Page 113 - SAMRC AnnualReport 2025-26

P E R FOR M A NC E I N FOR M A T I ON

collaboration and promoted applications for A key area of innovation is the use of unsupervised
bursary opportunities, helping to remove barriers learning techniques in South Africa’s national
to further training. HIV surveys to identify hidden behavioural and
health profiles that are not detectable through
The Unit further contributed to early pipeline
development by hosting high school learners conventional single-outcome models. This work is
through the GenS programme in Durban, Pretoria, directly relevant to strengthening HIV prevention
and Cape Town. In addition, Masters students strategies and informing policy dialogue.
in Statistics graduated as SWRNAC scholarship The Unit is also advancing methodological innovation
recipients under the supervision of Prof. Tarylee through the integration of machine learning
Reddy and Associate Prof. Nonhlanhla Yende-Zuma. algorithms, including random forests, gradient
boosting, and Super Learner, within causal inference
AI in Medical and Health Research frameworks to address HIV testing non-response

The Biostatistics Research Unit is advancing health in South Africa’s national HIV prevalence surveys.
research through the application of artificial These approaches enable more flexible modelling
intelligence (AI), machine learning, and advanced of complex relationships and aim to improve
data-driven methodologies, particularly in the participation and data quality in population-based
analysis of large and complex national survey surveys, thereby strengthening the evidence base
datasets. for national HIV monitoring and decision-making.
The Unit’s commitment to data-driven innovation
is further reflected in the establishment of a
dedicated Machine Learning Working Group, which
provides a structured platform for biostatisticians to
collaboratively develop, evaluate, and responsibly
apply AI and machine learning methods to real-
world public health data.
In addition, the Data Management Division
contributes to the ALIGN project by mapping
relevant health data sources to improve
understanding of available information assets.
The Division is also responsible for developing the
Market Intelligence Hub, a centralised platform
designed to consolidate and provide the National
Department of Health (NDoH) with access to critical
health market data to support evidence-informed
decision-making.

SA M R C A N N U A L R EP O R T 2 0 2 5 / 2 0 2 6 111