What digital infrastructure did for healthcare abroad, and what it means for Jamaica

In October 2025, Hurricane Melissa tore through western Jamaica. At Falmouth Hospital, staff spent the weeks afterward drying patients' medical records on tarpaulin sheets. Some records were declared irretrievable. Years of diagnoses, prescriptions and histories, gone with the water, because they existed in exactly one place: on paper. [1]

That image is the clearest argument for digital health infrastructure anyone will ever write. But it is not an isolated failure, and Jamaica is not an isolated case. Health systems everywhere ran on paper until, one by one, they stopped. What follows is what happened when they did, drawn from the countries that went first. Every number in this article traces to a named source at the end.

The world has already voted

The World Health Organization adopted its Global Strategy on Digital Health in 2020 and extended it through 2027. By the end of 2024, 129 member states had national digital health strategies, up from 85 a decade earlier. [2]

The economics are stark. A joint WHO and ITU analysis found that an additional investment of 24 US cents per patient per year in digital health interventions could save more than 2 million lives from noncommunicable diseases over the next decade, and avert roughly 7 million hospitalizations. [3]

Rwanda: logistics as medicine

Rwanda made the most famous bet. In 2016 it became the first country to run national medical drone delivery, moving blood to rural hospitals through Zipline. A peer reviewed study in The Lancet Global Health examined 12,733 blood orders across 20 hospitals: mean drone delivery took just under 50 minutes, 79 to 98 minutes faster than road estimates, and blood wastage from expiration fell 67 percent within a year at served facilities. [4]

A Wharton School study conducted with Rwanda's Ministry of Health, as reported by Zipline and Knowledge at Wharton, found facilities served by the network cut in hospital maternal deaths from postpartum hemorrhage by 51 percent. [5] In 2023 alone the network moved 28,754 units of blood. [6]

India and Africa: access at scale

India's Practo showed what happens to access when healthcare goes online. During 2020, the metro to non-metro split of its online consultations shifted from 75:25 to 60:40, and when it launched consultations in 15 Indian languages, two thirds of those came from users outside the big cities. [7] The platform now publishes patient reported outcomes: 78 percent of telemedicine users reported recovery within three weeks. [8]

In Kenya, the M-TIBA mobile health wallet grew from 140,000 users in 2016 to 4.7 million by 2021, connected to more than 3,000 healthcare facilities. [9] In West Africa, Helium Health reported more than 1,000 healthcare facilities and 10,000 health workers on its electronic records tools, serving over a million patients. [10]

The clinic level evidence

National numbers can feel abstract. The clinic level evidence is not.

A Cochrane systematic review of eight randomized trials found appointment attendance rose from 67.8 percent with no reminders to 78.6 percent with SMS reminders. [13] A broader synthesis across 29 studies put the reduction in no shows at roughly a third. [14] And a cluster randomized trial comparing digital records to paper in primary care clinics measured about a 30 percent reduction in time spent on health information management per visit. [15]

For a country that, per 2016 World Bank data cited by the Jamaica Observer, has 0.5 physicians per 1,000 people, that last number is the one to sit with. [17] When clinicians are scarce, every administrative minute reclaimed is clinical capacity created.

The honest lessons

The record abroad is not a parade. Two failures matter as much as the successes.

Babyl Rwanda covered 2.7 million users and its telemedicine measurably outperformed in person care on several quality measures, including around 70 percent fewer unnecessary lab tests. It still shut down in 2023, overnight, when its foreign parent company went bankrupt. [11] Clinical success does not survive unsustainable ownership.

And in late 2025, M-TIBA confirmed a breach in which roughly 4.8 million records were illegally obtained. [12] Data protection is not a compliance checkbox bolted onto health infrastructure. It is the infrastructure.

The lesson from a decade of digital health abroad: build it durably, govern the data properly, or do not bother.

Jamaica's turn

Jamaica knows all of this. The Auditor General's 2023 performance audit records that a 2014 electronic patient records pilot was never rolled out to its 27 target sites, and that hospital systems still "operate in silos." [16] The current effort is more serious: the eCare electronic health record system went live at May Pen Hospital in January 2024, the first facility nationally, and now runs in 13 public facilities under the US$148 million Health Systems Strengthening Programme. [15b] That is real progress. It is also 13 facilities out of several hundred.

The public rollout will take years. The private layer does not have to wait. Foresythe Informatics, a Jamaican healthtech company, is building the health informatics layer for the Caribbean: clinic records, patient booking, AI assisted clinical documentation. It is in the process of delivering exactly this class of change to Jamaican clinics now. [18] Every clinic that moves off paper this year is a clinic whose records survive the next hurricane, whose patients get reminder messages, and whose clinicians spend their scarce hours on medicine.

The gap is execution, not evidence

Nobody needs another pilot study. The evidence phase of digital health ended years ago, in Kigali and Kisumu and Bangalore. What remains is engineering: pipelines that move data reliably, records that are governed properly, systems that are built to outlive their vendors.

That is infrastructure work. It is what we do. If you run a clinic, a lab, or a health organization in Jamaica and you are still running on paper and spreadsheets, tell us what you are trying to solve.