# Integrating forest data and health facility surveys to optimise risk-based malaria surveillance in the Philippines

## TL;DR

This study shows how combining forest data with health surveys can improve malaria surveillance in the Philippines by making it more efficient and effective.

## Contribution

The novel contribution is integrating environmental and health data to optimize risk-based malaria surveillance in forested areas.

## Key findings

- Health facility-based surveys increase the probability of detecting malaria infections due to broader spatial coverage.
- Routine malaria diagnostics show decreased sensitivity in forested areas, highlighting the need for targeted methods.
- Risk-based surveillance using forest data is three times more effective at detecting malaria foci than routine methods.

## Abstract

Malaria transmission is highly spatially heterogeneous. Within Southeast Asia, forested landscapes are associated both with increased malaria transmission and reduced healthcare access. Identifying environments with malaria foci is a priority for control and elimination programmes.

Here, we integrate health facility and environmental data to identify optimal surveillance approaches across a forested district in the Philippines. We conducted convenience surveys of health facility attendees utilising tablet-based applications to geolocate participant residences. Malaria infection was assessed using both routine (microscopy and rapid diagnostic test) and molecular methods. Integrating remote-sensing derived data, we assessed how fine-scale environmental factors influence the spatial distributions of malaria infections, diagnostic sensitivity and health-seeking behavior. We evaluated costs and probability of detecting malaria foci for multiple surveillance approaches using different diagnostic methods and target populations defined by landscape data.

We demonstrate that health facility-based surveys increase the probability of detecting malaria infections by increasing numbers of individuals screened and spatial coverage of surveillance systems. We additionally show sensitivity of routine malaria diagnostics varies spatially, with the decreased sensitivity in forests. By targeting diagnostic methods to high-risk environments, we developed a model approach for how to use landscape data within disease surveillance systems. Risk-based surveillance incorporating forest data is highly cost-effective and increases the probability of detecting malaria foci over three-fold compared to routine surveillance.

Together, this illustrates the essential role of environmental data in designing risk-based surveillance to provide an operationally feasible and cost-effective method to characterise malaria transmission.