Improving disease modelling capacity key to beating malaria

Mosquito in a swamp.(Courtesy)

As we observe World Malaria Day, it's a good time to look back on our journey so far – the progress we have made and the long road ahead. Beyond the horizon is a Kenya where malaria epidemics are no longer surprises but something we can predict and prevent. Where interventions are not just widespread but also smart, targeted, and cost-effective. That future is within reach, and it starts with epidemiological modelling.

Despite all the progress we have made, malaria still continues to cast a long shadow over Kenya, threatening millions. The 2024 World Malaria Report estimates that 3.3 million cases were recorded in 2023—a drop from previous years, but still far too many. About 70 per cent of Kenyans remain at risk, and the transmission risk varies across the country, from a 19 per cent prevalence in lake-endemic regions to less than 1 per cent in low risk areas.

The combination of the new worrying trends, such as new outbreaks in previously low transmission areas like Turkana County and the recent discovery of the urban-adapted invasive species, Anopheles stephensi mosquito, has made the ="https://www.standardmedia.co.ke/health/health-science/article/2001516956/why-kenya-must-reinforce-its-fight-against-malaria-now">fight against malaria in Kenya< more difficult. The presence of this malaria transmitting vector is now confirmed in nine African countries which complicates the situation further.

The detection of the Plasmodium vivax parasite in northern Kenya adds another layer of concern about resurgence and complexity in transmission. Clearly, a one-size-fits-all approach won’t work. We need smarter strategies.

During my time as a field epidemiology resident at Kenya's National Malaria Control Programme in 2023, I encountered the idea of sub-national tailoring for intervention optimisation, a powerful concept made possible by epidemiological modelling. Simply put, it means using data and simulation to select the right interventions at the right places at the right times.

However, the modelling process hit a roadblock because there were not enough local experts. This problem stems from an over-reliance on foreign consultants. These types of structures do not understand our local health systems and the communities they serve. This ultimately leads to model outputs that, although they look good on paper, miss the reality of disease patterns and dynamics on the ground.

This gulf of knowledge and lack of experts is unfortunately not only specific to Kenya. This is a fact much more pervasive in the rest of the malaria-endemic nations in Africa. Countries lack sufficient local modelling capacity which means they have to depend on other sources for planning, which is often disconnected from local context and becomes problematic at implementation.

While great progress has been made in collection of data, with over 90 per cent of health facilities in Kenya reporting timely submission of malaria reports, the country is still striving to become malaria-free. However, we still have not unlocked the maximum potential of data collected, and that is where modelling comes into play.

Epidemiological modeling takes into consideration the seasonal transmission, climate changes, insect profile, and varying levels of intervention coverage over time. It facilitates 'what-if' analyses for understanding how different interventions—vector control, chemoprevention, case management, and vaccination—work in unison. It can also determine how to allocate limited resources for maximum impact.

In Kenya, only 29 per cent of households have enough insecticide-treated nets, 22 per cent of pregnant women take the recommended three ="https://www.standardmedia.co.ke/health/health-science/article/2001516963/new-vaccine-targets-mosquito-weak-spot-offering-hope-for-malaria-fight">doses of malaria prevention<, and 36 per cent of children with a fever are tested at health facilities.

If we have the right data and expertise, modeling can go a long way in helping us reduce malaria cases. It can also save us from throwing money at strategies that don’t deliver and even show us just how much impact certain interventions could have.

For Kenya to fully leverage the power of modelling, there is a need for locally trained professionals who understand the math as well as the local health landscape. Besides that, we should build local university capacity by establishing specific modelling training courses and strengthen data systems to facilitate high-quality analytics.

Creating career opportunities in public health for disease modellers can also strengthen collaboration between modellers, decision-makers, and healthcare providers. This ensures that models lead to real, actionable change on the ground.

Modelling can support other diseases such as dengue, cholera, and newly emerging diseases, thus making Kenya resilient to any new health threats. This is a proactive investment in the health security of our nation.

To make this vision a reality, we propose that the Ministry of Health sets aside at least 5 per cent of its annual budget over the next five years to go towards developing disease modeling programmes and creating jobs in the public sector for modellers.

Universities and research institutions need to step up too. They need to create disease modeling programmes at both Master’s and PhD levels that fit Kenya’s unique needs while also keeping in line with global best practices.

Moreover, local and international funding agencies ought to focus on backing research, training, and building up our data infrastructure. This is key to developing a sustainable modeling ecosystem that can really make a difference.

Kenya’s battle against malaria is not just about having the right tools at our disposal. It’s about knowing how to use them in the best way possible.

Dr Githinji is a research fellow for Malaria Interventions Modelling at the Center for Epidemiological Modelling and Analysis, University of Nairobi