Learning from
observational data
Summit | South Africa
3 – 4 Sep 2026
Table Bay Hotel, V&A Waterfront, Cape Town
3 Sep 2pm – 5pm Executive Forum
3 Sep 6pm – 9pm Networking Dinner
4 Sep 9am – 5pm Workshop
Summit | South Africa
3 – 4 Sep 2026
Table Bay Hotel, V&A Waterfront, Cape Town
3 Sep 2pm – 5pm Executive Forum
3 Sep 6pm – 9pm Networking Dinner
4 Sep 9am – 5pm Workshop
The Learning from Observational Data Summit is a curated, invitation-led forum for senior healthcare leaders, policy makers and research practitioners. It is designed to bridge the gap between methodological excellence and real-world executive decision-making, enabling the translation of observational data into actionable, decision-grade evidence.
Convened by Katalitix and delivered in collaboration with the South African Medical Research Council (SAMRC), University of Pretoria (UP), African Institute for Mathematical Sciences (AIMS), Stellenbosch University (SU), University of Cape Town (UCT), University of the Witwatersrand (WITS) and the University of KwaZulu-Natal (UKZN), this is not a typical academic congress, nor an AI summit or a vendor exhibition. It is a high-density, healthcare-specific forum. This event creates a platform to reposition causal inference as a core capability for decision-grade learning, while also enabling researchers and analysts to engage directly with leading experts in the field to build methodological depth and adopt best practices.
South African healthcare sits on a significant body of observational data. The methods to learn causally from it exist. The gap is that these two worlds rarely meet at the decision-making table.
Large-scale observational datasets exist across hospital groups, medical schemes, EHR systems, registries, and public health platforms.
The result: high-cost healthcare decisions made under inference uncertainty, or worse, on the basis of invalid inference.
Yet advanced causal inference is not systematically embedded in operational decision-making. Instead, public policies and corporate strategies continue to rely on descriptive analytics rather than causal understanding.
A substantial proportion of current research and analytics in healthcare remains focused on associations and predictive modelling. While valuable, these approaches do not answer the causal questions required for decision-making.
Dialogue between executive decision-makers and methodological experts is insufficient and largely unstructured.
Prof Ruth Keogh
Professor Biostatistics
LSHTM
Prof Dr Wim Delva
Founder & Managing Director
Katalitix
Dr Nonhlanhla Yende-Zuma
Specialist Statistician
SAMRC
Prof. Innocent Maposa
Associate Professor Biostatistics
Stellenbosch University
3 September 2026
Format:
Executive forum – strategic framing, keynote, applied case studies, executive dialogue sessions, structured networking.
Who attends:
CEOs, COOs and CFOs of hospital groups, medical schemes and administrators · CMOs and Clinical Executives · CIOs and Chief Data and Analytics Officers · Heads of Strategy · Policy Directors · Funders and Regulatory Leaders.
Objective:
Creation of awareness, strategic partnerships and ideation of applications.
Get ahead of the inference curve
Understand how causal methods are already changing the quality of healthcare decisions, and what it means if your organisation is not using them.
Access the right room
Engage peer executives from hospital groups, schemes, funders and policy bodies in a structured, high-trust setting.
Leave with a plan
Every session is designed to produce strategic clarity, not just awareness. Pre-booked strategic meetings ensure your time investment produces tangible follow-through.
Registration & coffee
Welcome: Learning from Observational Data: An imperative for effective, equitable and efficient healthcare
Prof Dr Wim Delva
Managing Director, Katalitix
Keynote: Chances, choices and challenges: tackling causal questions using observational data
Prof Ruth Keogh
Professor Biostatistics, LSTM
Better longitudinal adherence to antiretroviral therapy among virally suppressed people with HIV is associated with reduced occurrence of Serious Non-AIDS Events
Prof Innocent Maposa
Associate Professor Biostatistics, Stellenbosch University
Causal Inference in Practice: Lessons from Clinical and Public Health Research in South Africa
Dr Nonhlanhla Yende-Zuma
Specialist Statistician, SAMRC
The Talent Is Here: Partnering to Grow Africa’s Data Science Ecosystem
Prof Karin-Therese Howell
Executive Director, AIMS South Africa
Learning Fast and Slow from Observational Data with Ancient and Modern Methods
Prof Dr Wim Delva
Managing Director, Katalitix
Closing remarks and wrap-up
Prof Dr Wim Delva
Managing Director, Katalitix
18:00 – 21:00
NETWORKING DINNER
La Parada, V&A Waterfront
4 September 2026
Format:
Hands-on workshop — methodological deep dives, code walk-throughs, cross-institutional dialogue, closing synthesis.
Who attends:
Biostatisticians · Epidemiologists · Health Economists · Clinical Researchers · Data Scientists · Masters, PhD and Postdoctoral Fellows
Objective:
Capacity building, methodological dialogue, cross-institutional collaboration and reinforcement of applied expertise.
Bridge the research-to-decision gap
Present, discuss and pressure-test methods in a setting where the downstream decision-making context is taken seriously.
Build cross-institutional relationships
Connect with peers across academia, clinical research and the private sector in a focused, collaborative environment.
Influence the applied agenda
Contribute to a national conversation on how causal inference can be embedded in healthcare operations, and where your expertise fits.
Registration & coffee
Welcome
Prof Dr Wim Delva
Managing Director, Katalitix
Introduction to causal questions and trial emulation
Prof Ruth Keogh
Professor Biostatistics, LSTM
Estimating effects of one-time treatments (point treatments)
Prof Ruth Keogh
Professor Biostatistics, LSTM
Estimating effects of longitudinal treatments
Prof Ruth Keogh
Professor Biostatistics, LSTM
Advanced topics in causal inference
Prof Ruth Keogh
Professor Biostatistics, LSTM
Q&A session on the datasets, causal questions and methods from the audience
Prof Ruth Keogh
Professor Biostatistics, LSTM
Closing remarks and wrap-up
Prof Dr Wim Delva
Managing Director, Katalitix
Prerequisites for Day 2: We will assume knowledge of regression (linear, logistic, Cox) and basic epidemiological concepts. Experience of R would be beneficial, though code will be provided.
If your application has been accepted, you will have received an email with your personal application code and a payment link.
Select the sessions you were accepted for, enter your application code, and proceed to secure payment. Your place is confirmed once payment is received.
Haven’t applied yet? Submit your application →
From your acceptance email — or click the link in that email to fill this in automatically
✓ Code pre-filled from your email link
🔒 Payments processed securely by PayFast
