How can Non-Governmental Organisations (NGOs) and International Development Organisations (IDOs) ensure insight-driven response plans during Covid times?
5 recommendations to data-driven decision making
More than ever, the availability and optimisation of data is a primary need in a public-private ecosystem.
The COVID pandemic is accelerating innovation, galvanising the development of a raft of new and more pervasive public private partnerships between international organisations, governments, regulators, healthcare providers, the life science industry, tech and consumer health businesses. These collaborations take many forms, from R&D clusters, to healthcare delivery models, bringing stakeholders together like never before to tackle the pandemic, improve population health, and help economies to survive and ultimately thrive. These trust based partnerships have led to a wave of innovations and the sharing of knowledge and IP across value chains.
For instance, during summer of 2020, the WHO, together with Gavi, and the Coalition for Epidemic Preparedness Innovations (CEPI), launched the COVAX initiative to help build and upscale vaccine manufacturing and supply capabilities and provide countries worldwide with equitable access to two billion doses by the end of 2021. COVAX is one of three pillars of the Access to COVID. In addition, a private-public initiative was also launched to support Yemen as COVID-19 threatens to worsen the world’s worst humanitarian crisis. In the country about 24 million people need some form of humanitarian assistance and protection to survive. In April, a group of multinational companies and the United Nations have joined together to launch the International Initiative on COVID-19 in Yemen (IICY).
The stakes for global health are high: beside the direct effects of the COVID-19 pandemic, disruptions to supply chains are also compromising efforts to eradicate longstanding global health issues, such as Malaria, Tuberculosis (TB) and HIV.
Vincenzo Chiochia, Director Consulting
The pandemic has uncovered the fragility of the national health and related data systems. According to the Deloitte study “Future Unmasked”, by 2025 there will be a shift to prevention, including vaccines, genetic testing and digital therapeutics, enabled by radically interoperable data, advanced technology and analytics, secure open platforms; and new value based business models that deliver savings to the broader health system.
According to a recent report published by the Global Fund, about 60% of countries report moderate to high disruption to lab services for HIV and TB diagnostics. In the fights against HIV, community-based services are strongly impacted, with key populations most affected. Attendance in health facilities has decreased due to fears of COVID-19 infection and financial hardship associated with lockdown. Moreover, transportation issues prevent people from accessing services and hampers the transportation of samples to reference labs.
In this challenging context, the prompt availability and exchange of epidemiological data as well as data related to key medicine stocks and lab service availability is a primary need for International Organisations to act with adequate and insights-driven response plans.
How can International Donor Organisation prepare from a data and analytics perspective, in order to act adequately?
Here 5 concrete steps to accelerate the journey towards data-driven decision making in a public-private ecosystem of contributors:
Build and grow a data and analytics landscape in the Organisation having internal data exchange and interoperability with partners as guiding architectural principle. Modern data architectures start with the user needs and journeys, and build components such as data-hubs and data flows accordingly. In an International Organization context, users can be both internal (e.g. staff members in HQ as well as in the regional offices) as well as in partnering organizations, such as academic institutions, field operators and country level administrations. Therefore, modern data architectures should include solutions for user friendly data discovery, access and dissemination. For instance, data catalogues can improve the visibility of data assets across the organization and foster correlation analysis across various programmatic areas. In addition, catalogues give the opportunity to create custom data sets for dedicated needs under a well governed and controlled framework.
Promote metadata standards, e.g. via APIs and common data dictionaries: Data alone can be hard to interpret and correlate without a solid metadata description. Therefore, datasets should always be complemented with clear indicator definitions and descriptions of the statistical methodologies. In addition, as reproducibility and transparency of statistical estimates is key for scientific studies, organizations should be ready to provide solutions for sharing data reduction, aggregation and analysis techniques with the wider public. Finally, secure data exchange and data quality standards can be more effectively deployed and managed by means of Application Programming Interfaces (APIs). As the data exchange with third-party institutions increase and risks of data breaches intensify, organizations should avoid files exchange via unsecured communication channels such as unencrypted emails.
Leverage partnerships with cloud providers to accelerate the enablement of advanced analytics and AI capabilities at scale, as the technological landscape evolves rapidly. Cloud platforms nowadays can provide the building blocks for developing, deploying and manage machine learning models at scale and can accelerate the journey to AI adoption quite substantially. In addition, in the age of ever expanding dataset sizes and variable workloads, international organizations can benefit from secure and scalable architectures provided by cloud environments, that adapt to the data processing requirements on demand.
Foster collaboration across the ecosystem of partnerships by promoting open platforms for data modelling, predictive analytics and scenario planning. For example, organisations can accelerate collaboration with academia and private companies by providing cloud-based test and sandbox environments for secure and on-demand data analysis and prototyping, without adding extra loads on operational IT systems. For instance, data scientists may need access to disaggregated data in a staging area or in a dedicated sandbox where they can combine the organization data with third-party, economic and demographic datasets for scenario planning and modelling.
Complement the in-house talent for data analytics by creating broader partnerships with tech companies, professional service providers, academia as well as computational science centres of excellence. For insights-driven decision making, precise risk assessments and scenario planning, organization will increasingly need data explorers, data engineers and data scientists in addition to data consumer profiles, which may not be readily available within the institution.
International organization play a key role in protecting vulnerable populations at a very critical time. As the COVID-19 emergency compromises humanitarian efforts worldwide, decisions and response plans will increasingly leverage broader networks of partnerships involving the private sector, governments and academia. By having these 5 key aspects in mind as guiding principles, institutions can more effectively navigate the uncertain circumstances, minimizing the time between field data gathering and decision making in a rapidly evolving ecosystem of stakeholders.