From Fred Ezeh, Abuja
Dr. George Komolafe, a clinical informatics specialist, in this interview challenged government and healthcare service providers to embrace Artificial Intelligence (AI) and other digital tools for timely and effective health care delivery. Excerpts:
You mentioned clinical decision support in your recent presentation. What exactly does it mean, and why should Nigerian hospitals care about it?
Clinical Decision Support (CDS) is software that help physicians make better and informed decisions at the point of care by giving them patient-specific information and knowledge that has been intelligently filtered. In Nigeria, where there are many patients and little specialized knowledge in many fields, CDS can help make sure evidence-based recommendations are followed to reduce medication errors, and expedite triage, all of which will improve patient outcomes and save costs.
How does AI fit into the picture?
AI is responsible for introducing predictive analytics and machine learning into the CDS. An AI model trained on local maternal health data, for instance, might identify high-risk pregnancies before problems occur. This translates into early intervention in settings with limited resources, which is a huge benefit when there is a shortage of specialized outreach.
Nigeria still struggles with reliable power and internet challenges. What infrastructural hurdles must we overcome to maximise the opportunity?
Broadband connectivity and dependable power supply are essential for the effective usage of the machine. Rural clinics frequently lack backup generators, but many tertiary hospitals do. While infrastructure catches up, a hybrid strategy that uses on-premise servers with offline capabilities and periodic synchronization when connectivity resumes could fill the gap.
Reliable data quality is often a barrier. How does that affect AI adoption?
The quality of AI models depends on the quality of the data they use to learn. Accuracy is decreased, and bias is introduced by incomplete or non-standardized medical records. Important first steps include training staff on consistent data entry and establishing common coding standards, such as adopting ICD-10 and SNOMED CT (where practical). These would go a long way in fixing the challenge.
What about patient privacy? There’s Nigeria Data Protection Regulation (NDPR). How do you align CDS/AI projects with it?
Patient consent and data minimization are required under the NDPR. So, de-identifying records before analysis and obtaining explicit consent for data use are prerequisites for any AI project. To supervise the creation and application of algorithms, we also require open governance frameworks such as institutional review boards or ethics committees.
Doctors are sceptical of “black-box” algorithms. How do you build trust among clinicians?
Transparency and ability to explain are very essential. Before adding AI, we often begin with rule-based CDS modules like automated drug-interaction checks. Frequent feedback sessions in which medical professionals examine the rationale behind an AI model’s recommendation, demystify the procedure, and foster trust.
What role should governments play in this regard?
They need to fund pilot projects and establish explicit digital health policies. Establishing a national interoperability framework will enable safe data sharing between various labs and hospitals. Uptake can be accelerated by incentives, such as matching grants for hospitals that use certified CDS/AI tools.
Are there any home-grown success stories to reference?
Yes, at Lagos University Teaching Hospital (LUTH), it reduced the time for antibiotic therapy by more than two hours by using a basic AI module to predict sepsis in intensive care unit patients. To help community clinics refer patients more quickly, another Kaduna collaboration trained a machine-learning tool on local radiographs to flag possible TB cases.
How can partnerships with international tech firms and NGOs help?
Global partners often bring technical expertise and seed funding, while local teams provide context and facilitate implementation. For instance, a U.S. university can partner with a Nigerian teaching hospital to co-develop an AI triage tool and train local engineers in machine learning so the solution can be maintained in-country.
Where do you see AI and CDS in Nigeria in the next decade?
Every level of care, from primary clinics to tertiary centres, should have integrated digital health platforms that feed into a central public health dashboard, in my opinion. Personalized chronic-disease management, early outbreak detection, and resource allocation will all be facilitated by predictive analytics. AI-driven decision support has the potential to become as commonplace in Nigerian hospitals as stethoscopes with continued investment in infrastructure, capacity building, and regulation.