Last month, the World Health Organization (WHO) sent joyous shock waves through the global health community and the Global South by officially recommending that the new RTS,S/AS01 (RTS,S) malaria vaccine be adopted into widespread use among children in sub-Saharan Africa and other regions with moderate to high P. falciparum malaria incidence. Now the international community, through Gavi, the Vaccine Alliance, has just stepped forward to help finance the rollout of the world’s first malaria vaccine.
It has been true for some time that sub-Saharan Africa bears the largest malaria burden in the world, with children shouldering the largest proportion of deaths. Over 90% of global cases are on the continent, with children under the age of 5 years constituting a staggering two-thirds of all malaria deaths. This is due to a confluence of factors, not least because of the widespread prevalence of P. falciparum (the most deadly species of malaria parasite), and a very efficient mosquito that spreads it (Anopheles gambiae). The economic impact of malaria is estimated to cost Africa $12 billion every year—a figure that also factors in costs of healthcare, absenteeism, days lost in education, decreased productivity due to brain damage from cerebral malaria, and loss of investment and tourism. The introduction of an effective vaccine offers a beam of hope in the fight to mitigate the massive toll this centuries-old disease inflicts on Africa.
In recent years, countries within the Southern African Development Community (SADC) region of Africa have been making outstanding strides toward the challenging goal of malaria elimination. This commitment to elimination is the guiding mandate of Elimination 8 (E8), which focuses on the following countries: Angola, Botswana, Eswatini, Mozambique, Namibia, South Africa, Zambia, and Zimbabwe. Through E8, these ministerial bodies have resolved to coordinate their efforts to collaboratively work toward the shared regional goal of malaria elimination.
To that end, in 2017, E8 contracted with MENTOR to deliver indoor residual spray (IRS) in border districts of southern Angola. The MENTOR Initiative is a registered nonprofit organization devoted to reducing deaths and suffering from tropical diseases. In Angola, MENTOR has been a stable National Malaria Control Program (NMCP) partner supporting vector control, case management, and surveillance activities since 2004. MENTOR has established offices across Angola and has full country reach in its operations, making it the ideal partner for E8 to engage for expanded IRS interventions. Since 2017, MENTOR has implemented IRS with reported high coverage rates in all campaigns conducted. For its 2020 campaign, MENTOR aimed to improve IRS reporting systems and increase accountability to donors and NMCP. To realize this goal, they chose to pilot Reveal—a digital global good and open source spatial intelligence platform used to drive the delivery of life-saving interventions. MENTOR collaborated with Akros to configure and support the field pilot of Reveal v. 1 in Menongue District (Cuando Cubango Province).
Through Reveal, field teams and managers use detailed household and community maps, protocols, and a data-collection and dashboarding platform for decision-making support to plan, implement, and adjust interventions so they achieve the greatest impact. Akros’ partnership with MENTOR Initiative to pilot Reveal v. 1 in Menogue District to support IRS delivery first included assessment and scoping to understand three key elements: intervention and data collection and management protocols, resource availability, and the environment within which Reveal was to be implemented and sustained. Akros then conducted system configuration to align IRS standard operating procedures (SOPs) with the Reveal platform, set-up mobile application and coverage dashboards, and configured feedback loops and tasking for unique user roles. Following that, the Reveal team participated in campaign planning, including enumeration and base map preparation and work to plan and validate in-field assumptions. Finally, the Reveal team supported implementation in December 2020 and provided post-implementation support.
Maximal coverage defines the effectiveness of vaccination campaigns. Over one year after the start of the devastating COVID-19 pandemic, countries around the globe are vaccinating those ages 16 and older, while vaccine-producing companies are conducting vaccine trials for children. Positivity and hope infuse vaccine distribution efforts, but as these initiatives ensue, attention is being drawn to the challenges of vaccine distribution, namely, those left behind.
Mass vaccination campaigns are critical to the introduction of new vaccines, to providing doses to those who may have missed routine doses, and to giving a second opportunity to those who may not have developed immunity. In each instance, with greater coverage comes stronger, more resilient communities. However, zero-dose children, or children who have not received any routine vaccinations, are often missed by these campaigns. With every child left unvaccinated, communities’ vulnerability to vaccine-preventable diseases escalates. Fortunately, in bracing for future vaccination efforts, we can look to previous initiatives to guide our efforts. In particular, the potential of geospatial data and technology to ensure all, including zero-dose children, are included.
From June through to December 2020, Akros, in partnership with Johns Hopkins University, Macha Research Trust, and the Zambia Ministry of Health, utilized spatial intelligence and the Reveal platform to identify and vaccinate zero-dose children following a nationwide Zambian vaccination campaign for measles and rubella.
We are thrilled to announce Akros has been awarded a Grand Challenges Explorations Grant, an initiative of the Bill & Melinda Gates Foundation! Grand Challenges Explorations (GCE) grants support impactful innovations striving to remedy critical global health and development problems. With this grant, Akros will incorporate into Reveal the ability to integrate data from human movement models. Reveal is a web-based mapping platform which uses spatial intelligence to ensure all receive life-saving interventions. With the ability to integrate human movement models with Reveal, decision makers and field teams will be better able to predict where people will be at different times of the day and seasons to ensure no one is missed with lifesaving resources.
The existing Reveal Platform, improves health campaign coverage by utilizing spatial intelligence and context-appropriate technology. Presently, Reveal maps communities at the household level and offers intervention teams a streamlined interface to plan, implement, track and monitor campaign coverage. Relative to traditional approaches in which local health teams aggregate population data by hand and process it in hard copy, Reveal’s user-centered technology offers a more accurate population count and implementation system—designed to include even the most remote of households.
Delivering health campaigns at high coverage rates can be challenging—particularly in places where frequent movement is common. Permanent relocation may swell or shrink a population, influencing critical resource distribution. Seasonal migration may redistribute a population for months at a time. Daily or weekly movements may make certain individuals more likely to be left out of a campaign. In these situations, health workers may arrive at a household to deliver interventions, but instead find the family has shifted, for even just a few months depending on fishing or farming needs. Take for instance, Nchelenge District in Northern Zambia.
At the time of writing, 33.7 million cases of COVID-19 have been reported worldwide. Regardless of socioeconomic standing, health systems around the world have shuddered beneath the weight of an international pandemic; leading to overflowing ICUs, overextended health care resources, and disrupted critical supply chains.
Accordingly, international attention and funding has turned to global public health and preparedness. The World Health Organization (WHO) has estimated approximately US$1.7 billion total funding is needed to adequately respond to COVID-19 until December 2020. As of September 21, 2020 WHO reports receiving 79.5% of their goal, with an additional 4% expected from pledges—a combined US$1.51 billion raised in the span of a few months. Resources have been rightly and urgently mobilized to offer aid now, but as a responsible global health community, we must look toward the future and set in motion plans to meet anticipated gaps and needs.
If we are to respond effectively and reach the most vulnerable populations, our future interventions to protect people from COVID-19 transmission will inevitably rely on community health structures to disseminate aid and vaccines. If the systems in place are not adequately equipped to respond, those interventions will fall short. Further, if we lack good data on the population and location of communities, getting resources to all those in need will be even more challenging. However, presently, population data are often inconsistent, outdated or quite coarse. Compromised by unclear boundaries or moving populations, the resulting data typically offers only a blurred picture of communities, making it challenging for public health teams to allocate resources effectively.
Zambia, along with a handful of countries within Southern and Eastern Africa, is on track to reduce malaria cases 40% by 2020. Relative to 2015, the country’s progress so far translates to as many as 700,000 cases prevented annually. Such tremendous strides speak to the success of preventative interventions, such as improved access to indoor residual spraying (IRS) and insecticide treated nets (ITNs), but the work is far from over.
The citizens and communities across Zambia rely on annual district planning to determine where, how much of, and what interventions are needed in a particular population. In answering these questions, district teams begin their microplanning processes, determining on a local level the nuances of the year’s malaria interventions. These teams must deeply understand the communities they serve. Who is at greatest risk of infection? Where do they live? What settlements ought to be targeted? And what resources are needed to bring a community out of harm’s way? Lacking this information, district teams cannot fully grasp the extent of preventable malaria cases and consequently limit their capacity to act.
Traditionally, teams of local community health workers aggregate this information on foot and process it in hard copy. This costly and time consuming work flow jeopardizes the data’s accuracy, totality, and speedy delivery to key decision makers. System bottlenecks, limited resources, or a lack of confidence in the data can then undermine the quantitative foundation of an intervention. Weary decision makers might turn instead to outdated data and an imprecise understanding of the population they aim to serve. What ought to be a concrete step in the year’s plan bends to inefficacious circumstances.
Accurate quantification of a population, and the ability to locate this population with precision, are fundamental requirements for reporting the true coverage and effectiveness of public health interventions—such as childhood immunizations, indoor residual spraying (IRS) for malaria, or mass drug administration (MDA) for neglected tropical diseases.
Public health interventions, however, often rely on field teams to locate rural villages or even homes on the ground. In areas where there are no street address systems, or where homes are not mapped, manual searches often result in groups of households being missed; thus preventing the delivery of services to those in need. When service coverage is subsequently reported as a function of the population found, the impact and effectiveness of an intervention may be overstated.
Spatial Intelligence and the Reveal Solution
The transformative field of spatial intelligence is revolutionizing digital health and public health more broadly. Artificial intelligence (AI), digital maps, and spatial modeling are powerful, burgeoning toolsets; but until more recently, they have not benefited field workers and large-scale, labor-intensive campaigns. Now, the power of these digital tools is being accessed by field workers in rural, underserved communities.
Reveal, an open-source platform and global good, uses spatial intelligence to help field workers effectively navigate and deliver life-saving interventions to people who previously would have been missed, increasing the true coverage of interventions and improving health outcomes for vulnerable populations.
Supporting an IRS Campaign in Zambia
Satellite imagery was enumerated to establish a baseline understanding of structure count and spatial distribution in several districts. These were layered with risk maps to target high-risk regions, which enabled users to identify eligible households and assign teams to priority areas.
Using Reveal’s mobile and map-based interface, field workers were able to navigate through communities, identify targeted households, and collect intervention data against eligible households in a coordinated manner within and across teams. The near real-time feedback of data, as a result of the mobile application’s offline and peer-to-peer (P2P) syncing functionality, inspired increased teamwork and cohesion as the campaign progressed and teams worked toward a common goal.
Through dashboards, map-based visualizations, and built-in feedback loops, intervention managers were able to actively monitor campaign progress toward targets, in a given spray area and as a whole, thus facilitating data-driven course correction to optimize performance and maximize impact.
With the support of Reveal, Siavonga District increased its absolute coverage of IRS from 51.5% to 75.5%, while Sinazongwe increased from 31.5% to 61.9%. These changes in coverage were possible due to a better understanding of resource needs. In other words, the use of Reveal allowed districts to better understand the size and distribution of the target population, thus impacting planning and implementation.
On a sweltering summer day in Siavonga, a lakeside jewel of Zambia’s Southern Province, I sat down with Mr. Bisael Phiri, the District Surveillance Officer for Siavonga District, to get his perspective on the malaria situation in his region and at large in Zambia. A breeze lazily floating through an open window and a small desktop fan provided the only reprieve from the oppressive heat as we got down to brass tacks about the progress made in the last few seasons of malaria interventions and how Reveal has impacted that progress.
Mr. Phiri has been working in public health for several years, and is a big picture thinker when it comes to tackling malaria in Zambia. His motivations to work in this field are straightforward, “I know the kind of impact that good health can have on people’s lives. I wanted to do this work to help change the environment so it doesn’t constitute a danger to the public.” His work is based in one of the select districts of Southern Province that utilizes Reveal for their indoor residual spraying (IRS), mass drug administration (MDA), and reactive IRS malaria interventions through a PATH MACEPA and Akros-implemented program. Reveal is a powerful, open-source platform that aids in household-level intervention management and data authenticity by using spatial intelligence and smart planning tools. Mr. Phiri stresses that learning how to use this technology is self-explanatory and that the dashboards, which are tailored to his administrative level, greatly assist his day-to-day responsibilities. In his own words, “the dashboard is important to view on a daily basis because it identifies the status of various teams and shows us where we need to go and where the problem areas are. When we have this information, we can change our strategy based on how well our teams are performing, and change targets midway if need be. For instance, if I didn’t receive updated data from a certain district, I would not know that there’s an issue there. Now because I have these data, I would be able to make a quick plan for how best to move forward with that district.”
Beyond program planning, Mr. Phiri’s work is strengthened by the hard proof the platform provides that work is being done where it is supposed to be done.
Deep in Zambia’s Southern Province, in a town a three-hour drive away from the district’s largest city and economic hub (Siavonga), lies Manchamvwa Health Facility. This clinic serves as the focal point for the health needs of hundreds of people who live in the surrounding villages, and as such, is often overwhelmed with the many health needs of its patients. Malaria season in particular tends to put a great strain on the facility, with peak periods in previous years seeing anywhere from 100 to 200 cases per month.
Over the last couple of years, the Government of the Republic of Zambia (GRZ), with the assistance of Akros and PATH’s Bill and Melinda Gates Foundation-funded Malaria Control and Elimination Partnership in Africa (MACEPA), have been working with the National Malaria Elimination Program district staff to overcome these numbers and improve the health of the local community by using geospatial technology to optimize indoor residual spraying (IRS) campaigns. Recently, the two organizations teamed up again to be the first to ever use Reveal’s spatial intelligence approach to maximize reach and ensure accountability in a mass drug administration (MDA) campaign that distributed antimalarials to the doorstep of each community member in three districts of Southern Province.
The recent history of malaria in Southern Province is one of resounding progress thus far. Due to its proximity to Lake Kariba’s glistening, still water, it is unfortunately a heavily malaria-burdened region by nature. But malaria in this region is highly seasonal, linked to the annual arrival of rainfall from December to April, leaving ample overgrowth and standing water—prime mosquito-breeding real estate. This seasonality provides an attractive window through which most interventions have taken aim. The result has been an impressive decrease in prevalence of malaria parasitaemia among children less than five years of age, from 15.5% in 2006, to 5.5% in 2010, and 0.0% in 2018.1,2 Trends like these make Southern Province appealing as a prime candidate for malaria elimination. However, despite overall improvement in the province’s malaria burden at large, districts directly adjacent to the lake are still at higher risk, as malaria cases have shown to be persistently high in some health facilities despite ongoing interventions.
To propel Southern Province closer to elimination, in 2014 MACEPA supported the national program with a malaria MDA research study in the Southern Province districts lining Lake Kariba, an area with an estimated population of 300,000 people. The rapid malaria reduction in the study area resulted in Zambia adding MDA to its arsenal of interventions in 2017. The country’s experience of malaria MDA—two rounds with one month in between doses—has shown it to be an effective intervention in areas with a strong foundation of vector control, case management, and surveillance. Recognizing that MDA campaigns are most effective when every household and individual in the targeted region are reached, MACEPA engaged Akros for its technical expertise in introducing Reveal as a novel approach to maximize the impact of MDA for malaria control and elimination in this area.
As a spatial epidemiology PhD student, I was drawn to questions about how the environment relates to and facilitates vector-borne disease (diseases that are spread by vectors like mosquitos). These questions and interests tend to lead spatial epidemiology graduate school students (like I was) straight into the land of building spatial models. We effectively try to understand how measures like wetness, greenness, and elevation may combine mathematically to tell us where high numbers of mosquitoes live. If those mosquitoes live near human hosts, or even animals, there may be greater risk of vector-borne diseases.
So, I too built a lot of maps and models during graduate school. I mapped the risk of West Nile virus (WNV) in Colorado, USA. At the time, WNV had, somewhat shockingly, erupted in that region of the US. I also modeled human plague in Uganda—effectively developing maps to precisely depict areas at high and low risk of plague transmission. “Target interventions where it’s red” was the more-or-less summary, where red equaled high-risk areas. Point made. Thesis closed. Safe on the shelf.
But here I sit on the other end of the world, far away from hallowed academic halls that are often lined with dead dissertations and theses like mine. Here in southern Africa, disease transmission is much more tangible. Before, I read about death rates due to malaria and HIV from my school in Colorado, US. Here, I witness the impact of those death rates every day when I drop my kids off at school—new graves being dug closer and closer to the road. I am involved with a local school that is inundated with orphans and vulnerable children—even from one of the more affluent regions of Zambia. In this environment, high morbidity and mortality rates are incessant. Help is highly dependent on securing increasingly limited resources. Navigating the challenging logistics of getting those resources to the right people at the right time and in the right place are often broken. However, despite the acute need to target limited resources, mapping approaches like the one I developed in school are rarely seen nor used to inform interventions.
It is time for the public health community—both globally and locally—to do business differently. It’s time to more appropriately lean on the idea of spatial intelligence through epidemiological and map-based approaches to inform the practice of intervention planning and delivery. Academic, math-based modeling can lend a good understanding of where and how we should focus our limited resources to save the most lives. “Why aren’t these approaches being actively used?” you might ask. Part of the challenge is a lack of tools and finite planning approaches to translate maps and models into operational, boots on the ground, public health programming decisions. Questions like, “Where are all the houses located?,” “Which houses exactly should receive the intervention based upon the model output?,” and “Has the intervention effectively reached everywhere it was targeted?” are challenging to assess, particularly in regions like here in southern Africa, where so many areas consist of rural villages with no addresses.