Dr. Abdisalan Noor: How billions of data points can help beat malaria

Dr. Abdisalan Noor entered the field of malaria in 2000, just in time to see it transform from a neglected area of research to a key global focus with that year’s Abuja Declaration , as well as an influx of funding following the creation of the Global Fund to Fight AIDS, Tuberculosis, and Malaria, and the Presidents’ Malaria Initiative, among others. As the former head of the Strategic Information for Response Unit at the World Health Organization’s Global Malaria Programme, he is part of the generation of malaria experts that helped prevent about two billion malaria infections and save 12 million lives around the world since 2000.

Over the past seven years, however, he and his colleagues have witnessed the progress against malaria plateau. Exemplars News spoke with Dr. Noor, who currently serves as Visiting Professor at the Harvard T.H. Chan School of Public Health, about the progress made to date and what it will take to regain the momentum and achieve global goals.

You originally wanted to focus your research on geospatial ecology, hoping to map out biomass, water resources, and livestock density to develop a resource management plan to reduce conflict among the pastoral communities in northern Kenya, where you grew up. You ended up using these same skills to fight malaria. Can you tell us about this background and how it helps your understanding of malaria?

Dr. Noor: I obtained a degree focused on geospatial engineering where you learn about mapping and spatial data, temporal data, and the dynamics that manage spatial-temporal systems. Through this, I learned the value of thinking about problems in a system, to think holistically, as opposed to just focusing on singular aspects.

Malaria, being a largely climate-sensitive disease, is environmentally driven. It is also influenced by health systems and social determinants. It has, therefore, spatial and temporal patterns. These patterns inform the epidemiology of the disease and the effectiveness of interventions. I have focused a great deal on that in my research and global health work.

One of my first jobs in malaria research was to map malaria prevalence in communities to develop an empirical map of malaria risk in Kenya. Following this work, I then developed the first national database of health facilities that included information on location, type of facility, health services provided, and population catchment around each facility. This became the basis of the first mapped Master Health Facility List in Kenya. When this information is linked with malaria risk and interventions in communities, you increase the power of that data to make malaria control decisions.

Ever since then, my focus has been on using analytics to inform our understanding of the epidemiology of malaria. This includes developing national strategic plans, prioritizing resources, evaluating the impact of interventions, and making decisions to ensure the fight against malaria is continuously improving.

Why is malaria surveillance important and why has it suffered from underinvestment?

Dr. Noor: You cannot run an effective public health program if you do not have a good surveillance system. Whether it's malaria, polio, health emergencies, TB, HIV. The data from such systems is at the core of making the best possible decisions to give you the best possible outcomes.

I started in malaria when there was a raging epidemic globally and there were hardly any interventions in communities in the lowest income countries. People were being treated based on fever and the drugs used were failing due to parasite resistance. You only knew people were dying of malaria because inpatient facilities were flooded.

The introduction of rapid diagnostic tests at scale, starting around 2012, not only ushered in a new era of malaria case management, but also of information of increasing quality. Since then, more than 3.5 billion malaria rapid diagnostic tests have been distributed. That represents billions of data points. At the same time, the digitization of information systems grew exponentially. Not only are you getting more data, but you're also getting more high-quality data, which in turn improves the quality of decisions. Add to this to the over 60 malaria indicator surveys conducted since 2005, and we are considerably better off than we were at the turn of the century.

Unfortunately, and somewhat understandably, investment in surveillance systems and use of data still has a long way to go, as the focus for many years has been on getting the commodities out there to save lives, without a corresponding clarity in our understanding of how much burden of disease you are dealing with, or how much benefit you are accruing from all of these investments. To a great extent, we’re still relying on heavily modeled data, particularly for Sub-Saharan Africa, to tell us something about the progress we are making on the burden of disease in many countries.

The upfront expense in creating fit-for-purpose surveillance systems seems a little daunting, but these systems are the one guaranteed thing that will provide long-term and sustainable benefits in the fight against malaria. These systems allow us to work better, smarter, and be more adaptive, with many benefits to the wider health system.

What’s new in malaria surveillance that you're particularly excited about?

Dr. Noor: Three things strike me as really exciting.

First, there's an increasing appetite to use granular data to make decisions. That is critical to how well your surveillance system grows, because if you don't use your data, you can be guaranteed your system will not improve.

This includes both the routine use of data on a daily, weekly, and monthly basis to improve programmatic activities, detect epidemics, and respond to those epidemics, as well as for medium or long-term strategic planning and priority-setting to define national goals, approaches, resources, and funding needs.

This requires a combination of data visualizations, geospatial methods, health economics methods, and mathematical models. This mixed-methods analysis has been the focus of a lot of the work we've been doing over the past five years in scaling up the subnational tailoring of malaria interventions.

The second thing that I'm excited about are the increasing investments in the digitization of surveillance systems. Digitization can help countries better map where health facilities are and where people are, as well as transmit data on interventions and disease as quickly as possible.

The third thing I'm excited about is the emergence of technological and methodological innovations in geospatial and mathematical modelling, including the use of artificial intelligence.

What tools and strategies can we use to get back on track in the fight against malaria?

Dr. Noor: Progress lies, first, in doing things we've been doing better and smarter. We have new tools, like new generation nets and vaccines. They will contribute a lot, but they're not a silver bullet and we will always need to discover and develop new tools.

I will summarize five ways forward.

First, while there has been progress, we still don’t have the adequate granular level data we need to make better decisions, to know what we’re fighting, where we’re fighting it, and how to improve our response. You cannot treat large regions within countries as if they have one malaria challenge. You have to accept the problem is local and has significant variations.

Second, the complexity of the malaria system from the vector side, the human side, and the parasite side means that you're going to deal with things you never expected. We must be quick on our feet to adapt to change. We are nowhere near close to the dynamism, for example, that the health emergencies sector shows when responding to unexpected outcomes and emerging threats. We must avoid inertia.

Third, we need to invest in capacity, particularly in human and institutional capacity. The human capital, the people who do the job daily, need to know how to do things better, design locally, detect threats, and adapt to those threats. I see increasing capacities within systems in terms of guidelines, protocols, and procedures as well as digitization of processes, but we are not where we need to be.

Fourth, all of this takes money. Annual malaria funding has basically remained the same over the past 10 years while the population in malaria endemic countries has increased rapidly. We need more funding, but we also need to do more with what we have, be imaginative in the way we fund the malaria enterprise, and make a better business case for malaria investment, including better governance and accountability mechanisms.

Fifth, we need unity of purpose, especially at global leadership levels so as not to lose sight of the goal of malaria eradication. We need to genuinely converge around a common plan, set aside unhelpful institutional agendas, and improve governance across all levels, so that we pay attention to specific malaria needs, but also the important synergies with health and other systems.

What do you see as the role of operational research and studying positive outliers in the malaria field?

Dr. Noor: It's critical. When you're faced with the most challenging situations, the positive stories become even more important. Telling those positive stories is a good way to impress upon countries that success can be realized.

That’s why I am working with the Exemplars in Global Health team to examine countries and regions that have been positive outliers in the use of subnational data and analysis to fight malaria. I don’t want to pre-empt our research but there are a few great examples in this area.

In Africa, Kenya, where I worked, and Tanzania, followed by countries like Rwanda, Guinea, Nigeria, Mozambique, and Zambia, have been great examples of using data to tailor interventions sub-nationally and achieving some level of impact. Another example is Senegal, which has gone the route of combining traditional epidemiological and programmatic data with genomic data to understand the malaria epidemiology better and make timely decisions and respond to threats.

Outside of Africa, another strong example is the Greater Mekong Subregion’s regional malaria elimination initiative, launched to stop the spread of artemisinin resistance. From the initiative’s inception in 2012, one of the core recognitions was the value of using granular data at the village level to respond to malaria. They not only identified villages at risk, but also mapped high-risk groups such as mobile and migrant workers, including forest workers. By doing that, they could design their interventions in a very targeted way. Over the past ten years, they reduced the overall burden of malaria by nearly 80%, and the specific burden of P. falciparum by over 90%.

This interview has been edited for length and clarity.