The growing use of digital technologies in healthcare allows for a more comprehensive understanding of the health of individual patients and of populations and can improve the efficiency of providers. This in turn could enable more accurate diagnoses, timely treatments, and personalized care or interventions at a population level.
To achieve these positive outcomes requires a wide range of systems and tools—from electronic health records to varied management systems—to be “interoperable.” That is, these systems need to be integrated and able to share or use the same data so that they can work together seamlessly.
Our prior research, and research from others, suggests that such integration can lead to benefits: for example, accurate electronic health records, along with tools such as e-prescriptions, could result in up to a 15 percent increase in efficiency within healthcare systems by 2030 (depending on implementation, digital tools and systems selected, existing scale, the spending base, and costs of system implementation).1 Such gains would allow healthcare workers to concentrate more on high-value medical care and less on administrative tasks.
Achieving this data interoperability can be challenging, and there is no one-size-fits-all solution: country context is a critical factor, and many countries have accordingly adopted different approaches.
In this report, the McKinsey Health Institute, as part of a consortium with Exemplars in Global Health and other partners, outlines a common set of options for how countries might build an interoperable system that fits their specific context and strategic aims.
This article is a condensed summary of the full report. It focuses on key design dimensions grouped under strategic approach, architectural and technical design, and user engagement, and examines the digital health experience in three countries that have very different contexts and approaches: Canada, Estonia, and Tanzania (Exhibit 1).
These three examples underscore how the specific financing context, locally available resources, and governance models strongly influence the path toward interoperability. Each country exhibits unique relevant contextual factors:
- Canada has a large and decentralized public healthcare system, in which provinces provide healthcare services, including the establishment of their own health interoperability architectures. This system is guided by Canada Health Infoway, a centralized, not-for-profit entity.
- Estonia has a centralized government-owned structure that uses the national digital infrastructure.
- Tanzania’s system, which has included support from donors, has been developed with a focus on use cases integrated at various levels of care, with plans to eventually incorporate all use cases into the Health Information Mediator.2
These countries’ varying experiences could provide lessons and options for others, particularly low- and middle-income countries currently looking to build and develop their own healthcare interoperability systems.
Digital interoperability in healthcare
Simply put, interoperability allows different organizational and digital systems to interact—or digitally communicate—through the sharing of data.3 The benefits of data interoperability play out in a variety of ways for different stakeholders that interact in the healthcare system:
- Patients and communities could benefit from more coordinated and integrated care and potential empowerment through access to their own digital health records.
- Clinicians might benefit from greater efficiency of data collection and better overall coordination with healthcare professionals and organizations.
- Healthcare organizations (public and private) may heighten efficiency by eliminating duplicative testing and repeat visits.
- Health authorities may find it easier to access high-quality aggregate data to formulate public policy and more cost-effective methods of ensuring compliance.
- Researchers might increase their access to big data and improve the identification and tracking of new treatments or interventions.
In Canada, for example, early estimates from 2018 suggested that full adoption of interoperability could result in annual healthcare savings of approximately CA $4 billion (approximately US $3 billion),4 or 1.6 percent of the country’s total healthcare expenditure.5 These savings would stem primarily from reduced hospital admissions and a decrease in redundant laboratory and imaging tests.
In Estonia, research estimates that digital solutions powered by interoperability could lead to substantial process improvements such as reduced time in repeat prescriptions and reduced time in waiting for emergency services.6 While Tanzania is still in the early stages, stakeholders express enthusiasm about the enhanced healthcare outcomes observed in the use cases implemented. For instance, an evaluation of one specialized hospital data-sharing case revealed substantial time savings for clinicians when data was automatically processed rather than summarized manually.7
Given such potential benefits, many countries have started taking steps toward building health information exchanges (HIEs) or similar digital health architectural frameworks that serve as the basis of interoperability. In 2023, 57 percent of countries participating in the Global Digital Health Monitor8—which tracks and evaluates the use of digital technology for healthcare in 67 countries, including 60 low- and middle-income economies—were building or beginning to implement high-level national digital health architectural frameworks. Additionally, 55 percent of countries have set and published data standards.9
Building a basic interoperability platform is not enough: realizing the full benefits will require countries to adopt a robust approach to implementation. As of 2023, only 13 percent of countries had a complex data architecture in place, while just 18 percent broadly adhered to established data standards that are routinely updated.10
Design choices for building interoperability
Countries face a common set of choices when making decisions about the development of interoperable systems. Here, we identify and describe seven dimensions, categorized under strategic approach, architectural and technical design, and user engagement (Exhibit 2). Countries often choose hybrid approaches in their design strategies, and there might be additional dimensions relevant in other contexts.11 The three focus countries have made varied choices, which we describe in detail in the full report.
Strategic approach
1. Governance and operating model
Governance structure. Key choices here include the degree of centralization, which may depend in part on the health system itself. A national structure led by a single entity may offer streamlined decision making across the system. Decentralized systems often have subnational governance with national guidance on key topics such as standards and suggestions for HIE architecture that can help guide decision making and investment. For example, Estonia’s TEHIK (Health and Welfare Information Systems Centre) is a single centralized government-owned company that manages the Estonian Health Information System (EHIS), which supported the rapid uptake of interoperability.12 Similarly, Canada Health Infoway offers general guidance on interoperability to provinces, such as suggested system architecture, while the specific design and implementation of interoperability systems remain under provincial jurisdiction.13
Leadership engagement. Engagement at the highest levels of government is crucial for driving decision making and aligning stakeholders, especially in the early stages of implementing interoperability systems. Ministerial-level engagement supports ongoing implementation efforts. Effective multistakeholder engagement involving government agencies, providers, payers, health tech service providers, digital and health lenders, donors, the medical sector, and others is essential but complex to orchestrate. Early engagement and frequent coordination are key to successful implementation and long-term sustainability of interoperability systems. Using each actor’s strengths—such as political capital, funding access, and sector-specific knowledge—can help establish use cases and foundational interoperability layers.
Among the three focus countries, the Tanzanian government has coordinated closely with the Development Partners Group—a donors’ group composed of 17 donor countries and five multilateral agencies that has been in operation since 1961.14 Canada’s e-pharmacy program, PrescribeIT, gained momentum only after the pharmaceutical sector became involved and championed the program.
2. Timing and phasing and 3. Financing
In countries with highly centralized decision-making power, more availability of financial resources, or both, countries may start their interoperability system journeys by building foundational elements such as the interoperability layer or spine.15 Estonia’s e-health system has evolved over time: planning to set up the national health information system began in 2000, and the system launched at the end of 2008 along with a first set of projects including electronic care summaries, digital imaging, and e-prescriptions.16 The systems were subsequently connected to the preexisting digital government interoperability exchange platform. Adopting new systems can be time-consuming. In Estonia, for example, widespread use of the EHIS by patients and clinicians only gained momentum in 2014, about five years after its go-live date and following substantial investments (Exhibit 3).
In countries with less centralized decision-making power, fewer available financial resources, or both, it may be necessary to build support for the system and secure funding by first building use cases with defined parameters and established standards. The interests of civil society and providers may influence the choice of use cases. In Tanzania, the health interoperability system was implemented in phases—for example, with certain use cases such as improving access to and use of data across specialized hospitals and children’s vaccination data in the health information system and levels of care—and digital solutions have been gradually connected to the Health Information Mediator (HIM).17
In countries where there is a subnational, decentralized governance structure, jurisdictions could accelerate and shape their interoperability health systems according to their needs and available funding while national entities provide guidance on architecture and data standards. In Canada, for example, there are marked differences among interoperability systems based on province priorities, existing digital infrastructure, and access to funding and human capital. For example, some provinces, such as Alberta (MyHealth.Alberta.ca), and Quebec (Carnet santé Quebec), have more advanced patient portals than others.
Architecture and technical design
4. Architecture, data, and infrastructure
Data standards and platform. Countries have the flexibility to implement either international or national data standards and can opt for open-source or proprietary platforms. A common approach is to adopt international data standards—which could be customized to meet specific national needs—alongside open-source platforms. Given the ease and flexibility in their use, these choices may be more cost-effective and encourage local capacity building and innovation over time. The best way to proceed may depend on the extent of legacy systems, local capacity to innovate, and stakeholder preferences.
Standards implementation. Often countries aim to adopt one standard, or a set of standards, to improve scalability. However, this can be difficult when legacy systems are in place. In such cases, the country might provide a time-capped allowance for providers and other organizations to adopt new standards and implement a central transformation system—a terminology server that transforms different data standards (often legacy standards) into unified standards to facilitate data exchange.
Data storage and security. In more centralized countries, central storage with a central authority that verifies identity may offer a streamlined approach for meeting countries’ data security expectations. Decentralized countries might still decide to have central storage at the jurisdiction level.
Hosting. Hosting on premises may better align with government data sovereignty priorities and data management protocols, while also operating independently of connectivity issues. Conversely, cloud hosting could offer potential cost advantages, depending on implementation.
For example, in Tanzania, the HIM used international standards, but they were customized for the country’s needs. Most of the system (including HIM) has been built on open-source platforms, and hosting is done on premises to comply with legal mandates of data management. In Estonia, the EHIS stores data centrally on premises to streamline data management as part of its core functions.
5. Identification mechanism
Countries with more mature digital ecosystems might use existing national digital IDs as an identifier in the healthcare systems, while other countries might use a combination of other existing IDs to build a robust health identification system. Countries might also migrate to a national digital ID over time. For example, Estonia provides citizens with a state-issued digital identity through e-ID, a comprehensive digital identity program for citizens. Tanzania aims to have a single ID number (Tanzania Universal Identity Number Jamii Namba) for its citizens by the end of 2025.18 (In the meantime, Tanzania uses multiple legacy identifiers to ID citizens.) For countries with substantial migrant populations, ensuring access to an identification mechanism compatible with the healthcare system could enhance access to integrated, quality healthcare for these groups.
User engagement
6. Adoption
A combination of mandates, where feasible, and both financial and nonfinancial incentives may drive adoption of the system across users including providers and vendors. In Tanzania, health facilities with sufficient connectivity in the public sector are mandated to capture data and share it with the system, depending on the level of care. For example, facilities at the district level with sufficient connectivity are mandated to share their captured data with the appropriate repository for district and primary care facilities.
A dual approach of providing data rights for patients while building trust and incorporating user feedback may support broad adoption. Estonia has a consent-based model under which patients can choose whether to disclose their data and monitor visits to their HIS.19
7. Capability building
Up-front investment in directly building capabilities within government and across providers may support initial adoption and provide a baseline of capabilities to better enable ongoing lighter-touch or train-the-trainer approaches. At Ontario Health in Canada, tech-savvy clinicians who are involved in crafting the implementation plan of digital solutions that rely on interoperability also support onboarding of their peers in clinical settings, according to experts we interviewed for this research. Estonia developed a train-the-trainer approach with introductory meetings and seminars for developers and vendors to help clinicians understand and implement e-health solutions.
Choosing a path forward
Establishing and maintaining a successful health data interoperability system can take various forms depending on national context and specific digital health ecosystems. Context challenges include insufficient connectivity, shortages of skilled personnel, a reliance on donor funding, and challenging healthcare governance structures. Challenges for digital health ecosystems include low adoption by providers and vendors and insufficient adoption of standards. The most effective strategy to address challenges will vary according to each country’s context and capabilities and the availability of funding. While there is no single path forward for all, multiple paths can lead to similar, positive outcomes from greater interoperability.
