This article draws on “Bridging the Last Mile: Scaling AI-Driven Supply Chains and Connected Diagnostics for Equitable Primary Health Care,” a VillageReach-led session at the second Africa Primary Health Care Forum in Abuja, Nigeria, moderated by VillageReach’s Dapo Awobeku. Watch the full conversation on YouTube. Quotations have been lightly edited for clarity.
Digital health has acquired a familiar reflex: When a system struggles, commission another tool.
The result can resemble a digital archipelago. One platform records vaccinations. Another monitors stock. A third collects diagnostic results. Partner dashboards multiply, each with its own login, reporting cycle and training package. Somewhere inside this arrangement, a health worker is expected to keep every island supplied with data.
Many of these tools address real needs, but their cumulative weight creates a different problem.
Health workers may enter similar information several times while caring for patients, managing commodities and completing routine reports. Health facility staff receive streams of data that rarely connect. Devices arrive through projects with separate plans for connectivity, maintenance and replacement. A promising pilot performs well until its funding, technical support or staff incentives reach an end date.
Now artificial intelligence is entering this already fragmented environment. Its ability to find patterns, support forecasts and improve decisions could be profound. Yet AI draws its power from the systems around it. Fragmented information produces a partial picture, while unreliable workflows weaken the data used for analysis.
As Omofome Ikogho of Development Delivery Partners (DDP) observed during the APHCF session:
“You cannot automate a broken system.”
That sentence deserves a place at every digital health procurement meeting.
For VillageReach, this captures one of the defining challenges of the next phase of digital health: not how quickly new technologies can be adopted, but whether they strengthen the systems around them. Do they fit frontline workflows? Connect with existing platforms? Produce information people can act on? Remain viable after external support ends? As AI creates new possibilities, those fundamentals become more, not less, important.
Design around the work
Dr. Muknaan David Nshe of the Christian Health Association of Nigeria (CHAN) started with a practical test for digital innovation: Does the system make a health worker’s day easier?
CHAN supports hundreds of faith-based health facilities, many serving rural and peri-urban communities with limited staff. Its work with an inventory management platform brings pharmacy functions into a common system. Health workers can record stock movements, manage commodity information, support financial processes and generate monthly requisitions from data entered during routine work. The platform gives health workers more time for patient care by integrating several administrative tasks into an existing workflow.
“Design systems around the workflow of the health worker, and not around technology,” Nshe said.
The lesson is straightforward: Start with the work, not the technology. Developers need to understand how a clinic opens, how commodities are counted, when records are updated, who approves a request and where delays usually occur. A digital tool earns its place through daily usefulness.
Adoption becomes more durable when health workers can see what the system does for them. A dashboard viewed only at the national level may feel remote to the person feeding it data. Information that helps a facility identify low stock, prepare a requisition or follow a child’s immunization history offers an immediate return.
Invest in the machinery behind the screen
Digital health attracts attention through software. Much of its success depends on electricity, connectivity, functioning devices and practical technical support.
“It is more glamorous to come up with a new app or a new solution,” said Muhsin Mufti of eHealth Africa. He pointed instead to the behind-the-scenes infrastructure that allows health workers to use these systems consistently.
Even the most sophisticated application has limited value if a facility cannot charge its tablet, afford connectivity or access ongoing technical support.
These conditions call for shared foundations. Mufti drew on humanitarian response models in which a dedicated provider establishes connectivity that several organizations can use. Similar arrangements could serve health facilities through common power, internet and device services. Partners could build on shared infrastructure rather than reconstructing it project by project.
Such an approach requires coordination that extends beyond the health sector. It also asks funders to value the pipes and wiring of digital health as much as the application visible on the screen.
Give existing systems room to work
Akolade Jimoh of Sydani Group described using the World Health Organization’s Strategic Toolkit for Assessing Risks to help one state anticipate public health threats. The exercise supported preparation for hazards that later occurred, including flooding and a cholera outbreak.
The example points to an often-overlooked form of innovation: getting more value from systems that already exist. Investment in integration, data quality, user support and analytical capacity may unlock more value than another standalone platform.
VillageReach is seeing the same principle in Nigeria’s nationwide rollout of Last Mile Vaccine Stock Visibility. Rather than introduce another standalone platform, VillageReach and DDP, whose Omofome Ikogho joined the APHCF panel, are supporting the National Primary Health Care Development Agency to strengthen an existing DHIS2-based reporting tool, adapt it around frontline workflows and prepare it to connect with other existing systems tracking vaccines, cold-chain equipment and temperature data. The aim is a more connected flow of information that helps government teams identify and respond to supply risks.
Sydani also used digital mapping to identify settlements omitted from previous immunization plans. A resource-tracking tool showed where partner activity was concentrated and where communities received little support. The technology helped reveal the combined footprint of many programs, giving planners a clearer basis for allocating resources.
Interoperability makes this kind of intelligence possible. Technical standards matter, alongside agreements about ownership, access and responsibility. Someone must have the authority and resources to act when connected data reveal a stock risk, an underserved settlement or duplicated investment.
As LMVSV reporting becomes more consistent, the same data can support predictive analytics. Patterns in consumption and recurring shortages could help teams identify emerging risks earlier — an example of more advanced analytics becoming useful because the underlying system has first been made stronger.
Build for public ownership
Digital health projects often treat government transition as a final stage. Omofome Ikogho of DDP described a different model:
“When we talk about handing over, it shouldn’t just be as an exit strategy.”
Government ownership begins while the system is being conceived. Policies and oversight arrangements shape its future. Procurement rules affect whether public agencies can continue paying for services or equipment. Recurring costs need a credible home within domestic budgets. Government teams require meaningful control over configuration, data governance and future development.
Existing structures can anchor this work. State logistics working groups, for example, can guide implementation and connect digital tools to routine decisions. A platform becomes part of the health system when the people responsible for that system can direct its use and sustain its operation.
Make complexity earn its place
The next generation of digital health will include predictive supply chains, connected diagnostics and AI-supported decisions. Each advance will place greater demands on the quality of data and the coherence of the surrounding system.
Progress may therefore look less like a procession of new applications and more like careful consolidation. Governments can define a smaller set of trusted systems, funders can support common infrastructure and national and regional architecture, and technology partners can improve interoperability while building around the realities of frontline care.
For VillageReach, that is the opportunity presented by this next phase of digital health: not simply to accelerate the creation of new technologies, but to become more disciplined about where technology adds value.
The next breakthrough in digital health may not be another tool. It may be the discipline to use fewer, better-connected ones. Doing so would create space for technology to fulfil its central purpose: helping health workers and decision-makers respond sooner, allocate resources wisely and bring reliable care closer to every community.