In December 2025, a search was conducted across several databases, including PubMed, Google Scholar, Cochrane Library, and Google Search. Specific terms, outlined in the Box 1, were systematically combined using Boolean operators to capture a comprehensive set of relevant literature, policy documents, and global health reports on digital health transformation in LMICs. Additionally, more publications were identified through citation chasing of the selected studies. Targeted hand-searching of the websites and the publications of international organizations—such as the WHO, the World Bank, the United Nations Children’s Fund (UNICEF), and the United Nations Development Programme (UNDP)—was also carried out to locate relevant strategic frameworks, policy documents, and technical reports for inclusion in this review.

This narrative review includes studies and reports that met the following criteria: (a) Economic setting: The studies must have been conducted in or relevant to LMICs, as defined by the World Bank classification at the time of publication [11]. (b) Topic relevance: Publications should focus on digital health interventions, systems, or policies. This includes mHealth, telemedicine and telehealth, EHRs and health information systems (HISs), AI and data analytics in healthcare, digital tools for disease surveillance, service delivery, or health workforce support, ML, PM, AI-driven data analysis, big data analytics, and internet of medical things (IoMT). (c) Health system relevance: Studies must address health-system-level outcomes, such as access to care, quality of services, efficiency, equity, governance, workforce capacity, or health system resilience. This includes case studies, but excludes any disease-specific pilot projects. (d) Study type: Eligible studies include peer-reviewed original research articles, review articles (narrative, scoping, or systematic reviews), and policy-relevant grey literature from international organizations (e.g., WHO, World Bank, UNICEF, UNDP). (e) Time frame: Publications should be dated between January 2015 and December 2025 to reflect the contemporary evolution of digital health in LMICs. (f) Language: Publications must be available in English.

Studies were excluded from the review if they met any of the following criteria: (a) Economic setting: Studies conducted exclusively in high-income countries that do not explicitly discuss their relevance or applicability to LMICs. (b) Limited digital health relevance: Publications that primarily focus on general information and communication technologies without clear applications in healthcare. This category includes studies that describe digital tools without assessing their implementation, usage, or implications for health systems. (c) Clinical focus without health system relevance: Studies that concentrate narrowly on clinical efficacy or technical performance without addressing health system integration, implementation, policy relevance, or broader service delivery implications. (d) Non-scholarly sources: Opinion pieces, editorials, commentaries, or blog posts. (e) Time frame: Publications that were published before 2015. (f) Language: Publications that are not available in English.

The table below presents the study protocol, illustrating the literature identification, screening, selection, and synthesis process undertaken in this review (Table 1).

The author independently screened articles and grey literature identified during the initial search by reviewing their titles and abstracts. Publications that met the predefined inclusion criteria were selected for further assessment, while those that did not were excluded. The full texts of potentially relevant publications were then reviewed for final selection. The selection process was iterative, meaning that additional exclusions occurred during the synthesis and writing stages as thematic categories developed and relevance to the review objectives was reevaluated. Duplicate publications were identified and removed to prevent overlap in the synthesized evidence. A PRISMA flowchart was used to visualize the study selection process (Figure 1).

Data extraction and synthesis were conducted using an iterative narrative review approach. A data extraction table was created to systematically organize relevant information from the included literature (see Supplement Document). This table included details such as title, authors, journal or organization name, publication type, and the primary focus of each study. Following data extraction, the findings were analysed using a manual inductive thematic coding approach. Initial codes were generated from the extracted data based on recurring concepts, patterns, and issues identified across the literature. These codes were reviewed and grouped into broader thematic categories through an iterative process of comparison and refinement.

The findings were subsequently synthesized through thematic narrative analysis. Studies were grouped based on recurring concepts, digital health domains, health system functions, implementation contexts, reported challenges, and proposed opportunities or future directions. Initial thematic categories were developed based on the frequency, relevance, and consistency of topics identified in the literature. Key themes included interoperability, AI, big data analytics, PM, IoMT, blockchain applications, governance and regulation, workforce capacity, and equity-related challenges.

To enhance the reliability of the thematic analysis, the emerging themes were cross-checked against the original publications throughout the synthesis process. Studies assigned to each theme were reviewed repeatedly to ensure consistency between extracted findings and the thematic categories. Themes were refined through iterative comparison across studies, and only themes supported by multiple publications were retained in the final synthesis. In addition, the identified themes were compared with existing global digital health frameworks, including the WHO digital health strategies and related policy guidance, to assess their conceptual consistency and relevance. Particular emphasis was placed on themes relevant across multiple LMIC contexts, reflecting broader sustainability and health system implications. The final thematic structure was refined iteratively to ensure coherence, conceptual clarity, and alignment with the review’s objectives.

Interoperability is essential for digital healthcare systems [26,27]. EHRs play a vital role in improving health service delivery by centralizing patient information, improving data accessibility among providers, and enabling patients to access their own medical data [6,28,29]. Furthermore, integrating EHRs with biobanks and genomic databases has demonstrated significant potential to advance PM initiatives [26].

Big data analytics is crucial for transforming healthcare practices and research. It facilitates the acquisition, management, analysis, and integration of large volumes of diverse health data, both structured and unstructured [36]. This process leads to improved data-driven clinical decision support, optimized care delivery, and enhanced evidence-based health planning. Key areas of focus in this field include medical image processing, signal processing, and genomics [42]. High-dimensional imaging data are utilized to train deep learning models for diagnostics, risk assessment, and treatment planning. Additionally, continuous medical signals allow real-time monitoring, predictive analytics, and risk assessment. Advances in genomic sequencing further enhance AI-driven PM by enabling large-scale pattern recognition and risk stratification [36,42]. An example of effective implementation in Armenia is the development of a national cancer registry that incorporates genetic and genomic testing data. This registry is accessible to all clinics and laboratories nationwide and provides vital support for cancer patients. It is used for genetic testing and counselling for individuals with a high genetic predisposition and plays a role in PM initiatives [43]. By integrating various datasets into interoperable digital ecosystems, big data analytics empowers LMICs to implement predictive, personalized, and population-level interventions that address gaps in their health systems.

PM represents a significant shift in the prevention and treatment of diseases. It combines clinical, genetic, genomic, and epigenetic data to predict individual vulnerability, disease progression, and response to therapies [44,45,46,47]. Advances in AI and ML, along with interoperable data ecosystems, are essential for the successful implementation of PM [44]. By utilizing genomic sequencing and advanced analytics, clinicians can more accurately predict and tailor individualized treatment plans, reduce trial-and-error prescribing, and minimize adverse drug reactions [45,47,48]. In LMICs, PM offers considerable promise and substantial structural challenges. Key opportunities for PM in these regions include managing infectious and non-communicable diseases and applying pharmacogenomics [46,48,49].

Recently, there has been a shift toward “Precision Public Health,” which extends the principle of PM from focusing solely on individuals to implementing population-level interventions [39,50]. This approach uses integrated data on genomic and environmental factors, as well as social determinants of health, to enhance outbreak prediction, assess climate-related health risks, and guide targeted planning and resource allocation [39,50]. Considering that climate change disproportionately affects LMICs, it is vital to incorporate environmental and climate data into the digital ecosystem to improve preparedness and resilience [41]. However, to fully realize the potential of these innovations, we need strong data governance frameworks, ethical oversight, interoperable data ecosystems, and ongoing investment in local genomic, digital, and computational infrastructure, as well as the development of workforce capacity [38,39,46,50]. Meeting these requirements will pose a significant challenge for many LMICs in the foreseeable future.

The concept of a “hyperconnected healthcare ecosystem” has become a reality in many countries, including several middle-income countries (MICs). IoMT refers to the network of medical devices—such as wearables, implantable sensors, and smart hospital equipment—that provide real-time data [51]. By integrating these connected devices with an interoperable health ecosystem, IoMT facilitates a more responsive, data-driven approach to PCC. IoMT enables remote care in several ways, including: (a) testing and tracing for disease spread for infectious diseases, (b) real-time health monitoring, particularly for chronic disease management, (c) utilizing healthcare monitoring devices like smart pens, smart wearable devices like smart watches, smart implants, smart tooth brushes, sleep trackers, and loneliness detectors, (d) developing smart hospitals and smart operating rooms, (e) implementing smart medication-automated medication management, (f) tracing digital biomarkers, (g) providing robotic assistance, (h) supporting virtual planning and simulation- and navigation-guided surgery; (i) offering mobile and smartphone applications, (j) utilizing voice assistants; (k) enabling ambient assisted living; (l) detecting adverse drug reactions; and (m) facilitating telemedicine [51]. Furthermore, the integration of Metaverse technology into IoMT, through augmented reality (AR), mixed reality (MR), and VR offers opportunities for surgical training, rehabilitation, remote collaboration, and enhanced patient engagement [52,53,54,55].

Blockchain technology offers a promising solution for enhancing data security, patient-centred interoperability, and supply chain integrity within the digital health ecosystem [27]. By facilitating decentralized, secure digital data ledgers, blockchain has the potential to improve medical record management, streamline insurance claims processing, and support biomedical research [56,57]. This technology not only supports institution-driven interoperability models but also promotes a patient-centred interoperability approach, allowing individuals to access their own health information. This model is particularly relevant in LMICs, where internal migration is common, care pathways are often fragmented, and referral systems may be unreliable [27,56,58]. By shifting the control from institutions to individuals, blockchain-based systems could enhance transparency, trust, and accountability in health information exchange (HIE).

As digital health systems face an increasing threat of cyberattacks, blockchain provides enhanced protection of patient identity and data integrity through cryptographic validation and distributed consensus mechanisms [56]. Additionally, blockchain-based pharmaceutical traceability systems can help combat counterfeit medicines by tracking products throughout the supply chain, which is a critical concern in many LMICs [59,60].

Inadequate digital infrastructure remains one of the most significant barriers to digital health transformation in LMICs. Many of these countries continue to struggle with unreliable power supplies, limited broadband coverage, and unstable internet connectivity. These issues hinder the advancement of HISs, cloud-based EHRs, and AI-enabled applications [27,61,62]. In addition to connectivity limitations, LMICs also experience a “hardware gap” across all levels of their health systems, which restricts the potential for digital transformation [25,62]. Although mobile phone penetration is relatively high in many LMICs, the quality, speed, and affordability of data services are often insufficient for supporting data-intensive applications, such as real-time remote monitoring or AI-driven decision-making [17,18].

Furthermore, disparities in infrastructure and access to services are widespread in most LMICs, exacerbating inequities in digital health adoption [16,63,64]. A shortage of skilled personnel is another critical barrier to progress [65]. If foundational infrastructure gaps are addressed, investments in AI, interoperable health ecosystems, and DPI risk deepening exist inequities, limiting scalability, and preventing digital health innovations from achieving the intended impact in LMIC health systems [18,64,65,66].

The digital transformation in healthcare relies not only on the availability of digital infrastructure but also on workforce capacity and digital literacy of the workforce. A significant challenge identified in many LMICs is limited “digital readiness” [27,67]. Health workers often lack adequate training in digital applications, data interpretation, and AI-assisted decision-making, which diminishes the effectiveness and sustainability of digital health interventions. Additionally, there is a shortage of personnel with advanced technical and informatics skills, further hindering system maturity and preparedness for AI implementation [65]. Inadequate training in these areas leads to a lack of human resources, adversely impacting user safety and the overall effectiveness of health services [65].

Moreover, the outcomes of digital technologies can vary significantly among different population groups within a country or region [63]. Digital health initiatives risk exacerbating existing inequities by disproportionately benefiting individuals with higher levels of education or better digital skills. Patients with low literacy or limited experience using digital platforms may struggle to navigate applications or identify misinformation, which can ultimately erode trust in health systems [16,63].

Data governance and regulatory frameworks pose significant barriers to digital health transformation in many LMICs. Many of these countries lack comprehensive policies regarding data protection, privacy, interoperability, and ethical oversight of digital health technologies [27,61,68]. Although consent mechanisms are often included in digital applications, users frequently lack awareness and understanding of the potential risks associated with the use or misuse of their data [68].

The integration of AI into healthcare has intensified these ethical and regulatory challenges. Key concerns include algorithmic bias, a lack of transparency in “black-box” models, data privacy and safety risks, and unclear liability for AI-driven clinical decisions [25]. Most LMICs lack a legal framework to govern AI accountability, particularly when models are trained on datasets from high-income countries that may not accurately reflect local populations and contexts. Weak governance further exacerbates these challenges, leading to inefficiencies, errors, waste, and corruption that undermine trust and compromise equitable service delivery [69,70]. As digital health initiatives evolve from pilot projects to national strategies, the absence of unified legal and governance frameworks has become a critical obstacle.

Robust governance frameworks are essential for the ethical and secure use of AI in digital health systems, particularly concerning patient data access and processing. Such frameworks may include strong safeguards for data privacy, cybersecurity, and informed consent, along with clear accountability for AI-driven decisions. Without these governance structures in place, the risk of misusing sensitive health data, experiencing algorithmic bias, and encountering opaque decision-making processes increases, disproportionately affecting vulnerable populations in LMICs [70]. Additionally, weak regulatory environments may exacerbate existing health inequalities if AI systems are trained on non-representative datasets or deployed without adequate validation in local contexts. These challenges can undermine public trust in digital health systems and hinder adoption by both healthcare providers and patients [69]. It is crucial to develop transparent regulatory standards, ethical guidelines, and context-sensitive governance mechanisms to ensure that AI contributes equitably and safely to strengthening health systems in LMIC settings [25,69,70].

A growing concern in digital transformation in healthcare is the potential to unintentionally widen existing inequities. Algorithmic bias occurs when AI systems amplify disparities related to socioeconomic status, ethnicity, gender or sexual orientation, religion, disability, or geographic location. This poses significant challenges in many LMICs [27,65,71]. Disparities in access to technology raise important questions about equity [66]. AI models require large, high-quality, and representative datasets; however, LMIC populations are often underrepresented in these datasets. This underrepresentation leads to reduced accuracy and reliability for marginalized or low-income groups. Additionally, economic barriers, such as high costs for smart devices, internet connectivity, and data plans, create a “digital entry fee” that prevents nearly half of the LMIC populations from accessing digital health services. This situation undermines the goals of the WHO and the World Bank, which aim to advance digital transformation and achieve UHC [15]. Without a clear equity framework, digital health initiatives risk reinforcing the very disparities they seek to address.

Sustainability remains a critical challenge in digital health ecosystems in LMICs. Many initiatives in these settings are small-scale projects funded by donors that often lack clear pathways to scale up to the national level or to integrate into existing health systems. Various factors, such as reliance on external financing, high implementation costs, limited local technical expertise, corruption, and uncertain returns on investment, continue to hinder long-term viability [1,7,72]. To address these challenges, there is a fundamental shift toward sustainable and diversified financing models. Blended finance has emerged as a particularly promising approach, strategically combining public funding with private capital to mobilize private capital flows towards emerging and frontier markets, such as healthcare digitalization. By improving risk-adjusted returns, blended finance makes the digital health market more attractive to private investors and is increasingly utilized for public–private partnerships (PPPs) [73]. An example of a successful blended-finance model focused on healthcare innovation in Africa is Villgro Africa. It combines grant funding with impact investment, reducing risk for early-stage investors [74]. The open-source logistics management information system (LMIS) in Nigeria demonstrates a successful transition from donor-reliant models to private-sector stewardship through a PPP [74]. Multilateral development banks, such as the World Bank Group, the African Development Bank, and the Asian Development Bank, play a catalytic role in structuring and maximizing the use of these financial mechanisms. Additionally, the strategic alignment of Donor Investments in digital health is essential for coordinating fragmented funding around shared national priorities.

However, financial innovation alone is not sufficient. Achieving long-term sustainability requires leveraging local resources and strengthening multisectoral collaboration. Indonesia serves as a relevant example of effective multisectoral coordination in the development and national-scale implementation of digital health applications. This approach has enabled coordinated engagement across government ministries, the private sector, civil society, and development partners, demonstrating the achievements possible through coherent multisectoral governance [75].

Capacity-building is equally fundamental. Sustained investment in digital health competencies at the local level is crucial for the effective adoption, maintenance, and long-term sustainability of digital systems within local contexts. Without meaningful user engagement, adequate training, and seamless integration into clinical workflows, digital health interventions may face low uptake, user fatigue, and eventual abandonment. Overburdened healthcare providers often resist digital tools they perceive as adding to their administrative workload rather than improving service delivery. Thus, ensuring sustainability in digital health requires not only financial innovation but also institutional strengthening and workforce development to support long-term system integration [53].

Digital health has evolved from merely a collection of technological interventions to a foundational element for strengthening health systems in many LMICs. Its focus has expanded not only to improve service delivery but also to restructure the HIS, governance models, and frameworks for patient engagement.

Traditionally, many health systems in LMICs have relied on paper-based records, resulting in fragmented data flows, insufficient continuity of care, and limited evidence to inform planning and decision-making. The expansion of EHRs and interoperable digital platforms presents an opportunity to enable real-time data access, improve care coordination, and enhance system responsiveness and resilience, especially during public health emergencies [6,12,62]. For example, Angola has improved Tuberculosis (TB) surveillance through the Electronic Medical Record System (EMRS), resulting in better patient care [83]. In Rwanda, the introduction of a national HIE that uses a unique patient identifier has successfully ensured continuity of care across public and private facilities, marking a significant advancement in the country’s digital health transformation [84]. Similarly, Brazil has established a unified national health system to create a comprehensive database for epidemiological, administrative, and clinical purposes [85]. In addition to developing infrastructure, it is essential to enhance digital literacy and workforce capacity [62,72]. Building competencies in data management, digital applications, and AI-supported decision-making is critical for ensuring the effective adoption and long-term sustainability of digital transformation initiatives [6,18,72]. Lastly, policy alignment and governance are vital for supporting these advancements.

A critical future direction for digital health transformation in LMICs is to align policies and strengthen governance mechanisms. As digital health ecosystems become more complex, data-driven, and interconnected, establishing coherent regulatory and institutional frameworks becomes essential. These frameworks are crucial for protecting data privacy, ensuring ethical deployment of AI, strengthening cybersecurity, and enabling secure cross-border data exchange [27,61,70]. Global coordination platforms, such as the Global Initiative on Digital Health (GIDH), provide valuable opportunities to harmonize standards, promote interoperability, facilitate knowledge exchange, and support countries in developing context-specific digital health roadmaps [86]. Strengthening the alignment between global frameworks and national implementation strategies will be vital for ensuring that digital transformation efforts are scalable, secure, and equitable.

An important future direction for HISs is the transition to “loose coupling” architectures. Unlike rigid, monolithic platforms, loosely coupled systems enable modular design, enabling various applications to work together through shared standards and open interfaces. This approach enhances flexibility, scalability, and adaptability—qualities that are especially critical in LMICs, which often face resource constraints and rapidly changing health priorities [15]. Promoting local innovation within interoperable ecosystems allows governments, startups, academic institutions, and private-sector entities to develop digital solutions tailored to local needs. A prime example is the Rwanda Health Information Exchange (RHIE), an interoperability framework that facilitates communication and data exchange across healthcare systems. By integrating healthcare data in accordance with global standards, RHIE has improved data accuracy, accessibility, and real-time sharing of patient information, including visit summaries, diagnostic results, and treatment histories. This integration supports continuity of care and informed clinical decision-making [84].

PPPs can further accelerate innovation, expand technical expertise, and mobilize investment that aligns with national health strategies and public interest objectives [15]. Rwanda’s focus on engaging the private and academic sectors through PPPs has been crucial to its success in transforming healthcare through digital solutions. These partnerships ensure that the health workforce becomes digitally literate and is prepared to utilize the technology integrated into Rwanda’s health systems [84].

The future of digital health revolves around moving away from isolated, pilot-based initiatives and transitioning toward integrated national digital ecosystems aligned with UHC goals. A noteworthy example of this is Indonesia, which successfully implemented a national-scale digital management system in response to the COVID-19 pandemic. This system encompassed detection, prevention, treatment, and monitoring [75]. Resilient health systems will increasingly depend on interoperable, scalable, and people-centred DPI to enhance service delivery. This approach closely aligns with the core objectives of the WHO Global Strategy on Digital Health, which emphasizes global collaboration, strong governance, and systems that prioritize people, all empowered by digital technologies [6,27,61,72].