As part of our commitment to social enterprise, we were thrilled to collaborate with the World Health Organisation’s (WHO) Western Pacific Regional Office (WPRO) on a groundbreaking project. Our mission? To map the population health landscapes across ten countries in the region: Cambodia, Lao People’s Democratic Republic, Mongolia, Malaysia, Papua New Guinea, The Philippines, Solomon Islands, Tonga, Vanuatu and Vietnam.
This ambitious project aimed to provide a detailed assessment of the region’s health status through key indicators, including:
Population vulnerability
The WHO defines vulnerability as the degree to which individuals, communities, assets, or systems are susceptible to hazards, shaped by an array of physical, social, economic, and environmental factors. To estimate and map this vulnerability, we examined contextual factors particularly relevant to low and middle-income countries. These included age demographics (such as populations over 60 years old), degree of wealth, access to clean water and sanitation services, time taken to reach the nearest health centre, among others. Each country was analysed both independently and within the broader context of the Western Pacific region.
Life expectancy at birth
Life expectancy, as defined by the WHO, represents the average number of years a newborn is expected to live, assuming they experience the current mortality rates throughout their life. This metric offers a snapshot of a population’s overall mortality level. In this project, we mapped and assessed life expectancy, a variable often only available at low resolutions – typically at the provincial, state, or national levels – in low-income countries.
Vaccination coverage in children
Vaccination is recognised as one of the most effective and cost-efficient public health interventions for children. Despite this, approximately 14.5 million children were unvaccinated in 2023 – so-called zero-dose children. To evaluate vaccination coverage for children aged 12 to 23 months, we used diverse data sources including motorised time travel to healthcare facilities, nightlights in the area, road proximity, under-five child mortality rates, population density, and the proportion of children experiencing malnutrition or stunted development. The model’s performance could be validated in regions where reliable vaccination coverage was available.
Challenges and outcomes
Throughout this project, we encountered several intriguing challenges. For instance, our models had to be trained on a relatively small number of samples. Additionally, we grouped countries by geographic proximity to gather sufficient samples for learning. However, this approach required careful consideration of the geopolitical and socioeconomic differences that could influence the health indicators we were investigating.
Ultimately, our efforts culminated in the creation of a customised geoportal that visualised all outputs, which was then shared with key stakeholders.The results of the WHO WPRO project were overwhelmingly positive. By leveraging contextual data, we were able to model population vulnerability, life expectancy at birth, and vaccination coverage among children at a high resolution – down to the smallest administrative level in each country. This fine-grained approach allows healthcare organisations to identify high-priority areas that may have otherwise gone unnoticed, enabling them to tailor much-needed interventions at a subnational level.