ESA signature

Ali Ahmad |

EO-guided telemedicine planning to strengthen healthcare accessibility in Bhutan

According to the most recent national census conducted in 2017, Bhutan had a population of 727,145 inhabitants, unevenly distributed across the country, with a significant share living in rural and remote mountainous areas. In this context, ensuring equitable access to healthcare remains a persistent challenge. Rugged mountains, deep valleys, dispersed settlements, and limited transport connectivity significantly affect how easily people can reach medical services. Delays can mean the difference between timely treatment and preventable complications, particularly for maternal emergencies, acute infections, and chronic disease management.

To address this challenge, the European Space Agency, through its Global Development Assistance (GDA) Public Health activity, partnered with the Asian Development Bank, the Ministry of Health, Royal Government of Bhutan and Diginove to develop an Earth Observation (EO)-based service supporting telemedicine deployment. The initiative integrates satellite data, national datasets, and geospatial modelling to map population distribution, healthcare facilities, and travel times under realistic terrain and transport conditions. This spatial evidence base helps identify where telemedicine can most effectively reduce geographic barriers and strengthen connections between isolated communities, local health centres, and hospital-based specialists.

Close collaboration with ADB specialists and Bhutanese stakeholders ensured the EO products addressed operational needs, delivering validated, decision-ready insights for planning, resource allocation, and targeted digital health investments.

When Geography Shapes Access to Care

Bhutan’s healthcare system has expanded considerably, with over 240 health facilities distributed nationwide. However, physical geography continues to impose constraints, particularly in remote areas where settlements are located in steep terrain and often require walking long distances before reaching a road. Seasonal hazards such as landslides or snowfall can further disrupt connectivity. 

As a result, accessibility is not simply a matter of distance. Travel time is shaped by terrain, infrastructure, and available transport options. This is especially relevant for specialised hospital care, which is concentrated in fewer locations.

Figure SEQ Figure \* ARABIC 2: Health Facilities Distribution Map of Bhutan. The right panel illustrates the spatial distribution of the national healthcare network, classified by facility type: Primary Health Centres (PHCs), hospitals with 10 beds, and hospitals with more than 10 beds.

Why Earth Observation Matters

Earth Observation provides a consistent, nationwide evidence base that complements existing health information systems and other data sources. Satellite imagery enables spatial granularity for the identification of settlements, land cover classification, and population modelling in areas where field data collection is limited.

In the Bhutan case study, EO forms the backbone of an integrated analytical framework combining population estimation, terrain-aware accessibility modelling, and service delivery analysis at scales relevant to health planning. This approach highlights where healthcare access is strong, where it is constrained, and where telemedicine can provide the greatest benefit.

Turning Satellite Data into Health Intelligence

A) Mapping Where People Live

The first component focused on understanding population distribution. Using recent Copernicus imagery, settlement footprints were extracted and combined with open-source data and census totals through demographic modelling. This produced a spatially detailed population estimate for 2025, capturing both urban concentrations and dispersed rural communities.

Figure 3: Population Estimate and Density Map of Bhutan (2025). This map presents the estimated population distribution, derived from EO-based demographic model. Population density is expressed in persons per square kilometer. The color scale ranges from light yellow (low density) to dark purple (high density), clearly illustrating concentration gradients between rural and urban areas. The dashed rectangle marks the Thrimpu District, which is shown in the zoomed inset to provide a clearer view of densely populated settlements and spatial clustering patterns.

B) Modelling Realistic Access to healthcare

The second component assessed how easily populations can reach healthcare services under real-world conditions. The analysis used a hybrid travel scenario reflecting typical patterns in Bhutan: people walk across terrain to reach the nearest road, then continue using motorised transport.

Slope effects were incorporated to account for variations in walking speed due to terrain steepness. Water components such as rivers and lakes are considered impassable barriers for travel in both Walking and motorised modes. This ensures that travel times more accurately reflect the country’s landscape and mobility constraints.

Figure 4: Overview of land use in Bhutan for 2025. The land-use map classifies areas into water, forest, rangeland, cropland, flooded vegetation, bare soil, urban areas, and roads. It provides an overview of how natural and human landscapes are distributed across the country. Water components such as rivers and lakes are considered impassable.

C) Revealing the Gap between Basic and Specialised Care

The service distinguishes between two accessibility scenarios:

  • Access to all healthcare facilities (primary care network): gives a picture of how well Bhutan’s PMCs serves the population
  • Access to hospitals (specialised care): shows where people face greater difficulty reaching advanced or referral-level care.

This distinction highlights how primary healthcare centres help mitigate geographic barriers, while also revealing remaining challenges in accessing hospital-level services.

Key Insights: What the EO Analysis Reveals

A) Strong primary healthcare coverage: Most of the population can reach basic services within relatively short travel times, suggesting Bhutan’s primary care network effectively provides first-contact care and could serve as an anchor for telemedicine hubs.

B) Uneven access to hospital-level care: Remote mountainous regions experience significantly longer travel times to specialised services, creating risks for delayed diagnosis, emergency referrals, and continuity of care for chronic conditions requiring specialist management.

C) Spatial disparities across districts: Central areas benefit from better accessibility, while remote regions face greater constraints.

D) Priority communities identified at local level: Areas such as Lunana, Lauri, Laya, Serthig, and Tsamang show the greatest access limitations and are strong candidates for telemedicine deployment.

Figure 6: Districts for Telemedicine Deployment. Top panel: Least accessible districts for both scenarios (all facilities and hospitals only), showing the distribution of health facilities and population accessibility classes. Bottom panel: Geolocation of the least accessible blocks identified as suitable for telemedicine deployment. Values reflects the accessibility score of each bloc computed as weighted sum of population percentage per travel time range.

Targeting Telemedicine Where It Has the Greatest Impact

EO-based accessibility analysis enables targeted telemedicine planning by identifying areas where travel times to hospitals are highest. In these locations, telemedicine can strengthen connections between populations and health centres, as well as between primary health care providers and specialists, supporting diagnostic consultations (e.g., radiology reads, ECG interpretation), chronic disease monitoring (e.g., diabetes, hypertension follow-up), maternal and newborn care (antenatal consultations, postnatal check-ins), and clinical referral decisions – bringing specialist expertise closer to communities. This reduces the need for long-distance travel, along with associated transport-related emissions, while improving continuity of care and overall system efficiency.

Unlocking Broader Value for Health Planning

Beyond telemedicine, the EO-derived datasets support wider decision-making across sectors. They can inform infrastructure planning, resource allocation, referral system design, and emergency response strategies, particularly in remote and high-risk areas. 

Building Sustainable and Transferable Solutions

In Bhutan, EO services support the identification of priority areas for telemedicine under the ADB initiative, guiding investments where needs are greatest. Based on transparent methods and globally available datasets, the approach is transferable to similar contexts and can be updated over time as infrastructure and population patterns evolve.

Key Takeaways

  • ESA’s GDA Public Health programme, in collaboration with ADB, the Ministry of Health of Bhutan, and Diginove, developed EO-based services to support telemedicine planning.
  • The approach combines satellite-derived settlement mapping, population modelling, and realistic travel-time analysis.
  • Results confirm strong primary health care coverage but highlight persistent gaps in access to hospital-level services in remote regions.
  • District- and block-level analytics identify specific underserved areas, including Lunana, Lauri, Laya, Serthig, and Tsamang, as strong candidates for telemedicine deployment.
  • The methodology provides a scalable, updateable, and replicable approach that can support longer-term health planning in Bhutan and be adapted to other countries facing similar geographic constraints.

Acknowledgment

The present work was carried out in close collaboration with the ADB and Ministry of Health personnel. We acknowledge the Asian Development Bank team, including Yves Barthelemy, Sonalini Khetrapal, Romelei S. Camiling-Alfonso, and Jae Kyoun Kim, as well as Kinley Dorjee from the Ministry of Health,Bhutan for their collaboration and support throughout this case study.

Ali Ahmad
Ali Ahmad

Ali is the Chief Technology Officer (CTO) at Diginove Sas, where he leads initiatives in Earth observation (EO) and geospatial data analysis. He holds a Ph.D. in image processing, artificial intelligence, and numerical image simulation applied to microscopy from INSA Lyon, France (2021). His academic journey also includes an M.S. in Physics from the University of Angers (2015) and a degree in Biomedical Engineering from Lebanese University (2014).

With 8+ years of experience in artificial intelligence, computer vision, and data science, Ali has developed deep expertise in image and data processing, machine learning (ML/DL), vision systems, and numerical simulation. His career spans both industrial and academic research environments, enabling him to bridge cutting-edge research with practical applications.

At Diginove, Ali focuses on leveraging EO and geospatial data to address challenges in health and environmental monitoring, climate change assessment and impact analysis, as well as urban and regional planning and statistical modeling. His work lies at the intersection of technology and societal impact, using advanced analytics to transform EO data into insights for sustainable development.

You might also like