoption of the FAIR Guidelines in Uganda's eHealth sector? Van Reisen and others note that the lack of adoption of the FAIR Guidelines in most research domains is partly because those domains do not see FAIR data stewardship as offering a solution to a problem that needs to be solved [4]. If the FAIR Guidelines are presented as a solution to the challenges faced by eHealth implementation in Uganda's eHealth sector, this may pave the way for the opening of a policy window for the adoption of the FAIR Guidelines.
Uganda's National eHealth Vision is structured around 13 eHealth pillars [14]. Acceptance of the FAIR Guidelines in Uganda's emerging eHealth sector is dependent on aligning the FAIR Guidelines with the National eHealth Vision. This can be done by exploring the barriers and opportunities in eHealth implementation and presenting the FAIR Guidelines as a workable solution. Eight eHealth pillars in which the FAIR Guidelines can help remove barriers to eHealth adoption/implementation were identified. These pillars also contain opportunities wherein FAIR can be accepted, adopted and subsequently implemented. These pillars include: 'Leadership and Governance', 'eHealth Enterprise Architecture', 'eHealth Services', 'Infrastructure', 'eHealth Information Assurance', 'Human Resources and Capacity Building', 'Change, Adoption, Business Process Re-engineering, and Transitioning', and 'Stakeholder Engagement'.
The 'eHealth Information Assurance' pillar relates to the security, confidentiality, privacy and integrity of electronic health data. For example, under this pillar, the Ministry of Health plans to review the National Information Security Framework and the Data Protection and Privacy Law to ensure that they address all the electronic health information security requirements in line with the existing regulatory environment, such as the Access to Information Act and international best-practices (e.g., the Global Health Information Privacy and Protection Statement [GHIPPS]) [15]. There is an opportunity for the FAIR Guidelines to be considered as part of this review, if they are presented as international best practice. The GO CHANGE pillar of FAIR [30] has a crucial role to play in this by leveraging/lobbying the relevant stakeholders (particularly those responsible for implementation of this pillar) to bring about alignment and synergy between the 'eHealth Information Assurance' pillar and the FAIR Guidelines. One of the strategic initiatives of the 'eHealth Information Assurance' pillar is "Develop and operationalize eHealth Information Security Guidelines and eHealth Information Privacy and Protection Guidelines, and monitor and enforce compliance to these guidelines by all stakeholders'' [15].
Due to security and privacy concerns regarding personal data, these data are usually completely closed off, making them totally inaccessible to both humans and machines. The FAIR Guidelines can be presented as a solution in this case, because they ensure the publication of rich metadata to facilitate discovery, even in the case of sensitive or personally-identifiable data, including clear rules regarding the process for accessing the data, which provides a high degree of 'FAIRness' even in the absence of FAIR publication of the data itself [1]. Therefore, by implementing the FAIR Guidelines, the 'eHealth Information Assurance' pillar can ensure that metadata is always public, even if the data are restricted or removed for privacy reasons, supporting the FAIR facets of 'Findability' and 'Accessibility'. In addition, regarding data anonymization/depersonalisation during aggregation due to privacy concerns, adopting FAIR can ensure that metadata is always public to enable 'Findability'. Another strategic initiative of the 'eHealth Information Assurance' pillar is "Develop data handling procedures and protocols to ensure consent/appropriate disclosure" [15]. In this vein, the FAIR Guideline of 'Reusability' requires meta(data) to be released with a clear and accessible data usage licence indicating dataset terms, who has the right to use the data, and how they can seek permission to use the data [1]. These principles can be built into the data handling procedures and protocols, thereby increasing the degree of FAIRness of the data.
Under the 'eHealth Services, Information Sharing and Data Management' pillar, the Ministry of Health's strategic objective is to establish an eHealth data, information and knowledge management, analysis and utilisation system. Under this strategic objective, there are the following strategic initiatives: develop data, information and knowledge sharing agreements that define the purposes for which health data may be shared between organisations; develop guidelines regarding data ownership/stewardship, data creation and collection, data quality, data audits and data analysis; create consistent national health data standards, definitions and dictionaries in accordance with the 'eHealth Enterprise Architecture'; develop a right-to-use policy that defines the purposes for which data can be used by health providers in the direct delivery of care to the patients/clients; and develop guidelines regarding the use of data analytics and health intelligence tools to access the vast amounts of digital health data, such as electronic medical records stored in clinical data repositories, as permitted by privacy regulations [15]. The FAIR Guidelines can be applied to all these areas, hence, this provides a perfect opportunity to adopt and implement them.
The electronic delivery of health services through mobile devices (mHealth) is another strategic objective of the Ministry of Health. Under this objective, the Ministry intends to develop and implement the required mHealth infrastructure used to develop mobile apps including the front-end, middleware (to connect the front-end with the back-end), and back-end (data access layer, server-side) [15]. Most digital health solutions use mobile devices to reach disadvantaged and underserved groups in the population. Therefore, if the FAIR Guidelines were adopted by the Ministry of Health in the development of mHealth infrastructure, such as application programming interfaces (APIs), these could subsequently be used by digital health providers to interface with the national HMIS, which would be an important step forward in the provision of FAIR Data and Services. Likewise the private sector, including mHealth app developers, can be engaged by the GO TRAIN and GO BUILD FAIR communities [31] on how to build mobile software that adheres to the FAIR Guidelines. Therefore, all the three pillars of GO FAIR-GO CHANGE, GO BUILD and GO TRAIN [31]should put maximum effort into this area, given that it has important implications for the acceptance and adoption of the FAIR Guidelines in the digital health/eHealth sector and the establishment of the Internet of FAIR Data and Services [32].

FAIR Equivalency with Regulatory Framework for Digital Health in Uganda
Under the 'Human Resources and Capacity Building' eHealth pillar, the National eHealth Strategy acknowledges that a sufficient number of personnel with eHealth competencies and skills is a critical factor in the successful implementation of eHealth in Uganda. The Ministry intends to carry out a needs-assessment to evaluate the skills and competencies of its health professionals with the goal of developing and enforcing an eHealth curriculum framework to be followed by different training providers in developing and delivering health training. The Ministry also plans to: develop accredited courses/qualifications in eHealth (such as in health informatics); provide basic ICT training for health workers at all levels (continuous professional development) through e-learning and digital resources for offline learning; develop a national eHealth knowledge repository; and establish national eHealth centres of excellence [15]. This provides a unique opportunity for the adoption of FAIR Guidelines, which can be embedded in the eHealth curriculum, so that health professionals are not only equipped with eHealth skills, but also with knowledge of the FAIR Guidelines, which will be crucial in the building of the Internet of FAIR Data and Services. Towards this, the GO CHANGE and GO TRAIN pillars [31] will have to engage in serious consultations with relevant stakeholders to convince them to include the FAIR Guidelines in the eHealth curriculum and to participate in the training of the eHealth workforce. This training may be carried out at established eHealth centres of excellence. The author proposes that a regional (East Africa) FAIR Centre of Excellence be established by the Africa GO FAIR Implementation Network, in which professionals from various domains/backgrounds can receive instruction on the relevance and application of the FAIR Guidelines in their domain. Such a regional FAIR Centre of Excellence would be akin to the incubation/innovation hubs in the ICT sector, where the GO CHANGE, GO TRAIN and GO BUILD pillars [31] can come together and collaboratively perform their respective responsibilities. Such regional hubs have been used before to provide capacity building in new scientific fields. An example is the Biosciences Eastern and Central Africa, International Livestock Research Institute (Beca-ILRI) hub, which is located in Nairobi, Kenya [33,34], where researchers from various sub-fields in the biosciences in the region can go to receive basic training in bioinformatics and genomics in order to create competencies in this new area. This regional FAIR Centre of Excellence can also give the FAIR Guidelines and GO FAIR movement much needed visibility.
Under the 'Infrastructure' eHealth pillar, the National eHealth Strategy calls for the utilisation of appropriate mature and emerging technologies to support eHealth development and enhance core eHealth services. Under this eHealth pillar, the Ministry of Health will assess the appropriateness of mature and emerging technologies for the support of health in the Ugandan situation, develop guidelines for such technologies, and continuously identify and review emerging technologies in healthcare to establish their applicability in the Ugandan healthcare environment. The National eHealth Strategy defines emerging technologies as new technologies that are currently developing, or will be developed, over the next 5 to 10 years, and which will substantially alter the business and social environment. Some of the examples of mature and emerging technologies mentioned are cloud computing, big data and open data, data analytics, smart systems, digital services and the Internet of things [15].
Van Reisen et al., explain that the motivation to listen to a message about the FAIR Guidelines is affected by the perceived usefulness of the message, and the objectives, opinions and beliefs the receiver has about the social process of data management [4]. Although the FAIR Guidelines are not, in themselves, a technology, standard or specification [1], the fact that the Ministry of Health is invested in continuously identifying and reviewing technologies to support eHealth development and enhance eHealth services, and by virtue of the substantial impact that the FAIR Guidelines can have on the Ugandan healthcare environment (e.g., solving the problem of fragmentation of eHealth solutions) [5], there is an opportunity for the adoption of the FAIR Guidelines. Therefore, the social entrepreneurs and the GO CHANGE pillar of FAIR that are tasked with convincing and motivating potential users of the FAIR Guidelines to embrace them [4] should take advantage of this opportunity to create a policy window for the adoption of the FAIR Guidelines.
The 'Change, Adoption, Business Process Re-engineering and Transitioning' eHealth pillar is concerned with establishing a comprehensive eHealth transitioning strategy to take care of rapid changes in technology. Rapid technological changes are a challenge to eHealth implementation, because different organisations have different innovation cycles, hence, adopt different technologies, which are usually incompatible [15]. This pillar seeks to manage the rapid changes in technology during eHealth implementation by enforcing adherence to the eHealth 'Enterprise Architecture', which ensures standardisation and interoperability, such that any new technologies can integrate with existing ones [15]. This is partly in line with the FAIR Guidelines, so there is an opportunity to fully align and conform this objective to the FAIR Guidelines. Moreover, the Ministry of Health Resource Centre (which is tasked with implementation of this eHealth pillar) is responsible for "keeping abreast with current trends and forecasting changes early enough to enable review of plans" [15]. The GO CHANGE pillar may, therefore, liaise with the Ministry of Health Resource Centre to motivate and encourage adoption of the FAIR Guidelines in this pillar.
The 'Stakeholder Engagement' eHealth pillar has the following roles: engage all eHealth stakeholders at the national, regional and international levels; carry-out mass campaigns and training to create awareness of eHealth services and applications among target groups; and establish a framework for measuring the effectiveness of engagement and awareness campaigns [15]. Here, there is an opportunity for the FAIR Guidelines to be included alongside eHealth during stakeholder engagement and awareness creation so that the mass campaigns advocating for adoption of eHealth also call for the adoption of the FAIR Guidelines.
Another way in which the policy window for the adoption of the FAIR Guidelines can be opened is if they are established as part of international best practices for good data management and stewardship. Many guidelines, protocols, procedures and regulations that have been established as international best practices are quickly adopted, because policymakers and stakeholders always carry out benchmarking and extensive consultations when establishing a new area (for example, eHealth). These stakeholders always seek to align with the international best practices when developing or reviewing guidelines, regulations, protocols and procedures. For example, if the FAIR Guidelines were to be established by the World Health Organization as part of international best practices in eHealth implementation, they would easily be adopted by all downstream institutions that seek to align their eHealth vision with international best practices. For example, Uganda adopted the eHealth Enterprise Architecture and Interoperability Framework to guide the implementation of its eHealth strategy after carrying out a benchmarking/review of eHealth experiences both internationally (Australia, England, Scotland, Northern Ireland, Denmark, Cuba, Philippines and Canada) and regionally (South Africa, Nigeria, Rwanda, Tanzania, Kenya and Ghana). The output from this review was the definition of the international best practice/criteria for the optimum eHealth governance and eHealth implementation solution for Uganda [14].

CONCLUSION
Implementing the FAIR Guidelines in the digital health/eHealth sector in Africa has the potential to solve problems-such as the lack of sustainability of digital health initiatives and fragmentation and lack of integration of health service solutions-as long as they are adopted and implemented with contextual awareness. The general objective of this study was to assess the governance framework for digital health/ eHealth policy in Uganda and explore the possibility of a policy window opening for the FAIR Guidelines to be adopted and implemented in Uganda's eHealth sector. The key policy documents relating to information, communication and technology (ICT) and health that are pivotal to the emerging eHealth sector were identified. The 12 policy documents obtained were then analysed using a comprehensive coding and labelling approach that employs a binary system using keywords and phrases to indicate whether or not the documents mention the FAIR Guidelines (that data be Findable, Accessible, Interoperable and Reusable), or the equivalent of 15 FAIR sub-criteria/facets, in order to determine the level of FAIR Equivalency. A FAIR Equivalency Score (FE-Score), which is the sum of the 15 FAIR sub-criteria/facets, was then obtained for each policy document. The analysis found that although none of the documents mention the FAIR Guidelines, 10 out of the 12 documents (83%) mention the equivalent of the FAIR Guidelines. The FE-Score, which was devised for this study as an aggregate score of the mention of the equivalent of FAIR facets in the policy documents, revealed that Uganda's National eHealth Policy, National eHealth Strategy and National ICT Policy Framework have the highest FE-Scores. The higher the FE-Score, the more in sync the policy document is with the FAIR Guidelines. These three documents are the most recently drafted of the 12 selected policy documents.
Whereas the FAIR Guidelines were not mentioned in any of the policy documents in Uganda's eHealth sector, 83% of the documents mentioned FAIR Equivalency efforts, such as the adoption of the National Identification Number as a unique identifier in Uganda's national eHMIS ('Findability'), planned/ongoing integration of the various information systems ('Interoperability'), and alignment of various projects with international best practice/standards ('Reusability'). These findings indicate that there is not a large gap between the position of the stakeholders that need to be convinced to accept the FAIR Guidelines and the foundation on which the FAIR Guidelines are based. Therefore, it can be concluded that favourable conditions exist for a push to adopt and implement the FAIR Guidelines in Uganda's eHealth sector.

LIMITATIONS OF THIS STUDY
Efforts were made to obtain all published documents relevant to the research objectives by searching the websites of relevant NGOs and government institutions. Key personnel in the designated ministries, departments and agencies were also contacted and formal requests for the relevant documents submitted. However, two documents that were considered relevant to this study were not included, because they were not available online and were not provided upon request. Regarding scope, this research article focused

FAIR Equivalency with Regulatory Framework for Digital Health in Uganda
on national-level documents published by the Government of Uganda (policies/policy frameworks, strategies, plans, legislation and programmes) in order to obtain the governance overview required by the research objective. Annual health sector performance reports, evaluation reports, survey reports and programme reports were not included.