Across the globe, healthcare providers everywhere are struggling amidst critical shortages of medical professionals and tightening budgets. But if these resource constraints are a universal struggle, they represent a systemic crisis in Africa. Consider the realities of healthcare across the African continent today. Africa carries roughly 25% of the global disease burden, yet operates with only 3% of the world’s healthcare workforce and 1% of global health financing. When you are faced with this level of systemic disparity, traditional solutions simply do not scale. For example, building more hospitals isn't an economically viable option; we cannot construct our way out of this crisis.
What we can do is think differently. Our focus must pivot entirely to proactive and preventative care. We have to move healthcare "upstream" to treat issues before they become critical. And given the rising costs of healthcare worldwide, this kind of proactive thinking focused on context-specific social determinants of health can move the needle in high-income countries as well. For many, there may be an assumption that innovation happens in high-income countries first — only slowly (if ever) making its way to Africa and elsewhere. But often the innovative ways we're forced to think hold lessons for the rest of the world.
Of course, this upstream intervention requires a robust technological foundation. At Aga Khan University, we utilized MEDITECH Expanse to collect, organize, and build a comprehensive repository of our electronic health record (EHR) data. This is the same EHR platform used in the US, Canada, and dozens of other countries, tailorable to all of these unique environments. Across the healthcare industry, it is estimated that roughly 97% of generated data goes unused. Expanse allows us to capture and structure this information to transform raw data into actionable insights, turning that neglected 97% into a powerful tool for proactive population health.
Local context is everything
But having data is only the first step — applying it correctly requires a deep understanding of the specific environment. Local context is absolutely everything. We have found that 60 to 80% of standard care protocols developed in high-income countries do not work for us. For us, the realities on the ground are fundamentally different. We serve different populations, we operate with vastly different resources, and our care pathways must take this context into account.
Take gestational diabetes as a prime example. In high-income countries, standard medical protocols suggest screening mothers between 24 and 28 weeks of pregnancy. This timeline relies on the assumption that the patient will have eight to ten antenatal visits throughout a pregnancy. In Africa, that assumption falls apart. Women present to our clinics much later, and you are incredibly lucky if you get them to go for four total visits. If we relied strictly on Western protocols, we would miss critical intervention windows entirely.
To solve this, we are using both structured and unstructured EHR data within Expanse to identify early risk signals much sooner — at the 16–18 week mark. We’ve mined clinical notes, patient histories, and demographic data to flag high-risk mothers and intervene during the few precious touchpoints we have with them. We have to innovate for our reality, because we know that the best healthcare innovations happen when we don't simply copy and paste solutions from high-income countries.
A continuous learning loop
Our push for localized innovation has allowed us to create a continuous learning loop. The EHR is not just a digital filing cabinet. We are using every single patient interaction as a learning opportunity, and that learning, in turn, informs and improves the entire health system.
A powerful illustration of this is our work in oncology and mental health. Cancer rates are rising across the continent, and through our repository, we discovered that 65% of our oncology patients experience significant psychological distress. Recognizing this issue, we integrated DART (Distress Assessment and Response Tool) protocols directly into MEDITECH Expanse. Now, when a patient exhibits signs of distress, the system automatically triggers bedside interventions for the care team.
Already, this data has revealed trends we didn't expect. For example, we’ve seen higher distress levels among patients in higher economic brackets, a nuance we might have missed without a data-driven approach. Surfacing this nuance allows us to automatically trigger a targeted intervention, utilizing data that would usually go unused. What's even more important about this is that it's not a medical determinant. It’s a mental health determinant that we often don't look at in acute care settings.
The ultimate takeaway is that every interaction is now an opportunity for us to learn and to make use of that insight. Data from every interaction populates our repository, and we can mine that data and look for trends, patterns, and anomalies. Those invaluable learnings are then incorporated back into the protocols we leverage in Expanse, which makes the system smarter and our care more targeted with every passing day.
The future and AI
Artificial intelligence is at the center of healthcare today, but most current AI models are trained using datasets predominantly sourced from high-income countries, with almost no representation from African communities. Relying on algorithms trained on different genetics, environments, and socioeconomic factors, will not yield results that serve our patients safely or effectively.
This is where our work takes on global significance. Working with MEDITECH Expanse to create a Tier 1 digitized EHR data repository means we're now able to partner with global organizations to provide data that has never previously been represented. So, for example, when a pharma company is doing research on hypertension or colorectal cancer, we can be at the table for the first time, because we've taken that data and organized it in a manner that is optimized for advanced research.
Everybody knows data and AI will shape the future of healthcare. But whose future will it be? Impact will be determined by the data that we use. And so for us, inclusivity and representation is also a key factor. By harnessing our own data, respecting our local context, and demanding a seat at the global table, we are ensuring that the future of healthcare includes Africa.
Learn more about transforming data into a strategic asset at Aga Khan University.