Dr Tendai Pfidze

16 August 2026

Digital transformation remains one of the most promising and difficult goals for healthcare systems across Africa. The potential benefits are substantial: better continuity of care, improved transparency, stronger linkage between levels of care, reduced administrative burden, better use of scarce resources, and the ability to turn routinely collected health data into actionable intelligence.

The promise is therefore not the problem. The challenge lies in how transformation is approached.

Too often, digital transformation is treated primarily as a technology acquisition exercise: procure an electronic medical record, deploy a national health information system, automate a form, build a dashboard, or introduce another application. Yet digital transformation is much broader than replacing paper with screens. Wurm et al. (2025) describe digital transformation as a fundamental organisational change process in which technology affects value creation, organisational structures and the way an organisation operates. A digital transformation strategy therefore needs to provide direction and coordinate the many interconnected changes involved, rather than merely define which technologies should be purchased.

For African healthcare systems, I believe two problems deserve particular attention.

First, we often attempt to digitise systems before sufficiently improving the underlying operating system of healthcare itself.

Second, the digital solutions that are implemented are frequently designed without a sufficiently holistic architectural view of the health system.

Digitisation cannot substitute for operational improvement

There is a temptation to assume that inefficiency is primarily a technology problem. It is not.

Healthcare organisations can have duplicated approvals, unnecessary patient movements, poorly designed referral processes, unclear responsibilities, inefficient queues, repeated data capture, weak communication channels and poorly defined performance measures long before a computer system is introduced.

Digitising such a process does not automatically improve it.

Sometimes it simply allows an inefficient process to operate electronically.

This is where established improvement disciplines remain surprisingly important. Plan-Do-Check-Act cycles, clinical audit, Lean thinking, Six Sigma, root-cause analysis, process mapping and other quality-improvement methods can identify waste, variation and bottlenecks without requiring significant capital investment.

This matters particularly in resource-constrained health systems. Before asking, “What software do we need?”, organisations should frequently be asking:

Digitalisation should then amplify an improved operating model rather than become a substitute for developing one.

Interestingly, the wider digital-transformation literature provides support for this relationship. Wurm et al. (2025) report that one of the organisations in their analysis benefited substantially during its digital transformation from experience gained through an earlier Lean transformation. They consequently argue that lessons from Lean and other forms of organisational transformation can inform digital transformation.

The implication for healthcare is important: process improvement, organisational learning and digital transformation should not be treated as separate programmes.

They should reinforce one another.

Technology cannot compensate for weak change management

Healthcare is also intensely human.

A technically excellent platform can fail if doctors see it as additional clerical work, nurses have to document the same information twice, managers continue demanding parallel paper reports, or staff do not understand why a new process has been introduced.

Technology changes workflows, responsibilities, information access and sometimes professional power relationships. That means digital transformation is inevitably also change management.

This is particularly important because healthcare applications sit inside complex socio-technical environments involving clinicians, administrators, patients, laboratories, pharmacies, insurers, policymakers, regulators and many other stakeholders.

The South African healthcare platform ecosystem examined by Mbanefo and Grobbelaar (2025), for example, illustrates healthcare as a network of interacting actors rather than a single organisation delivering value independently. Their study emphasises collaboration between healthcare providers, patients, insurers, policymakers, technology providers and other actors in creating healthcare value.

Successful digital transformation therefore requires more than system implementation. It requires stakeholder engagement, incentives, communication, training, governance and continuous adaptation.

The second problem: architecture without a view of the whole

The other major challenge is fragmentation.

A laboratory may have one system. Pharmacy may have another. Patient administration may operate separately. Disease programmes may introduce their own databases. National reporting may depend on yet another platform.

Each system may solve a legitimate problem.

Collectively, however, they can create a landscape in which information cannot follow the patient.

The opposite problem can occur as well: rather than excessive fragmentation, organisations may create excessive coupling by trying to force very different concerns into one enormous national system.

This is where enterprise architecture becomes valuable.

Enterprise architecture asks us to examine the organisation from several connected perspectives i.e. business, information, applications and technology, rather than beginning with an application. Mbanefo and Grobbelaar (2025) similarly describe enterprise architecture in terms of business, information, application and technology architectures, with business architecture establishing goals, processes and strategies while information and application architectures determine how data and applications support those goals.

Pingilili et al. (2025) make a similar argument: enterprise architecture can align technology with organisational objectives, simplify complexity and clarify the relationship between business and technology. Their review also highlights the importance of integrating information management with architecture so that technology investments improve decision-making, resource allocation and operational performance rather than existing as isolated IT projects.

This changes the digital-transformation conversation.

Instead of asking:

“Which national healthcare application should we deploy?”

we should first ask:

“What capabilities must the national health system possess, what information must flow between those capabilities, and which systems are best placed to support each need?”

The ministry and the clinician do not necessarily need the same system

Consider one common architectural tension.

A Ministry of Health requires information for population-level planning. It needs disease incidence, mortality trends, bed utilisation, medicine consumption, staffing information, programme performance, geographic variation and other aggregate indicators.

A clinician sitting with a patient has a fundamentally different problem.

The clinician wants to know:

• What happened during the patient's previous encounter?
• What medicines are they taking?
• What allergies do they have?
• What were their previous laboratory results?
• What investigations have already been performed?
• What needs to happen next?

The ministry is primarily interested in population-level intelligence and system oversight.

The clinician is primarily interested in individual patient care and an efficient clinical workflow.

There is overlap in the information required, but that does not mean the same application must optimise both experiences.

Attempting to satisfy both concerns through one tightly coupled monolithic national application can create conflicting requirements. Features needed for national reporting can increase friction for clinicians, while flexibility required by clinical environments can complicate national standardisation.

The better architectural principle may be:

A different architecture is possible

Imagine instead a national health information architecture built around a shared information layer.

Healthcare facilities could use applications appropriate to their clinical environments. A district hospital, specialist centre, primary-care facility or private clinic would not necessarily require identical user interfaces or workflows.

What would be standardised is their ability to exchange agreed health information.

These applications could communicate through clearly governed APIs and interoperability services into a national health data platform or data reservoir.

That platform could provide capabilities such as:

The ministry could therefore obtain the information required for stewardship, planning and policy without forcing every clinician to interact directly with a system designed primarily around ministry reporting.

At the same time, clinicians could use applications designed around clinical work while still contributing to a larger national information ecosystem.

This is closer to a platform architecture than a single national application.

The South African Discovery case studied by Mbanefo and Grobbelaar (2025) provides an interesting demonstration of this broader ecosystem principle. As the platform matured, integration, technology compatibility and interoperability became increasingly important, with multiple actors and services integrating around shared platform capabilities. The authors further emphasise that multi-stakeholder healthcare platforms require both data governance and interoperability standards to coordinate care effectively.

The exact architecture required by a public national health system would obviously differ from a private healthcare ecosystem. The architectural lesson, however, is valuable: integration does not require every participant to become the same system.

Decouple where independence creates value; integrate where information must flow

This distinction is fundamental.

Fragmentation is harmful when information that needs to move cannot move.

Decoupling is beneficial when different capabilities need the freedom to evolve independently.

Those concepts are not contradictory.

A pathology laboratory system should be able to specialise in laboratory workflows.

A pharmacy system should be able to optimise medication management.

A clinical application should focus on clinical decision-making.

A national analytics platform should optimise large-scale aggregation, reporting and analysis.

The architectural challenge is ensuring that they share information through deliberately designed interfaces and common information definitions.

This is why interoperability is not simply a technical integration problem. It begins with understanding which stakeholders need which information, for what purpose, and at which point in the healthcare value stream.

Only after answering those questions should APIs, databases and integration technologies be designed.

From digital projects to digital capability

The deeper change required in African healthcare may therefore be conceptual.

We need to stop thinking about digital transformation as a sequence of IT projects and start thinking about the development of digital health capabilities.

That includes the capability to manage patient information across organisational boundaries.

The capability to continuously improve healthcare processes.

The capability to integrate systems.

The capability to govern data.

The capability to analyse population health information.

The capability to introduce technology without disrupting clinical care.

And perhaps most importantly, the capability to continually adapt the system as clinical needs, technologies and populations change.

This aligns with the broader enterprise-architecture literature. Pingilili et al. (2025) argue that architecture should align business processes, information systems and technology infrastructure rather than allowing technology to evolve independently from organisational objectives.

Technology should amplify a well-designed health system

Africa unquestionably needs greater healthcare digitalisation.

But digitalisation itself should not become the objective.

Better healthcare is the objective.

Better continuity of care is the objective.

Lower waiting times are the objective.

Better clinical decisions are the objective.

Reduced waste is the objective.

Earlier identification of disease trends is the objective.

More intelligent allocation of scarce resources is the objective.

Technology is one of the most powerful tools available for achieving these outcomes—but only when it is applied to a system that has been deliberately understood and designed.

The sequence therefore matters.

Perhaps the biggest opportunity for African healthcare digital transformation is therefore not simply to digitise faster.

It is to architect better systems before and while we digitise them.

References