Medical Educators Use Delphi Consensus to Define Core Curriculum in South Africa

Medical schools around the world face a deceptively simple question that turns out to be extraordinarily difficult to answer: what exactly must every graduate doctor know, be able to do, and believe? As medical knowledge doubles at a staggering pace, curricula have ballooned to the point of overload, leaving students drowning in content while the health needs of their communities go under-served. A new study from the University of KwaZulu-Natal (UKZN) in Durban, South Africa, offers a rigorous, replicable answer to this challenge, and its approach is drawing attention precisely because it was forged in one of the most demanding healthcare environments on the planet.

The research, published in BMC Medical Education by Mergan Naidoo and Kimera Tamzin Suthiram of UKZN’s Department of Family Medicine together with Lucie Byrne-Davis of the University of Manchester, set out to achieve stakeholder consensus on the core knowledge, skills, and attitudes required during the clinical years of undergraduate medical training. Rather than leaving curriculum decisions to a small committee or tradition, the team systematically mapped what was actually being taught, verified it with the academics responsible, and then subjected every item to a structured consensus process involving the people who use and depend on the curriculum: academic clinicians, recent graduates, and primary healthcare nurses.

The technical backbone of the study was a sequential mixed-methods design. First, the researchers conducted a documentary analysis of module templates, course materials, and curriculum mapping data to identify the existing content of the clinical years. This produced an inventory of 724 distinct curriculum items. To organise this vast catalogue, the team categorised each item according to the AfriMEDS framework, the Health Professions Council of South Africa’s adaptation of the Canadian CanMEDS competency model, which describes the multiple roles a medical practitioner must fulfil, from medical expert and communicator to health advocate and professional. The mapping revealed a striking imbalance: 653 of the 724 items aligned with the Healthcare Practitioner role, while only 71 items addressed the other AfriMEDS roles, a quantitative portrait of how heavily curricula lean toward pure clinical knowledge at the expense of broader professional competencies.

Because documentary analysis alone can miss what actually happens in teaching, the researchers then asked seventeen academic clinicians to verify the mapped content, confirming that the inventory reflected genuine curriculum rather than archival artefacts. This verification step matters enormously for the credibility of the final consensus: participants in the later Delphi rounds were judging a validated list, not a guess. The verified list then entered an electronic Modified Delphi process, a well-established technique for building group agreement without the distortions of face-to-face hierarchy, in which dominant voices or seniority can silence dissent. By running the process anonymously through an electronic platform, the team allowed a nurse in a primary healthcare clinic and a professor of medicine to carry equal weight.

Twenty-one participants completed the first Delphi round and seventeen completed the second, with consensus defined a priori as at least seventy percent agreement on each item. The process was genuinely two-directional: participants did not merely approve or reject existing content but proposed eight additional items they believed were missing, while seven duplicate items were removed during consolidation. In total, 725 unique knowledge, skills, and attitudes items were considered across the complete consensus process. The final verdict was decisive: 698 items achieved consensus for inclusion, comprising 629 aligned with the Healthcare Practitioner role and 69 with the other AfriMEDS roles, while 27 items failed to reach the agreement threshold and were excluded.

What makes the agreed content compelling is how closely it mirrors the epidemiology of South African healthcare. The consensus list reflected the major health priorities encountered in primary care and district health services, including communicable diseases such as HIV and tuberculosis that remain dominant burdens in the region, alongside the rapidly growing tide of non-communicable diseases like diabetes and hypertension. Maternal and child health, mental health, trauma, and emergency care all featured prominently, as did essential procedural skills, the hands-on capabilities a young doctor in a district hospital must possess when specialist backup is hours away. In other words, the curriculum that emerged is not a generic international checklist but a portrait of what a South African doctor actually faces on the ward and in the clinic.

The study’s context is central to its significance. Defining core curriculum content is a persistent challenge everywhere, but it is acute in resource-constrained settings where medical schools must balance an ever-expanding global knowledge base against pressing local healthcare priorities, often with limited faculty time and teaching infrastructure. Content overload is not merely an inconvenience; it dilutes learning, crowds out deep understanding, and can leave graduates theoretically encyclopaedic but practically unprepared for the conditions they will treat. By anchoring every curriculum decision to documented health system needs and a national competency framework, the UKZN team has demonstrated a way to make those trade-offs explicit, transparent, and defensible.

The methodological machinery behind the study also deserves attention. The researchers used REDCap, the Research Electronic Data Capture platform, to run the Delphi rounds, and drew on the institution’s Learning Opportunities, Objectives and Outcomes Platform for curriculum mapping, showing how existing educational technology can be repurposed for systematic curriculum reform. The Delphi technique itself, with its iterative rounds and quantified agreement thresholds, converts a notoriously subjective debate into a measurable process: every item’s fate is recorded, every dissent is visible, and the final curriculum carries an audit trail from initial mapping through verification to consensus. This is the kind of procedural rigour that curriculum committees, often operating under time pressure and political pressure, rarely achieve.

The implications extend well beyond Durban. The authors argue that their methodology and resulting framework offer a practical approach for other medical schools, particularly in low- and middle-income countries, to align curricula with community health needs while maintaining competency-based educational principles. Globally, medical education is shifting toward competency-based models and entrustable professional activities, frameworks that specify what a doctor should be trusted to do rather than merely what courses they have attended. But such frameworks only work if their content is defined locally and legitimately, which is exactly what this study delivers: a stakeholder-informed, consensus-verified core that respects both international competency standards and the realities of district hospitals and community clinics.

There is also a quiet democratic principle embedded in the work. By including recent graduates, who remember what the curriculum failed to teach them, and primary healthcare nurses, who see firsthand where new doctors stumble, the study widened the circle of curriculum authority beyond the professoriate. The result is a core curriculum with 698 items that earned the confidence of the people who teach, practice alongside, and depend on South Africa’s future doctors. As medical schools worldwide grapple with information overload and calls to decolonise and contextualise their teaching, this South African study offers a template: map what you teach, verify it with those who teach it, and let a structured consensus of stakeholders decide what truly belongs at the core.

Subject of Research: Stakeholder consensus on core curriculum content for clinical years of undergraduate medical education

Article Title: Gaining stakeholder consensus on core curriculum content for the clinical years of undergraduate medical education at the University of KwaZulu-Natal, South Africa: a mixed-methods study

Article References: Naidoo, M., Suthiram, K. T., & Byrne-Davis, L. (2026). Gaining stakeholder consensus on core curriculum content for the clinical years of undergraduate medical education at the University of KwaZulu-Natal, South Africa: a mixed-methods study. BMC Medical Education. https://doi.org/10.1186/s12909-026-10629-5

Image Credits: AI Generated

DOI: 10.1186/s12909-026-10629-5

Keywords: medical education, core curriculum, Delphi technique, mixed-methods research, competency-based education, AfriMEDS, curriculum mapping, South Africa, University of KwaZulu-Natal, primary healthcare, undergraduate medical training, stakeholder consensus