Nigeria: The WhatsApp Chat That Could End a Health Crisis #CPHIA2025
Durban, South Africa — Every day in Nigeria, hundreds of women and thousands of children lose their lives to preventable causes, yet in some of the country’s most remote corners, a smartphone notification from HelpMum is now helping rewrite that story.
In a conversation, HelpMum shared how its journey began and why maternal and child health remains such a pressing challenge in Nigeria. Founded in 2017 by Dr. Abiodun Adereni, HelpMum is a Nigerian social enterprise on a mission to reduce maternal and infant mortality in rural, semi-rural, and urban communities across Africa. In areas where trained medics are hard to find, care is unaffordable, and cultural barriers persist, the organization uses mobile technology, low-cost innovations, research, policymaking, and artificial intelligence to address deep gaps in healthcare.
The country remains among the most dangerous places in the world to give birth. According to the UN and WHO data, at least 75,000 Nigerian women died during childbirth in 2023, accounting for nearly 29% of all maternal deaths globally. This means one in every 100 Nigerian women dies in labor or shortly after delivery. For children, the odds are equally devastating: one in every 13 babies dies before their first birthday, and one in eight never reaches age five, according to UNICEF .
It is clear that these statistics reflect a grim reality.
Nearly 20% of maternal deaths in Nigeria are caused by weak health systems, financial barriers, and entrenched social norms. In addition, 2.2 million Nigerian children have never received a single vaccine, the highest number in Africa. For Dr. Adereni, the rising rates of maternal and under-five deaths were impossible to ignore. That’s where HelpMum decided to step in.
Today, HelpMum’s digital ecosystem includes Vaccination Tracker, WhatsApp chatbots (VaxBot and MamaBot), and an AI framework called ADVISER (AI-Driven Vaccination Intervention Optimizer). These innovations provide vaccination reminders, connect mothers to nearby health centers, and support data-driven resource allocation for immunization outreach. HelpMum also provides affordable clean birth kits, trains community midwives through e-learning programmes, renovates maternity homes, and delivers health messages in local languages to improve maternal outcomes.
The organization shared how its innovative tools are transforming maternal and child health, touching the lives of thousands of women and children across Nigeria.
What inspired you to focus on maternal and child health in Nigeria, and how did that journey lead to the creation of HelpMum?
The rising rate of under-5 deaths due to vaccine-preventable diseases and the increasing maternal mortality ratio gap led us to question and understand the reasons fostering the stated challenges, such as poor access to healthcare, misinformation, cultural barriers, and financial barriers stemming from high out-of-pocket healthcare expenditure. Therefore, this led to the development of our innovations with a goal to bridge the gap between healthcare access and women in a community in Nigeria or any African country. We have developed something to ensure that no woman has to die because she had no access to relevant healthcare information (HelpMum Chatbot; Vax AI and Mamabot AI), something as basic as a vaccination reminder to ensure her child gets vaccinated (HelpMum Vaccination Tracker), or a pregnancy risk not detected early enough due to poor ANC assessment (HelpMum Stratify AI). In addition to this, we also saw the need for more organizations to provide more solutions that solve maternal and infant healthcare challenges through the development of our open-source innovations, making maternal and infant healthcare data available to them for them to build upon (Vax Llama, Mamabot Llama and AI Language translator).
Nigeria has one of the highest numbers of zero-dose children in Africa. From your perspective, what are the biggest obstacles preventing children from getting vaccinated?
Some of the biggest obstacles that we have identified from research and from the deployment of our innovations include a leading cause, which is misinformation, limited access to health care, and poor healthcare infrastructure. Also, on a social level, we have identified cultural barriers and myths as a major barrier to children getting vaccinated. We have seen that in some communities, things as basic as visiting health facilities are seen as a weakness, especially among women. Others believe that the vaccines cause illnesses; hence, a major sector of the population starts to believe it. This limits the acceptance of vaccination and ultimately uptake in these communities. On a larger scale, we have noticed that most rural communities experience geographic and infrastructural limitations with respect to healthcare availability, which also limits people from accessing the healthcare needed.
Your abstract at CPHIA will explore the evaluation of language models in maternal health and vaccination guidance in Nigeria. Could you share the key insights or findings from this work and why they matter for improving health outcomes?
Our study compared fine-tuned language models for maternal health and vaccination guidance in Nigeria. The maternal health model, MamaBot-Llama, showed a 4.9% overall improvement in response quality, with notable gains in medical accuracy (+5%) and trustworthiness (+7%). It also cut critical errors by half and was preferred by clinicians in 78% of cases. However, the vaccination model underperformed due to weaker data quality. These findings show that when trained on high-quality, locally relevant data, AI can safely enhance health communication and decision support for mothers, but poor data can have the opposite effect
The development of open-source models like MamaBot-Llama and Vax-Llama is a step toward technological sovereignty. How do you plan to scale this approach beyond Nigeria, and what are the main hurdles in creating a continent-wide ecosystem of locally intelligent health AIs?
Our goal when developing our open source model was to support local and international collaboration on the premise of a drive for innovation within the healthcare space. The goal is to keep developing, refining, and improving our models to inform collaboration from a wider audience across different countries. The hurdles we have faced, especially with respect to vaccination-related content, stem from the different vaccine-related information that applies to different countries, their health systems. Hence, for scale partnering with researchers, organizations across different countries, starting with African countries, means a lot for our scale plan and we look forward to this.
How did HelpMum leverage AI, WhatsApp, and behavioral science to design solutions that address vaccine hesitancy, missed appointments, and poor recordkeeping?
HelpMum leverages AI, WhatsApp, and behavioral science to design solutions that address issues with vaccine hesitancy, missed appointments, and poor record-keeping. First, we have leveraged AI by fine-tuning and improving on AI models to provide health-centered information covering maternal and infant healthcare, specifically beyond general knowledge, which most AI tools provide. Also, we have implemented this on WhatsApp through our chatbot service, a widely used and accepted digital tool that most mothers have access to. Hence, we took this and embedded basic maternal and infant healthcare information into it to reduce misinformation, vaccine hesitancy, and, more importantly, improve vaccination uptake while changing health-seeking behavior.
With the immunization tracker and plans to reach 100,000 women, what does success look like for HelpMum in the next five years, and how do you plan to ensure long-term sustainability?
Currently, we have exceeded our reach by having over 180,000 unique users. In the next few years, we intend to reach over 1,000,000 women through government adoption/ partnerships that provide us with access to more health facilities and reach more states. We believe this is an effective plan to ensure the sustainability of our innovations. Also, it provides us with grounds for validation of the effectiveness of our innovations on a larger scale.
What are the biggest challenges and opportunities in integrating HelpMum’s digital solutions into Nigeria’s national immunization program?
The opportunity we see is that our innovations are already in line with national immunization programs. Hence, this makes adoption and improvement easy. Also, for most of our innovations, we ensure to carry relevant stakeholders along in good time through capacity building sessions and dissemination meetings. This opportunity also gives us access to penetrate the system effectively for state-level adoption.
Your organization supplies clean birth kits to mothers. How has the integration of mobile technology and low-cost innovations, like your clean birth kits, changed outcomes for mothers and newborns in the communities you serve?
The way the integration of mobile technology and low-cost innovations, like our clean birth kits has changed outcomes for mothers and newborns in the communities we serve is that it serves as a pregnancy to child birth journey companion where the mothers can use the MamabotAI to get relevant pregnancy related information and they can assess a health worker using the HelpMum Stratify AI to triage and assess a woman’s pregnancy risk factor and categorize her for further healthcare recommendation and follow up. During her delivery, this woman is also provided with a clean birth kit containing safe, sanitary items. After the delivery, the woman is not left alone in her journey to caring for her infant because she has access to the HelpMum Vax AI to ensure her child is not lost due to a vaccine-preventable disease stemming from a missed vaccination appointment. In most communities where we have had the opportunity to serve, most of the feedbacks we have received speak to our digital innovations supporting eventual vaccination uptake, intent to vaccinate, myth busting, reduction of neonatal and postnatal infection, while also reducing out-of-pocket expenditure for sanitary items purchase in preparation for delivery through the clean birth kit.
Why is training AI on locally curated clinical data so important for maternal health and vaccination programs in Nigeria?
Locally curated data ensures AI systems reflect real Nigerian health contexts, our languages, cultural nuances, and clinical realities. Without it, models risk giving unsafe or irrelevant advice. Our results show that high-quality, Nigerian clinician-reviewed data made MamaBotLlama safer and more accurate, while poor data quality led to failures in the vaccination model. This highlights that for AI to truly improve maternal and vaccination outcomes, it must be built and validated on local knowledge, not imported assumptions.
What are you hoping to be addressed at CPHIA 2025?
At CPHIA 2025, one of the things we expect to be addressed is core public health issues and potential solutions on a larger scale that can be adopted across all countries. We also intend to see how relevant Artificial Intelligence solutions can make a significant shift in the healthcare space, promoting acceptance and positive use.
Looking ahead, what is your vision for scaling HelpMum’s AI and digital innovations across Nigeria – and potentially other African countries – to ensure every child has access to life-saving vaccines? And personally, what motivates you to keep pushing this work forward?
For scale, we are looking at government partnerships to scale across Nigeria, and through our open-source API integration, we intend to partner with organizations across other African countries that are also focused on improving maternal and infant healthcare in their respective countries. We believe this helps us push an acceptance of the power of AI in solving core maternal and infant healthcare challenges.
By Melody Chironda
