FCV Health Knowledge Notes Digital Technology in Forced Displacement Settings: Middle East and North Africa (MENA) Case Studies SUMMARY A wide range of new technologies are progressively being used for data collection, project planning, implementation, and monitoring and evaluation in forced displacement contexts. This note highlights case studies from countries in the Middle East and North Africa Region and provides practitioners with specific examples of technology use. Why is digital technology useful in forced displacement Lebanon have been prone to more fragility because of an influx of refugee populations from neighboring countries environments? and the strain experienced by basic service delivery systems, including health. According to the UNHCR, 70.8 million people have been FCV settings in general often pose challenges to both local forcibly displaced worldwide, often due to conflict and health systems and service providers that deliver needed violence. While it is estimated that 90% of the disease health services. As in the case of many MENA FCV burden among forcibly displaced populations comes from affected countries, conflicts lead to forced displacement, communicable diseases, non-communicable disease and inaccessible areas, infrastructure damage, and unique mental illness are also prevalent among this population. health challenges. For example, in Yemen, war-related damage to health facilities, water and sanitation However, access to healthcare in host countries is often infrastructure have led to 18 million people lacking access limited due to lack of entitlement to basic services based to basic healthcare and over one million cases of cholera on a migrant’s legal status in the host country. As a result, in the past year. both development and humanitarian partners are focusing on improving access to health services through innovative For World Bank teams, FCV environments add more service delivery mechanisms. To address the needs, the constraints to planning, implementing, monitoring and World Bank Group has allocated US$2 billion for the IDA18 evaluating projects. Specifically, given the unique regional sub-window for refugees and host communities. In challenges of forced displacement, the teams are required addition, the Health, Nutrition and Population Global to understand the following: Practice of the World Bank Group currently implements 34 projects totaling over US$2.8 billion in fragility, conflict and (1) What health services are most necessary and violence (FCV)-affected settings. where they are needed; (2) How key health services can be delivered to The Middle East and North Africa (MENA) region has a those in need; and variety of countries affected by FCV. Some are in active (3) How teams can monitor and evaluate the services conflict like Libya and Yemen, while others are being delivered to target populations (i.e. forcibly experiencing protracted crises like Syria. Consequently, displaced populations). non-conflict countries in the region, like Djibouti and 1 Digital technology has proven to be useful and effective in addressing the numerous challenges in FCV settings. It allows teams to plan, implement and monitor tasks in remote and/or insecure environments. As a complement to the previous publication by the World Bank Group on the use of technology in FCV situations (Using Technology in Fragile, Conflict, and Violence Situations: Five Key Questions to be Answered), this note highlights operational examples and lessons-learned from four countries: Djibouti, Lebanon, Libya and Yemen. The purpose of this note is to showcase their successful experiences in integrating digital technology into health operations. The note also explores the risks and mitigation measures that are essential in using digital technology in FCV settings. What FCV specific challenges can be addressed by using digital technologies? Assessing Population posts throughout the country. Not only did it provide a 1 Movement and Settlements: location for all health facilities, but it allowed the team to determine catchment areas for each health facility. Djibouti Determining the catchment area is important as it allows health workers to create zones which help them to better Djibouti, one of the smallest countries in the MENA region, understand the populations in their areas and the services has an estimated 100,000 migrants in transit every year that the population receives. The catchment areas are also and hosts about 30,000 refugees from many of its useful to mobile clinic drivers in helping to direct them to neighboring countries, including Yemen and Somalia. underserved populations. Additionally, the country is home to a traditional nomadic population. Since there had previously been no The team was also able to monitor the number of rooftops disaggregated population data, one of the challenges faced in a particular service delivery zone based on the by Djibouti’s Ministry of Health, and health service catchment areas. Using technology, relevant imagery was providers, was determining where health services were broken into pieces and the probability of a building being needed as populations traveled. present was calculated and assigned as a score. Using this method, both permanent structures and settlements were The World Bank’s Towards Zero Stunting nutrition-focused detected. With this innovation, nomadic and migrant project in Djibouti addressed this challenge through the use populations that were difficult to track could now be of geospatial mapping, satellite imagery, and machine matched to nearby health facilities. Further, the innovation learning. The technology allowed the project’s team to helped the World Bank team to target certain areas which determine the location of 24 primary, secondary and held populations without nearby health and nutrition tertiary health facilities as well as 33 rural community health services. Satellite imagery and machine learning used to detect rooftops, designated by the green outlines, in Djibouti. Left: An urban area in Djibouti. Right: A rural area in Djibouti. 2 Delivery of Services and facility surveys. While the SARA was used as a basis to 2 Infrastructure Assessment: locate facilities, satellite imagery was used to identify the accessibility of facilities, by taking overhead pictures of Libya roads, entries to facilities, etc. Additionally, social media sites, like Facebook, were then used to analyze facility Once mobile populations have been identified, it is activity and to investigate complaints, helping the team to important to determine the services to which they have determine services available at different facilities. Facility access. Libya, a country in active conflict, not only has surveys of both patients and providers were done in public internal displacement, but an estimated 700,000 or more health facilities by teams with tablets that could upload data migrants coming from neighboring countries. In this setting, to a server for immediate availability. it was difficult for the World Bank team to determine which facilities were available to provide specific services since With these technologies, the team was able to map the many facilities had broken equipment, were damaged, or location of every public health facility in these cities, as well were in inaccessible areas controlled by various militias. as the services provided in these facilities. This helped the team to determine which populations had access to To determine the services available in two of Libya’s major particular services, such as antenatal care, immunizations, cities, Tripoli and Benghazi, the World Bank team or other specialized care. Determining the location and triangulated data from, satellite imagery and social media caliber of services in each area will allow the team to assist analysis, in conjunction with using more traditional data the Ministry of Health in determining where to allocate from the World Health Organization (WHO) Service funds for populations in need and where to scale up Availability and Readiness Assessment (SARA) and health services for those populations. Map of Tripoli, Libya showing all health care facilities. At each facility, the team was able to determine the type of services available. 3 Monitoring and 3 Evaluation: Yemen and Lebanon The fluidity and political instability that exists in Yemen, an active conflict country, renders traditional monitoring and evaluation mechanisms nearly impossible to implement when providing oversight to existing health services. Yemen is also different from other contexts in that the World Bank does not have any staff in the country and must partner with UN agencies like WHO and the United Nations Children’s Fund (UNICEF) to implement interventions on the ground. The challenges in this setting have afforded the World Bank and its partners the opportunity to use unique digital technologies for the oversight and monitoring of Lebanon has a different context in which the health system interventions. is not under strain because of its own conflict, but because of the conflict of neighboring Syria. Due to the influx of In Yemen, fuel is provided to hospitals as an energy refugees from Syria, it estimated that Lebanon’s population source. It is important to ensure that the fuel is in fact being grew 30 percent in just six years. The strain that this has used for hospitals and not redirected to militias or other placed on the health system not only makes the scale up causes. This need for monitoring led to the use of fuel of services necessary, but makes it essential to have a sensors. The fuel sensors, located in the fuel tanks, have proper grievance mechanism in place to monitor the the ability to upload real-time data to a web-based server, effectiveness of the services being provided. To address indicating the exact amount of fuel in the tank. This lets the this concern, the World Bank team supported the creation monitor ensure that the fuel levels coming from the tank of an automated module in which one platform was created are appropriate for a health facility and do not indicate to track and manage grievances. The platform pulls from evidence of stealing, redirecting fuel, or other related registration databases and other sources. The grievances corruption. Another similar technology is being developed are then uploaded to the Ministry of Public Health web- for fuel trucks in Yemen. These trucks will also include a based server for timely response. The method of pulling GPS tracking system that will upload the coordinates of the grievance data from multiple sources is considered to be a fuel trucks at all times. global standard. Fuel sensors are used in Yemen to monitor levels in oil tanks used to service health facilities. What are the challenges and risks to consider when using digital technologies? The primary challenge to using digital technology is that the World Bank teams are tasked with trusting information, often in situations that they cannot access. For this reason, it is always important to triangulate data from multiple sources to verify information and to ensure the validity of the data being collected. In the Libya example, the team found that there were facilities that technically showed the availability of certain services after using one source, but many of these facilities proved to be inaccessible based on satellite imagery data. This emphasizes the need to triangulate multiple data sources, particularly in FCV and forced displacement settings. Similarly, in the Djibouti example, to avoid the risk of bias from one data source, more than one technology was used to triangulate the location of mobile populations. While one technological method can be used as a base, it is important to triangulate with multiple sources for the best data verification. Additionally, project implementors should note that technological solutions cannot work in silos, but must work within the framework of a larger system in order to be effective. For example, in Yemen, the fuel sensors worked well because of the presence of a web-based server that allows remote monitoring of the system. In some countries, the social media analysis in Libya would be ineffective due to the lack of internet access. Further, it is important to note the legal framework of each country to determine what technologies and use of information is permitted prior to engaging in technological solutions. *Source: Using Technology in Fragile, Conflict, and Violence (FCV) Situations: Five key questions to be answered. 4 TABLE 1 Risks, Challenges, and Mitigating Measures for Using Technology Risk Country Technology Used Challenge Mitigation Measures  Using one data source alone introduced bias. For example,  Used multiple technologies to  Social media using social media analysis by triangulate data. analysis itself has the possibility of  For example, in addition to social Bias Libya  Satellite imagery introducing bias since only a media analysis, satellite imagery was  Facility surveys portion of the population has used to show if there was road with tablet access to internet and social access or damage to facilities. media sites.  Triangulation was used so that data was coming in from different sources to verify accuracy.  The satellite imagery showed  Satellite imagery  Using one data source alone Verification Djibouti buildings, but did not show where  Machine learning proved to be inaccurate. settlements of migrants were created.  The machine learning combined with satellite imagery showed where migrant settlements were located.  Areas where ground teams were collecting survey data with  The ground survey teams were given  Tablets uploading tablets was not always safe due letters of approval by the government Safety Libya to web-based to conflict. to show credibility in the field. server.  Surveyors were at risk for not  Teams pulled back from areas where always being trusted by staff conflict arose. and patients at the facilities.  The World Bank team ensured that all  Social media data used was publicly available. For  Adhering to consumer privacy Privacy Libya analysis example, social media posts and standards.  Satellite imagery satellite images are available for public consumption.  Social media analysis  In all cases, the World Bank used Djibouti,  Satellite imagery transparency with the governments to  Governments and beneficiaries Libya,  Facility surveys ensure that all data and processes Perception may perceive technological Yemen, with tablet were made available, so that there innovations as privacy invasion. Lebanon  Fuel sensors were no perceptions of spying or  Automated obtaining unauthorized data access. grievance system  Target populations move, so the  This requires a GIS information officer  Satellite imagery data on beneficiary locations to manage the incoming data and Sustainability Djibouti  Machine learning and catchment areas must be generate new maps as populations constantly updated. move.  Djibouti’s rooftop mapping costed USD$15,000 for the entire country. The cost here can vary based on the  Rooftop mapping size area that needs to be mapped. Djibouti,  Geospatial  Libya’s geospatial mapping and social Cost  Technology costs can vary. Libya mapping and social media analysis costed USD$50,000 media analysis to map the two major cities, Tripoli and Benghazi. This cost will vary based on the size of the area to be mapped. Authors: Sheila Carrette, Health Consultant, Health, Nutrition and Population Global Practice, World Bank Group Takahiro Hasumi, Health Specialist, Health, Nutrition and Population Global Practice, World Bank Group For more information on other HNP topics, go to www.worldbank.org/health 5