AMaSu: Amazon Malaria Surveillance

Malaria is one of the most serious parasitic diseases of the tropics, is caused by species of the genus Plasmodium and transmitted among humans from the bites of infected female mosquitoes of the Anopheles genus. The WHO has established an ambitious plan for control and elimination of the disease by 2030, and Latin American countries have made substantial advances towards that goal, particularly from 2000 to 2015, when symptomatic disease declined by 62% (from 1,181,095 cases in 2000 to 451,242 in 2015), and malaria-related deaths by 61.2% (from 410 in 2000 to 159 in 2015). Nonetheless, in 2016, a considerable increase in case incidence (875,000) was reported in the region. According to the most recent epidemiological report, Venezuela  accounted for 34.4% of the total reported cases in 2016 (240,613). The number rose to 411,586 cases in 2017 (a 71% increase). Incidence has been rising greatly since 2000 (increasing by 359% by 2015), particularly since 2010, where the Amazon region has been the most affected in the country (https://doi-org.ezproxyberklee.flo.org/10.1016/S1473-3099(18)30757-6). There has not been new official data on the incidence of malaria in Venezuela in the latest 4 years (https://www3.paho.org/data/index.php/es/temas/indicadores-malaria.html).

It is estimated that the indigenous peoples of the Venezuelan Amazon, with a population of more than 200,000 persons (INE 2011, www.ine.gob.ve), is one of the groups most affected by malaria and other endemic diseases. However, there are no clear reliable numbers that allow defining concrete health actions to combat the disease. There are no data sources to analyze the evolution of health, nor mechanisms for its collection. In this way, it is impossible to generate health policies that allow health workers to combat the disease. In particular, disease monitoring strategies are required, which involve the collection of data on its behavior, but also its processing and analysis, with artificial intelligence models to help in decision-making processes, which must be accessible to actors in the health sector to optimize their performance. This strategy must consider the collection and processing of data that allows for an integral vision to address the disease, and allow the generation of indicators of its behavior, in a context of poor connectivity, and where the involvement of indigenous communities in data collection is essential.

A malaria monitoring tool based on indirect surveys and data analysis is proposed to follow the evolution of the disease in the indigenous communities of the Venezuelan Amazon. The tool should estimate its incidence, detect outbreaks of the disease, predict its behavior, and diagnose its severity, among other things. To do this, a data collection approach with direct involvement of indigenous communities with difficult access will be used. 

The data collection will be based on indirect surveys, that allow a large coverage of the population with a small sample. The surveys will be distributed to different populated areas (indigenous communities) via a local representative, which will fill the survey and provide any additional relevant data. Then, this will be all sent via radio and other media to a regional aggregation center at Puerto Ayacucho. From there, the data will be sent to the project researchers for curation, processing, and analysis.

The project will use as a basis the techniques developed in the CoronaSurveys project for the COVID-19 pandemic of the IMDEA Networks Institute (design and implementation of surveys, advanced analysis algorithms, etc.), to detect and monitor the disease. This project has been based on the use of surveys with indirect answers to estimate the incidence of COVID-19, by applying the statistical method Network Scale-up Method. Those surveys were done online and made available worldwide, to track other parameters of the pandemic (e.g., deaths due to the COVID). The great advantage is that few survey responses are enough to obtain estimates of the incidence of a disease, as demonstrated in said project (https://doi-org.ezproxyberklee.flo.org/10.3389/fpubh.2021.658544).

Thus, the project will exploit the methodological framework developed in the CoronaSurveys project for monitoring the disease. On the other hand, the project will combine the data and estimates obtained through the surveys, with other  available data sources that are identified, to carry out more advanced analyses. These processes will use statistical algorithms and machine learning, to exploit all this data to study the evolution of the disease, and design strategies to control it. In this way, knowledge will be generated to help in the decision-making processes of the actors of the health sector to combat the disease. The outcome of this process will be used to estimate incidence, detect outbreaks, forecast evolution, and prescribe interventions. All this will be provided to health officials to be used in the fight against malaria.

The main objective of this project is to develop a tool to monitor malaria in indigenous communities with difficult access, almost in complete isolation, in the Venezuelan Amazon. In this sense, the main audience are decision makers in the health sector who can use the information generated (alarms, estimations, diagnoses), and evaluate the interventions proposed by the project. This project is especially interested in helping decision makers in geographic areas with limited communication infrastructure, but also with limited data collection. On the other hand, the target population of the project is the indigenous communities of the Venezuelan Amazon, which will see improvements in their health by means of the collected data and the outcome of its processing. For this reason, their involvement in the data collection processes via indirect surveys, and in their interpretation, is a challenge in the project. Finally, the dissemination of the results to the general public is essential, to increase its support. So, a health dashboard will be defined to make all our data and results public. Particularly, the collection of georeferenced data that can be represented in a dashboard helps with making the right decisions on issues of epidemic control and optimization of health efforts.

The project will achieve several impacts: (i) the information that will be provided will be useful for decision makers in the health sector and indigenous communities to combat malaria, in particular, because there are no official sources of information, (ii) It is a pilot experience that can be extrapolated in a second phase to other indigenous communities in Latin America, and to other communities in the world with difficult access later, (iii) it is a pilot experience of combining hard and soft technologies for the supervision of diseases, with appropriation of knowledge and participation of local actors (decision makers, indigenous communities, among others).

The project will have the participation of local indigenous organizations (ORPIA, among others), and with researchers who have been working on different aspects related to the area of health and local and indigenous rights in the Venezuelan Amazon (GTAI, Instituto de Zoología y Ecología Tropical of the Universidad Central de Venezuela, among others). All of them have extensive experience in various projects that promote the well-being of Venezuelan indigenous communities. Finally, the official actors of the health sector are involved (Directorate of Health of the Government of the State of Amazona, Regional Directorate of the Ministry of Indigenous Peoples, among others). Data collection will be carried out by local indigenous communities, articulated through the Venezuelan research groups involved in the project. In addition, the dashboard will be socialized to the official organizations by the entities that participate in the project, to promote its use by them.

Also, the IMDEA team has been working together on the CoronaSurveys project for more than two years. Its members have different skills and backgrounds, from Statistics to Psychology, including Computer Scientists and engineers. In this project, the team has developed several important skills such as: Deploying a computing system and mobile apps, creating an online survey for all countries of the world translated to several languages,  analyzing the collected data to estimate the incidence of the disease, among other things. Additionally, the team has participated in two data challenges, in which it had to merge the data produced in the project with external data, and develop knowledge models to forecast the behavior of the pandemic and to propose intervention plans to reduce cases. The whole system is ready and can be extended or adapted as needed, given the experience and know-how of the team. Everything is ready to deploy new versions of the survey to monitor new pandemic parameters or new health risks, and to integrate the resulting data with any other available data.

We are applying to this Challenge because we think the AMaSu project can be useful for communities that are isolated, both spatially and technologically, because new unconventional methods are required to monitor the evolution of local diseases, and to analyze the data to generate useful knowledge for decision makers and the communities themselves. This monitoring process is instrumental to be able to evaluate the performance and evolution (improvement or deterioration) of primary health care systems. Malaria has been chosen as a paradigmatic disease due to its current impact on the communities of the region. Thus, this is an extension of the CoronaSurveys project for the Malaria disease in communities of difficult access, which will allow its maturation in other real context, to think about its extrapolation to regions of difficult access, around an issue as sensitive as health.

For that, we need funding to fully develop the system proposed in this context, support to make it visible to stakeholders and decision makers, and financial support for the participation of the local partners. On the one hand, part of the funding will be used to maintain the computational infrastructure, and to enroll researchers and developers that can work in the project full time. On the other hand, the funding will be used to increase the local participation in the surveys, in order to have a steady flow of data (responses).

This question can be answered from two dimensions.

At the level of social health, because the use of indirect surveys to generate data on the evolution of Malaria will be combined with data analysis techniques that will exploit it to generate knowledge models (prediction, diagnosis, prescription, among others), and interpretability techniques, which allow decision-making processes based on data. These models will be appropriated by local actors (decision makers, communities, NGOs), for use in combating Malaria. In addition, the use of communication methods through radio frequencies, which allow communication and data exchange through voice and digital media (internet), allows an elementary connection and the exchange of information in indigenous communities.

At a technological level, to our knowledge, this is the first time that the Network Scaling Method has been used to monitor health in isolated communities. The previous times these techniques have been used, they have been limited to a round of surveys collected in a small geographic area (for example, a city) or massive cases (CoronaSurveys project). In our project, the indirect surveys will be managed by the local communities. The context of the project makes it necessary to define new solutions for new problems. For example, how to manage the variability of survey frequencies, or extrapolate them to nearby geographic communities, or integrate with public data sources on malaria. Our solution can lead to the extension of the low cost sampling and data analysis techniques developed in the CoronaSurveys project to regions with difficult access and very little data.

It is intended to impact the health system to improve care in communities with difficult access through a monitoring tool with high community participation. In 5 years we should have extrapolated its use to other communities in the Latin American Amazon, and eventually to other regions of the world that are difficult to access, but also for areas other than health.

We will also make an impact with an easily scalable and reusable tool in different geographical contexts, with low-cost data capture and analysis mechanisms. In 5 years, many of the tools, in particular, for building data analysis models, should have been automated, to quickly allow their reuse in different contexts.

The main metric that we would like to use to measure the impact of the Malaria solution is the control achieved over it. Also, we will be measuring the use of indirect surveys, scientific articles published, citations of our articles, data analysis models developed for the dashboard, software products launched, etc. These indicators will provide an idea of the impact achieved in the project.

Finally, also, we will seek to generate impact indicators that can be determined with our tool, established in the Sustainable Development Goals for the area of health, linked to guaranteeing a healthy life and promoting well-being for all at all ages. We will define several metrics for measuring the progress of the project, such as:

• Malaria prevalence/incidence

• Malaria local outbursts (early detection, geographical distribution, number).

• Population sampled.

• Frequency of collection of data and generation of estimates

• Stability of the data flow 

Finally, the expected results would be:

• 2 rounds of data collection and estimates

• Data collected from 20 isolated indigenous communities of the Uwottuja indigenous territory  

• Data collected from a total population of 20,000 people

The key technologies are based on the CoronaSurveys project, which uses the Network Scale-up Method that considers indirect surveys. In this context, several advantage provide this survey as i)  participant does not provide his/her own personal health data, but the data of his/her contacts, ii) no personal health data is provided, preserving privacy Also, as part of the CoronaSurveys project there are analysis algorithms to estimate daily the parameters of the COVID-19 pandemic. These algorithms use statistical and machine learning techniques, which will be extended to the context of this new project. Finally, as part of the CoronaSurveys, we have defined a methodology about how to complement the data collected via our survey with other data sources, in order to have more data to build the knowledge models. With the aggregation of data from several sources can be obtained better data driven models of the disease using a small number of surveys.   

These three aspects, the indirect surveys, the data analysis algorithms, and the integration of the survey data with other data sources, give our project a solid base, now applied to a new context, with specific challenging requirements.

At present, the collection of health data is carried out by health personnel established in certain indigenous communities. These data are stored in morbidity notebooks, to later be shared with the health authorities, with a frequency that could vary between one to two times a year.

Subsequently, the health personnel of the indigenous communities will share the material compiled from a network of amateur radio and authorized government stations that provide worldwide radio email using radio pathways where the internet is not present (https://www.winlink.org/), allowing low-cost connections in rural locations.

Our team for the project is made up of a diversity of people, both in the technology development team and those who will be deployed locally in indigenous communities. The team at IMDEA is from a great diversity of nationalities, with different social and cultural backgrounds. On the other hand, the local allied organizations associated with the project are the indigenous communities themselves, accompanied by the academic and non-governmental organizations with which they have been working for several decades on various issues, and allies from the public sector, in particular, from the field of health. They will be involved from the beginning of the project. In addition, other public sector actors and organizations that are interested in the project will be welcome.

The objective of the project is the design of social practices and useful technological solutions for local environments, which appropriate them for use in their daily lives. We hope that this project develops an inclusive, open, human-centered technological solution that allows the incorporation of new knowledge, particularly local knowledge, into it. Also, we aspire to extend this experience to new communities with difficult access, as a way to include socially marginalized groups.

Particularly, inclusion, diversity and equity will be guaranteed during the development of the project for four aspects: (i) The local roots of the solution to be developed, which implies including the knowledge, vocations, and capacities of indigenous communities, the potentialities of their environment, as well as the cultural and social practices that exist in them. (ii) The participatory decisions of all the actors involved in the project, which implies spaces for dialogue that enable debate between the different actors, and decision spaces that make it possible to achieve common objectives. (iii) The next aspect is local control over the development of the project, through mechanisms that allow the project to be audited and followed up, such as control instruments, impact indicators on the communities, among others. (iv) Finally, the last aspect to be considered in the project is the local enrichment derived from the execution of the project, which in turn provides feedback to the project execution team. In particular, in projects linked to the generation of products based on the application of knowledge (as in the case of ICTs), the local appropriation of knowledge is a fundamental indicator to observe local enrichment.

The foregoing will be guaranteed through a project development framework guided by participatory methodologies, centered on users, with leading roles of local allied actors throughout the process of specification, design and deployment of the final solution, in such a way that the technological, and therefore social, solution to be deployed, has the required local roots.

We will work with low-income organizations to offer social services focused on providing access to those who couldn’t otherwise afford it. Initially, in the domain of health monitoring. In this context, the solution must be low cost, with high operating efficiencies. In this way, there will be a service subsidization, such that will be defined other products or services with our monitoring platform (e.g., in marketing)  to an external market to help fund this social program. 

Also, based on this pilot, we hope that regional governments realize how easy, participatory and inexpensive this method of collecting and analyzing information is.

For this, the different aspects used/developed in the project will be exploited, such as the indirect surveys, the data analysis algorithms, the participatory methodologies, in order to define  any type of business that leverages its assets.

We will divide this part into two, grants and financial aid received for the development of the CoronaSurveys project, and grants and aid received to improve the well-being of indigenous communities

In the case of the CoronaSurveys project, the grants and aid received 

• Individual patrons and donors.

• Funded partially by the Comunidad de Madrid and IMDEA Networks Institute, under the CoronaSurveys-CM grant.

• Funded partially by the Regional Government of Madrid and the European Union through the European Regional Development Fund (ERDF) as part of the response from the European Union to the COVID-19 pandemic in the context of REACT-COMODIN-CM-23459 research project.

In the case of the projects linked to well-being of indigenous communities, the grants and aid received:

In recent years, the Regional Organizations of Indigenous Peoples of Amazonas (ORPIA) have received various funding from FORD foundation, AVAZZ, Ambassade de France, Rain Forest US in the area of prevention against the Covid-19 epidemic. Recently, the coordinator of indigenous organizations of the Amazon Basin, in conjunction with the International Union for the Conservation of Nature, are beginning to finance the Initiative “Amazonia for Life 80% by 2025”. which seeks to avert the tipping point in the largest forest on the planet. The indigenous peoples across the basin and allies are raising their voices to make a call to protect the Amazonia and safeguard our future (https://amazonia80x2025.earth/).