Read Also

When Murang’a county in central Kenya launched its telemedicine programme in September 2024, the announcement had the texture of a public health success story. More than three dozen health facilities connected to the internet. Specialist care accessible to rural residents who had previously spent half a day travelling to the county’s main hospital. A consultation fee of less than a dollar. Within a year, Governor Irungu Kang’ata could point to 25,000 patients reached and queues, officials said, that had visibly shortened.

None of that is false. But behind the appealing headline sits a question that no public document, procurement record, or official statement has yet answered: who is actually paying for it?

That question, deceptively simple and persistently unanswered, sits at the heart of why telemedicine in Africa keeps producing promising pilots and very few durable programmes. The continent has now accumulated a decade’s worth of launches, a library of congratulatory press releases and one rigorously documented case study that ought to change how every government, development finance institution and digital health entrepreneur thinks about what sustainability actually requires.

The continent’s best experiment

That case study is Rwanda. Babyl Rwanda, the local operation of UK-listed Babylon Health, launched in 2016 with an architecture specifically designed for low-resource settings: consultations accessible via basic feature phones using USSD and SMS, integrated with Rwanda’s national health insurance scheme and backed by a formal partnership with the government. By 2023 it had enrolled 2.5 million users, roughly 18 per cent of the national population, and processed 3.9 million consultations across 450 of the country’s 510 primary health facilities.

Rwanda had, by most credible measures, achieved what every digital health optimist imagined was possible: scale, insurance integration and measurable clinical improvement. Then, in August 2023, Babylon Health filed for Chapter 7 bankruptcy in the United States, having recorded a net loss of $221 million on $1.1 billion in global revenue the prior year. Within weeks, Babyl’s Rwanda operations were wound down. Nearly a fifth of Rwanda’s population lost access to digital health services because of decisions made in a London boardroom.

The aftershock was quantifiable. A subsequent interrupted time series study published in BMC Primary Care documented an immediate surge of 809 additional monthly respiratory infection cases presenting at physical health facilities following Babyl’s closure, with malaria adding 256 cases per month. Patients who had built a care habit around digital access reverted to facilities that had never been resourced to absorb them.

The architecture problem

Rwanda’s experience strips away the comfortable narrative that telemedicine in Africa fails because of connectivity gaps or digital literacy. Babyl solved the infrastructure problem. It failed because sovereign healthcare access for nearly a fifth of a population had been outsourced to a company’s ability to sustain losses on a global growth strategy conceived in a Swedish digital health company’s boardroom and financed by London capital markets.

This is not an isolated failure of corporate governance. It is the predictable consequence of a structural design choice repeated across the continent: governments invite private digital health operators to deliver public health functions, provide minimal or opaque public financing, and assume that the operators’ commercial incentives will sustain what public budgets cannot.

Return to Murang’a. Patients pay less than a dollar to the local facility. BYON8, a Swedish digital health company whose core commercial product is a subscription service priced at six to 20 times that per month, provides the platform. A local firm, HealthX, manages on-the-ground implementation. The county government provides political will and internet connectivity.

What no public document discloses is what BYON8 and HealthX are paid, by whom, or whether any contracted service fee exists at all. The programme may be running on partner cross-subsidy, on strategic goodwill, or on a budget line too small to appear in published county allocations. None of these are illegitimate ways to launch a programme. All of them are illegitimate ways to sustain one.

What transparency would actually require

The Murang’a programme has demonstrated something real: implementation feasibility, patient acceptance and a functional hub-and-spoke model with Muriranjas Subcounty Hospital as the central node. These are not trivial achievements. But 25,000 patient contacts is an output metric.

It says nothing about whether NCD control has improved, whether catastrophic health expenditure has been reduced, whether the people reached were the rural poor or the relatively better-connected within each catchment area, or whether the clinicians providing remote consultations are being compensated at a rate that will keep them at their screens in year three.

A genuinely transparent architecture would require the county government to publish its contracted obligations to private platform partners, including per-consultation or platform licensing fees. It would require a disclosed unit cost that includes the full supply-side expenditure, not only the patient-facing tariff.

It would require outcome metrics tied to clinical endpoints rather than consultation volumes. And it would require a formal sustainability plan that does not depend on a single political administration’s continued enthusiasm, a point worth noting when any program’s primary advocate faces an election cycle.

Rwanda, to its credit, is now redesigning its national digital health strategy using the Babyl experience as its primary evidence base. That is what institutional learning looks like. Ghana has published policy frameworks with similar intent. Most of the continent is still repeating the pilot.

The lesson Africa cannot afford to learn twice

The problem with digital health in Africa is not the technology, and it is not even primarily the infrastructure. It is that the financing architecture of nearly every scaled programme has been designed to impress funders and deliver political announcements, not to survive the departure of its champion, the bankruptcy of its vendor, or the exhaustion of a grant cycle.

Telemedicine works. Rwanda proved it works clinically. What has not been demonstrated, anywhere on the continent at scale, is a model where the public health function is financed with the same seriousness as the technology deployment.

Where the government is not merely the political host of someone else’s platform but the accountable, adequately-resourced guarantor of continuity. Where a corporate restructuring in a distant capital cannot, overnight, erase the care access of millions of people who had built their health behaviour around a promise.

Until health ministries and their development finance partners treat sustainability architecture with the same rigour they apply to the app interface, the pattern will hold: launch, scale, celebrate, collapse. The patients who restructured their health-seeking behaviour around the promise will absorb the consequences at physical facilities that were never built to receive them.

The most dangerous sentence in African digital health is not a frank acknowledgement of failure. It is “we reached 25,000 patients”, offered as a conclusion, when it is only a question.

Surgeon, writer and advocate of healthcare reform and leadership in Africa