Improving Clinical Data

Africa’s healthcare systems face a persistent challenge: fragmented and underutilized health data. The use of electronic medical record (EMR) systems is vital to resolving this, but offers the potential to do more! Under a Gates Foundation funded project, we have supported the 1beat EMR system – mobile, clinician-friendly platform that was co-designed by public sector clinicians – to not only capture, but  organize, and support real-time decision-making at the point of care for young patients at the Chris Hani Baragwanath Academic Hospital’s paediatric ward in Soweto, South Africa.

Through this grant, the 1beat EMR is being enhanced with a fine-tuned Large Language Model (LLM) and FAIR data stewardship to transform how patient data is used, stored, and valued. The project has incredible potential, with opportunities for further value creation through enabling University medical student training, pharmaco-vigilance studies, clinical trials, community surveillance, and even reducing the burden of medico-legal expenses on hospitals.

In 2025 we will be piloting the LLM-enhanced version of 1beat, now offering a simple diagnostic support tool and palm-of-the-hand access to hospital specific clinical guidelines for treatment. The system is being geared to improve clinical note-taking, and reduce clinician workload. The model is trained on local, de-identified clinical notes and paediatric guidelines, offering context-aware support that is culturally and medically relevant. Integrated directly into the 1beat app, the LLM serves as a co-pilot—not a replacement—to enhance the clinical decision-making process. In doing so we believe that we may have inadvertently created a tool that can be used to improve access to knowledge for medical students and interns, while simultaneously reducing time from admission to diagnosis and beginning treatment.

To ensure ethical use and future-proofing, 1beat has also been FAIRified (Findable, Accessible under governed access, Interoperable, Reusable). This alignment with global data standards enables patient data to be used in secure, federated environments without ever leaving their source, ensuring both privacy and usability for research and health innovation – In other words, the data is owned by the patient and the hospital and Wits University remain its custodians.

Why Does This Matter

Poor data quality can have life-or-death consequences. By enhancing data completeness and supporting diagnoses, this innovation strengthens care coordination, accelerates treatment decisions, and creates a foundation for high-quality, African-led clinical research. It also supports the development of a sustainable, locally governed health data ecosystem that can attract ethical investment.

Through monitoring this innovation, we hope to see that a locally created product is saving lives — and in doing so, proving that Africa holds the talent, insight, and determination to solve her most pressing challenges, on her own terms