Netcare is applying to the South African Health Products Regulatory Authority for approval of an AI system designed to predict early warning signs of life-threatening deterioration in hospital patients, following a clinical study in the group’s intensive care and high care units.
The system is an artificial neural network that applies a machine-learning algorithm to real-time patient data — including heart rate, blood pressure, oxygen saturation and respiratory rate — monitored continuously through Netcare’s digital health platform. The intent is to flag clinical risk before it would otherwise be detectable.
It builds on the quick sequential organ failure assessment, or qSOFA, score, an established measure used in critical care.
“Already, the qSOFA score has proven valuable in the critical care setting as an indicator of a patient’s potential risk of developing serious conditions like heart failure, respiratory failure requiring mechanical ventilation, infection, sepsis, or acute cardiac arrhythmia,” said Prof Reitze Rodseth, head of clinical data innovation and research at Netcare, in a statement on Wednesday.
The algorithm was developed by Telehealth Competence Centre (TCC) Clinical Solutions, a German clinical AI manufacturer that has worked with Netcare since 2023. Netcare is applying to Sahpra to become the South African distributor of the technology so it can be used more widely across its hospitals.
After development by TCC, the algorithm was tested over several years by researchers from Netcare, TCC, the University of KwaZulu-Natal and DigitalOn Tech. Netcare did not release the study data, sample size or performance measures.
New rules for clinical AI
“The remote digital intensive-care level of individual monitoring provides real-time, high-frequency data as a basis for AI-guided prediction of a person’s medical course, proactively identifying where preventive treatment may be beneficial to guide earlier intervention,” said Prof Christian Storm, CEO and co-founder of TCC and a specialist in internal medicine, intensive care and emergency medicine, in the statement.
Both partners said the security of patient data is central to the arrangement.
Read: The Popia problem with agentic AI
The application come after publication a regulatory framework a year ago. Sahpra published its first consolidated requirements for AI and machine learning-enabled medical devices in September 2025, drawing on frameworks from the International Medical Device Regulators Forum, the US Food and Drug Administration, the EU, Britain’s MHRA, Singapore’s Health Sciences Authority and the World Health Organisation.
Under those rules, any AI or machine learning-enabled device requires Sahpra authorisation before it can be used in South Africa, and importers and distributors must hold a medical device establishment licence. Devices are graded from Class A to Class D by risk, and systems that directly influence clinical decisions typically fall into Class C or D — the two highest tiers.
Manufacturers must comply with ISO 13485:2016 and submit a technical file covering software design, validation, risk analysis and cybersecurity, along with clinical evidence that includes real-world validation and analysis of how the model performs across patient subgroups.
Sahpra has not yet begun full product registration for medical devices under the 2016 Medical Devices Regulations, and has said it is engaging industry proactively in the interim.
What comes next
Netcare and TCC are working on further models. These include predicting the risk of renal failure and of respiratory difficulty requiring ventilation, both during a hospital stay and after discharge, as well as calculating the probability that a patient will be readmitted to critical care after leaving the intensive care unit.
“Building on the capabilities of the algorithm, we are now turning our attention to models that further protect hospital patients,” Rodseth said.
Group medical director Dr Anchen Laubscher said Netcare’s shift towards a data and AI-driven approach, under CEO Dr Richard Friedland, places it among global healthcare trends.
Read: South Africa’s AI policy collapse shows up in World Bank report
“Today, judicious use of AI has become an essential part of healthcare. The clinical indicators derived from this technology provide valuable data points to support doctors’ treatment decisions, often creating opportunities to prevent clinical deterioration,” she said.
Netcare has not said when it expects a decision from Sahpra. — © 2026 NewsCentral Media
- Subscribe to TechCentral’s daily newsletter
- Get breaking news alerts on WhatsApp
