How AI is transforming procurement at the point of care

After two decades of PATH investment in digital health, AI can now move procurement intelligence beyond national dashboards and into the hands of frontline health workers.

Tanzania’s Better Immunization Data (BID) Initiative introduced electronic records to increase immunisation coverage. Photo: PATH/Trevor Snapp

A patient should never miss treatment because the health system sent them to the wrong place. Yet this happens too often across low- and middle-income countries (LMICs). A person takes time away from work, pays for transport, waits in line, and reaches a facility, only to learn that the medicine, diagnostic test, or essential supply they need is unavailable.

Sometimes it is unavailable everywhere. But sometimes it is available at another facility or another pharmacy. The patient does not know that. The health worker may not know that either. And so, the patient is asked to come back, referred blindly, or quietly lost to follow-up.

This is not simply a procurement failure. It is a failure of information reaching the person who needs it when it matters most.

Investment in digital health

For more than 20 years, PATH has invested in digital health systems across Africa and other low- and middle-income regions. Much of that work has been technical and invisible: electronic registries, logistics systems, data governance structures, interoperability standards, peer learning networks, and the steady work of helping people use data in the flow of service delivery. These investments laid the foundation for the next generation of health system intelligence.

Tanzania offers a useful lesson. The Better Immunization Data (BID) Initiative, led by Tanzania’s Ministry of Health in partnership with PATH and funded by the Gates Foundation, was grounded in a simple premise: better data plus better decisions lead to better health outcomes. Beginning in 2013, BID supported countries in building their own systems to improve data quality and routinely use it to inform decisions. This included a standards-based electronic immunisation registry connected to the vaccine information management system, linking service-delivery data with stock and supply chain information.

That linkage matters because visibility is power. Before BID, health workers often had poor visibility into vaccine supplies and limited capacity to use data for routine decisions. The electronic registry helped prevent stockouts, reduce wastage, and support better planning for vaccine distribution.

For me, the real lesson is that digital tools alone will not solve all health system problems. Digital transformation works when technology is matched with governance, workflow redesign, user support, and a culture of decision-making. Data must be reliable, accessible, and actionable. And importantly, people must have the confidence, skills, and authority to use it.

AI and procurement reform

This is especially relevant to AI. Too often, health data are collected at the point of care, aggregated for reporting, and used far away from the people who generated them. The information flows to the district, the national programme, the donor, or the dashboard. But it does not always return to the nurse, pharmacist, records officer, or facility manager in a form they can act on.

This is the point at which AI becomes more than a technology story.

The opportunity is not simply to improve national dashboards, though that matters. Ministries should be able to ask their data practical questions: which facilities are at risk of stockout next month? Which districts are over-ordering? Which products are vulnerable to financing delays?

But the more important opportunity may be at the last mile. Imagine a health worker who can ask in real time: where should I send this patient for insulin today? Which nearby facility has the medicine in stock? Is this stockout likely to be resolved this week?

This is the natural next step when procurement, logistics, electronic medical records, and routine health information systems are connected, governed, and designed around service delivery.

For patients managing chronic conditions such as diabetes, hypertension, cancer, or HIV, this matters enormously. Care is not a single event; it is a chain of interactions over time. Every broken link increases the chance that someone drops out. Every unnecessary referral or missing medicine imposes real costs: transport, lost wages, childcare, fear, and fatigue.

Procurement reform is not only about price, tendering, warehousing, or supplier performance. It is about whether the system can make good on the promise of care. Accountability should extend to whether the right patient could access the right product at the right time and place.

Shared intelligence and regional capacity

AI will not solve this by itself. In fact, poorly governed AI could deepen existing failures. If data is incomplete, recommendations will be unreliable. If tools are designed far from frontline realities, they become another burden. If countries do not own the systems and rules, AI may reproduce dependency in a new form.

There is an African proverb, “Wisdom is like a baobab tree; no one individual can embrace it alone.” That is true of procurement reform, and equally true of AI. No single ministry, donor, technologist, or implementing partner can solve the last-mile access problem alone. It requires shared intelligence, local ownership, and people close enough to the problem to understand what the data cannot say on its own.

That is why regional capacity matters. As PATH reimagines how it works, the goal is not to create another distant technical unit issuing guidance from afar. It is to build regional centres of excellence for data, digital, and AI that brings expertise closer to ministries, implementers, health workers, and the communities whose lives depend on these systems.

The next era of procurement will be defined not only by smarter forecasting or more efficient supply chains. It will be defined by whether intelligence reaches the last mile in time to change a decision.

And sometimes, the most important decision is simple: do not send the patient to the wrong place.

---

The opinions expressed are those of the author and do not necessarily reflect the position of Re:solve Global Health.

Dr Samuel Wambugu is a global health informatics and digital transformation leader with more than 20 years of experience supporting health information systems, data governance, digital health strategy, and AI readiness across LMICs. At PATH, he leads digital health and informatics work focused on practical, country-owned solutions.