NOMFUNDO KHABELA | The missing question in SA’s NHI debate

South Africa has spent years debating — in policy forums, boardrooms and courtrooms — who will fund universal health care.

The more urgent and largely overlooked question, however, is whether we have built a system capable of delivering it and whether that system is equipped to support human-centred, AI-enabled care.

This week’s Constitutional Court hearings will test whether the National Health Insurance (NHI) Act’s procedural foundations can withstand legal scrutiny. The outcome is uncertain. But regardless of the court’s decision, any pause in implementation should not be seen as lost time; it is an opportunity to strengthen the system needed to make universal health care work.

The NHI debate in South Africa has, from its inception, been conducted almost entirely in the language of funding and ideology. Who pays? How much? Will the private sector survive? Will the state manage the funds? All very legitimate but second-tier questions.

The real question remains whether we have the clinical workforce, the data infrastructure and the incentive architecture that such a health-care system would require to function. The answer to that question, when examined honestly, should concern everyone who believes South Africa needs better health care.

When care becomes administration

In late 2024, Healthbridge conducted research which found that more than 80% of medical professionals cite inefficiencies in data collection and processing as a primary driver of burnout.

Doctors are spending five to 10 minutes of every consultation simply gathering patient history, time that should be spent on diagnosis, listening and delivering patient-centred care that actually improves lives.

This is the paradox at the heart of modern health care: the more data the system demands, the less time clinicians have to provide care, undermining the very shift towards human-centred care. Data mining can be performed more efficiently by AI than by a human, reducing the physician’s cognitive load.

Imagine a health-care system where your doctor’s success is measured by how well you stay, not how often you need treatment. That is the ethical imperative, and the right technology makes it achievable

Physicians are burning out, not from the weight of treating illness but from the burden of administration. One clinician captured the frustration with quiet precision: “I became a doctor to heal people, not to be a data miner.”

That frustration resonates in consulting rooms worldwide. I recently listened to a PwC-hosted Take on Tomorrow podcast in which all panellists, from global health economists to health-care technology founders in the Middle East, reached the same conclusion: health care must become more digital and, in doing so, become more human.

“AI with purpose” technology in service of the patient and the clinician rather than in service of the system is the conviction at the heart of what we should be building. Practices that invest now in integrated, secure, health care-specific, AI-enabled practice management technology will be best positioned to meet rising expectations and deliver connected, human-centred care in the years ahead.

But to understand why that matters, we first need to be honest about how serious the challenge has become.

The incentive problem we need to resolve

South Africa’s health-care system is built on incentives that actively reward the wrong behaviours. Under the current fee-for-service model, more procedures mean more revenue. Prevention pays less than treatment. Keeping patients healthy, the entire point of medicine, generates less income than managing their deterioration.

The alternative, value-based care, a model that rewards outcomes rather than volume, has been a compelling concept for decades. But its adoption has been constrained not by a lack of will but by inadequate data infrastructure capable of reliably measuring patient- and clinician-reported outcomes on which reimbursement can be based.

This is precisely where technology, used purposefully and embedded in clinical workflows, changes the equation. AI systems that aggregate data across the care continuum, track outcomes longitudinally and identify risk patterns before they become crises do not just improve care; they make a fundamentally better payment model viable for the first time.

Imagine a health-care system where your doctor’s success is measured by how well you stay, not how often you need treatment. That is the ethical imperative, and the right technology makes it achievable.

What the NHI debate is missing

South Africa does not need to resolve the ideological question of whether universal health care is desirable. That question was resolved decades ago. What it urgently needs is a serious, technically grounded national conversation about how to get there.

That conversation must include at least three things currently absent from the public debate:

• A workforce retention strategy that addresses the administrative conditions driving burnout. Doctors do not leave because they stop caring. They leave because the system makes caring unsustainable. Technology that restores clinical time, that handles documentation, aggregates history and translates data into insight, is not a luxury. It is a precondition for retaining the skills the NHI needs to function.
• An honest national audit of digital infrastructure readiness. While the department of health has made progress, including work on patient registration, fraud prevention and the draft NHI Fund regulations, the divide between public and private digital capability presents a genuine risk. With many patients already accessing services in both the public and private sectors, standardised digital infrastructure is essential to maintain patient outcomes.
• A commitment to implementing the proposed value-based care incentive model proposed in the NHI Bill, rather than simply replicating fee-for-service under a new payer. The NHI’s single-purchaser structure could, in theory, be a powerful lever for shifting to outcome-based payment. That potential will be wasted if the new fund reimburses procedures at scale without the measurement infrastructure to hold providers accountable for results.

The window the pause creates

Every month that the NHI is not being promulgated is a month in which the foundational work, data systems, workforce conditions and incentive architecture could be built.

It is estimated that full NHI implementation could take up to 15 years or more. If that is the realistic timeline, the question for policymakers, technologists and health-care providers alike is not “when does the NHI start?” but “what kind of system will be in place when it does?”

The shift to human-centred, AI-enabled care

The future of health care is grounded in a simple, daily reality: a doctor who finishes a consultation energised by connection rather than exhausted by documentation. A patient who leaves not only with a prescription but also with a clear understanding of their health. A system that measures success not by the volume of interventions but by the quality of lives improved.

The decade ahead will not belong to those with the most sophisticated algorithms but to those who understand that in health care, as in all things that matter, technology is never the focal point. The human always is.

• Khabela is a strategic relationships executive at Healthbridge.