The ability to guide each member toward the right action at the right time
In South Africa, Discovery Health's largest open medical scheme has turned artificial intelligence into a quiet instrument of behavioral change — not through mandate, but through the ancient art of the well-timed, well-placed reminder. Since early 2025, over 650,000 members receiving personalized health prompts are three times more likely to undergo health checks and twice as likely to pursue cancer screenings, with nearly a quarter of those checks uncovering previously hidden conditions. The experiment raises a deeper question humanity has long wrestled with: not whether people know they should care for themselves, but whether they can be gently, precisely reminded to do so before it is too late.
- Hundreds of thousands of people were quietly drifting away from preventive care — skipping screenings, avoiding doctors — until an AI began asking them, personally and specifically, to show up for themselves.
- Nearly one in four health checks triggered by the programme returned abnormal results, revealing silent conditions like hypertension and high cholesterol that would otherwise have compounded unseen.
- The scheme is racing to scale precision nudging across 2.1 million adult members, betting that specificity — telling someone exactly what their next health step should be — outperforms generic public health messaging.
- A decade of data shows mortality down 5.6% and cancer survival up 48%, but a surging 80% rise in mental health conditions among young people signals that the system's next frontier may be its most complex.
- The model's promise is real, but its reach remains bounded — a large, relatively affluent membership base that may not reflect the populations where behavioral health nudging is needed most.
Discovery Health Medical Scheme, South Africa's largest open medical scheme, has been running a quiet but consequential experiment in behavioral nudging since January 2025. More than 650,000 of its adult members have enrolled in a programme called personal health pathways, which uses artificial intelligence to send each person a tailored prompt toward their single most relevant next health action — a colonoscopy, a dental visit, a blood pressure check. The results have been striking: enrolled members are three times more likely to complete health checks and twice as likely to pursue cancer screenings than those outside the programme.
The specificity of the prompts appears to be the key. Rather than broadcasting generic wellness advice, the system identifies what it calls the "next best thing" for each individual. When members act on those prompts, the discoveries are often sobering — nearly 25% of health checks returned out-of-range results for conditions like high cholesterol and hypertension. For many, these were problems that had gone undetected and untreated.
The profile of who is enrolling adds texture to the story. Two-thirds already live with a chronic condition, but 38% had previously shown poor health-seeking behavior — people who avoided doctors, skipped flu shots, and neglected routine care. The programme appears to have shifted those habits meaningfully, with chronic disease patients showing improved adherence to clinical protocols and treatment plans.
Over a longer horizon, Discovery Health reports a 5.6% decline in member mortality over the past decade and a 48% improvement in cancer survival rates, with members now living an average of 7.1 years longer after a cancer diagnosis than they did ten years ago. New treatments have contributed, but so has earlier detection.
Not all the data is reassuring. Mental health prevalence among young people has risen 80%, a figure the scheme attributes partly to greater willingness to seek help, but one that points to a growing and layered burden. More members are managing multiple chronic conditions simultaneously, and AI prompts, however precise, operate within a health system already under strain. Whether this model can extend meaningfully beyond a large and relatively affluent membership base remains the defining open question.
Discovery Health Medical Scheme, South Africa's largest open medical scheme with 2.7 million members, has quietly conducted an experiment in behavioral nudging that is producing measurable results. Since January 2025, more than 650,000 of its 2.1 million adult members have enrolled in a programme called "personal health pathways"—a system that uses artificial intelligence to send tailored prompts encouraging people to take specific health actions. The early data suggests the approach is working: members who receive these prompts are three times more likely to undergo health checks and twice as likely to get cancer screenings compared with those who haven't enrolled.
The programme operates on a simple principle. Rather than sending generic health advice to everyone, it identifies what the scheme calls "the next best thing" for each individual—whether that's booking a colonoscopy, starting an exercise routine, or scheduling a dental appointment. The specificity appears to matter. When members do follow through on these prompts, the results are often sobering. Nearly a quarter of the health checks performed through the programme revealed results outside normal ranges for conditions like high blood pressure and high cholesterol. For many of these people, the screening caught problems that would have gone undetected until they became serious.
Who is enrolling in this programme tells part of the story. By the end of May, two-thirds of participants were already living with a chronic condition. But a significant portion—38 percent—had poor health-seeking behavior before joining. These were people who skipped annual flu shots, avoided dental check-ups, and generally stayed away from doctors. The programme appears to have shifted their habits. Among those with existing chronic conditions, adherence to clinical protocols improved markedly. They began keeping regular GP appointments, getting blood tests done, and following treatment plans more consistently.
Discovery Health CEO Ron Whelan framed the shift as one of precision. "We have long understood the importance of prevention," he said. "What has changed is our ability to deliver it with precision and personalisation." The company is not alone in seeing results. Discovery Life, an insurance subsidiary of the same parent company, reported in March that its clients prompted through similar systems underwent 18,000 health screenings in 2025, with more than 9,500 focused on cancer detection.
The longer-term picture is even more striking. Mortality among Discovery Health members has declined by 5.6 percent over the past decade, a shift the scheme attributes to behavior change, earlier disease detection, and improved treatments. Cancer survival has improved by 48 percent—members now live an average of 7.1 years longer after diagnosis than they did a decade ago. New treatments like biologics and immunotherapy are part of that story, but so is the simple fact that more cancers are being caught earlier.
Yet the data also reveals emerging challenges. Mental health prevalence among young people has surged by 80 percent, a rise the scheme partly attributes to greater awareness and willingness to seek help, but one that signals a growing burden. More members overall are living with multiple chronic conditions simultaneously—cardiovascular disease, cancer, and mental health issues layered together. The AI prompts can nudge people toward screenings and appointments, but they operate within a health system already stretched by complexity and chronic disease.
What Discovery Health has demonstrated is that technology can change behavior at scale. Whether that change translates into sustained health improvement, and whether the model can be extended beyond a large, relatively affluent medical scheme population, remains an open question. For now, the numbers suggest that sometimes people simply need to be asked—specifically, personally, and at the right moment—to take care of themselves.
Citações Notáveis
What has changed is our ability to deliver prevention with precision and personalisation, guiding each member towards the right actions at the right time.— Ron Whelan, Discovery Health CEO
A Conversa do Hearth Outra perspectiva sobre a história
Why does it matter that these prompts are personalized rather than generic?
Because generic advice gets ignored. When you tell everyone to get screened, most people don't. But when the system knows you haven't had a colonoscopy in five years and sends you a prompt about it specifically, you're three times more likely to book one. Personalization removes the friction of deciding what you need.
What surprised you most in these numbers?
That 38 percent of people who enrolled had actively avoided health care before. These weren't people who wanted to be healthy but couldn't access care. They were actively staying away. The programme didn't change their circumstances—it changed their behavior. That's powerful.
The cancer survival improvement is remarkable. How much of that is the AI versus better drugs?
Honestly, it's both. The new treatments like immunotherapy are real breakthroughs. But you can't benefit from a breakthrough drug if your cancer isn't diagnosed until it's advanced. Earlier detection through screening means people get those drugs when they're most effective. The AI enables the drugs to do their job.
What about the 80 percent rise in mental health prevalence among young people?
That's the uncomfortable part of the story. It could mean more young people are actually struggling, or it could mean they're finally willing to seek help and get diagnosed. Either way, it's a problem the health system wasn't built to handle. Prompts can't fix that alone.
Does this work for people outside a medical scheme?
That's the real question. Discovery's members are relatively affluent and have insurance. They have access to the doctors and tests the prompts are nudging them toward. In a public health system with long waiting lists, the prompt might just be frustrating.