Keywords

1 Introduction: The Promise and Peril of New Technologies

The last two decades have witnessed an explosion of mobile networks, cloud computing and new technologies introduced within the healthcare field. These technologies result to incomprehensibly large volumes of information generated, ingested and translated- which then have to be applied as part of routine practice, both at the individual level, e.g., at a patient-treating physician level, as well as at the population level, e.g., at a public health/governmental level. Furthermore, the handling of this data (from the generation to the end-use) needs to be governed by well-defined ethical and regulatory frameworks, to ensure adherence to national and international requirements.

While the above may be achievable within high-income settings, the same cannot be stated for low-and middle-income countries (LMICs), where there are many competing healthcare needs and the continuum of data generation to end-use may not be fully achievable or not present at all. Thus, while there is a clear promise for digital technologies to positively influence healthcare provision, there is a danger that regarding LMICs, such technologies may be over-promising given the local needs and contexts. Indeed, the COVID-19 pandemic has provided ample evidence within LMICs, for introducing digitalization efforts that have performed very well (as described in Chap. “Ubiquitous and Powerful Artificial Intelligence (AI)” for example for Indonesia), and others that have performed less well (as described in Chap. “Long-Term Digital Storage and Usage of Research Data: Data Pooling” for Latin America). Therefore, digitalization in LMICs, needs to follow a careful balancing act, between the promise and peril of new technologies, akin to similar introductions previously in healthcare, e.g., the introduction of radiology in routine healthcare practice.

Reading the chapters of the current book, it becomes clear that the pathway to digitalization in LMICs is highly dependent on the individual contexts and needs of the country. Chapter “The Emergence and Growth of Digital Health in Saudi Arabia: A Success Story” provides the example of a mature digital healthcare ecosystem from Saudi Arabia, with very high interoperability between different healthcare operating units. Chapters “The Digital Divide Based on Development and Availability: The Polish Perspective”, “Potential of Digital Health Solutions in Facing Shifting Disease Burden and Double Burden in Low- and Middle-Income Countries”, “Health Inequalities and Availability: Needs and Applications”, “Long-Term Digital Storage and Usage of Research Data: Data Pooling”, “Digitisation in Genetics and Diagnostics Laboratories in Armenia” and “Ubiquitous and Powerful Artificial Intelligence (AI)”, captured the experiences of Poland, Vietnam, Cyprus, Latin American countries, Armenia and Indonesia respectively. Thus, this book has managed to provide a balanced, global view of the different digitalization pathways followed, and a comparative read of those chapters will certainly provide insights into similarities and differences. If we consider the national examples as vertical implementations (i.e., for only one specific location), a number of chapters look into the horizontal implementations (i.e., across a particular field of operations). As such Chaps. “Biobank Digitalization in Low-Middle Income Countries (LMICs): Current and Future Technological Developments” and “Digital Healthcare: Technologies, Technical and Design Challenges” investigated the overarching risks and challenges, and the infrastructure risks, needs and opportunities respectively. Furthermore, particular mention should be made for Chaps. “Digitalization of Physical Health Data in Low- and Middle-Income Countries” and “Universal Internet Access Supporting Healthcare Provision: The Example of Indonesia”, highlighting technical and design challenges, and investments and incentives respectively. Therefore, taking together the above contributions, thematic units can be identified, forming the core of any future recommendations for digitalization of healthcare in LMICs. These are described below, though the relative prioritization would be context-dependent.

2 Emerging Common Themes in LMIC Healthcare Digitisation

2.1 Data Infrastructure

The availability of data infrastructure is crucial for the success of any digitalization effort in LMICs. The current chapters describe a highly fragmented picture across regions as well as individual countries. There is a high overall need for further data infrastructure development across LMICs, including the provision of financial incentives and/or direct public health investment for the expansion of existing infrastructures (Al Knawy et al. 2020; Zhang et al. 2022). However, as with digital health itself, the data infrastructure does not need to copy blindly what was implemented in high-income settings, but develop methodologies and data architectures that are LMIC-friendly and suitably tropicalized in terms of their performance and maintenance needs (Muinga et al. 2020).

2.2 Data Laws and Regulations

In most of the LMICs that have been mentioned in the different chapters of this book, it is evident that there don’t exist specific laws and regulations designed and implemented with digitalization of healthcare in mind. This does not mean that legal frameworks do not exist. Though the case differs between different countries, data protection legislation has been extant for clinical trials, teleconsultations, collection of biological samples and associated clinical data for research, etc. (Camacho-Leon et al. 2022; Vodosin et al. 2021; Purtova et al. 2015). It is the case that most of these regulations have expanded their remit and interpretation to include the introduction of digital applications in healthcare. Perhaps this is a sufficient course of action given the relatively restricted digital healthcare applications in LMICs and the many other competing interests. However, if digitalization is to reach its full potential, it would need to have clearly defined laws and regulations on the collection of data, regulation of access, sharing (nationally/internationally) and use, reporting of data analyses, and secondary use of collected data.

2.3 Education, Education, Education

Finally, and almost universally mentioned, the lack of education has been identified as a major barrier for digitalization of healthcare in LMICs. It should be noted though, that the term education is used as a blanket term, addressing a number of different aspects and needs (El Benny et al. 2021). The medical students and young healthcare professionals report higher levels of digital literacy than other sections of healthcare, and as such the education is more likely to be task-specific deepening already existing skills (Kuek and Hakkennes 2020). However, educational needs also address entire populations, especially in the case of COVID-19 pandemic where surveillance systems are thought to have under-performed due to the limited reach of the internet, and reduced digital literacy of the general population (Hennis et al. 2021). It has to be noted, that education is not an instantaneous activity, but would require persistence and structural incorporation of existing activities if it is to be effective in the longer-term. This is particularly true for LMICs, where in many cases the levels of digital literacy still remain at low levels, e.g., beyond 25% of the population.

3 Way Forward Is Context-Driven, Asymmetric, Culturally Sensitive and Locally Autonomous

During the COVID-19 pandemic a number of high-level discussions took place, looking into the post-pandemic landscape, and have been summarized in relevant publications (Kozlakidis et al. 2020; AlKnawy et al. 2023; Jazieh and Kozlakidis 2020). However, a second look is now warranted for these early viewpoints, as the implementation experiences from the field can inform and enrich the ‘building back better’ propositions (Adisasmito et al. 2023). Given the intense discussions that took place during the creation of this book, the following key factors are emerging as critical:

Context-driven: The digitalization of healthcare in LMICs cannot follow a single common blueprint. Different countries find themselves at different stages of financial and infrastructural development, facing different combinations of healthcare and socioeconomic pressures. Being unaware of the context can potentially create unintended consequences such as biases, discrimination, errors or unexpected results, and an overall lack of transparency with regard to how outcomes are achieved (Stahl and Coeckelbergh 2016). The consideration of context is not only important within national boundaries, but also at the regional level, and can influence the possibilities and reach of data access and sharing.

Asymmetric: Information asymmetry is one of the key features separating healthcare away from a traditional market economy definition which assumes that all parties have access to perfect information in terms of their decision making and negotiating power (Major 2019). In healthcare, patients typically lack the medical knowledge that healthcare professionals possess, and this causes information asymmetry. Digitalization of healthcare in LMICs has the power to reduce this information asymmetry (e.g., by involving the patient in the information translation and democratizing the decision-making process), or to further increase the existing information asymmetry by enlarging the digital divide and thus, enlarging the information availability only for the one party. In particular for a market that is in its initial growth stages, enlarging information asymmetry can have a detrimental impact on the rate of market growth and quality of services rendered.

Locally autonomous: The many different challenges within LMICs, necessitate that challenges would be most effectively addressed within a local context. This also impacts the digitalization model in LMICs, where applications may be preferably locally customized, and operating autonomously, while bearing the capacity of supporting federated data access, allowing for national and international data analyses (Loftus et al. 2022). As an inherent part of being locally autonomous, digitalization would also need to be dynamic (i.e., capturing temporal changes in physiologic signals and clinical events), precise (using high-resolution, multimodal data, interoperable and complex architecture), and finally in the case of machine learning algorithms able to learn with minimal supervision and execute without human input. Thus, autonomy transcends both the technical and hierarchical aspects of healthcare provision.

Culturally sensitive and inclusive: While digitizing healthcare generates large amounts of data, that data is only valuable when it’s accessible. The success of any application will be contingent on delivering data in a form the end-user, whether patient or care-giver, can easily understand and comfortably engage with. This requires focusing data-extraction on information relevant to the end-user local environment. In LMICs where education may be limited and newer technologies unfamiliar, special attention must also be paid to identifying alternate modes of communication that can more effectively reach the end-user. These may need to go beyond language-specific models and appeal to other traditions and modalities, e.g., audio and visual-only applications or culturally-responsive implementations. Combining finely-tuned datasets with local knowledge in readily accessible formats improves uptake and expands the application’s scope to reach a wider proportion of the population, including, crucially, marginalized communities.

4 Conclusion

This Chapter serves as the epilogue of a very rewarding process, addressing an existing knowledge gap for LMICs. The digitalization of healthcare in LMICs is a continuing trend that has accelerated by the COVID-19 pandemic. However, as with any new technologies introduced into routine practice, promises and perils entail. The optimal digitalization of healthcare in LMICs would need to address three major challenges: the data infrastructure (including financial support for widening digital access), the creation of relevant legal and regulatory frameworks (including implementation evaluation frameworks), and the education to be made available to multiple audiences. Finally, the way forward for the digitalization of healthcare in LMICs is anticipated to be: context-driven, asymmetric, culturally sensitive, and locally autonomous.