Pandemic Data Interoperability Domain

Ignacio has a wide variety of experience in engineering telecommunications projects in developing countries, which derives from his background as responsible of telemedicine projects in the NGO field, and researcher/teacher in the university. Prior to his current position in EyeSeeTea, he has been in charge of interdisciplinary projects for creating robust and efficient systems and infrastructures for health and telecommunications in connectivity-challenging environments. After a 3-years experience in the biocomputing center of the national biotechnology center of Spain, he dedicates his efforts to build EyeSeeTea as a software developer and director. Now EyeSeeTea is a reference in the epidemiological data with DHIS2 systems arena, working actively with big organisations like WHO, PSI, NRC or Medecins Sans Frontières making possible the evidence-based data-driven decisions in real-time.

Well-functioning health information systems are crucial for detecting and responding to pandemics such as COVID-19. Rapid decision-making is required within and across countries based on real-time data across disparate sources. Most digital systems used for COVID-19 cannot interact at present given they were not originally designed for bi-directional data exchange and compatibility Interoperability across digital platforms for COVID-19 must become a priority to ensure data-based decision making can become a reality.

We propose a hub-and-spoke interoperability layer to allow data exchange across systems related to COVID-19 case detection, surveillance and contact tracing. A common standard data structure modelling the minimum information needed to control pandemics, easy to translate for and from. The level of effort is then exponentially reduced and the data would be fully interoperable.

Pandemics are to be more frequents. A real-time data exchange between all systems would give humanity a chance to save millions of lives.

With over 12M cases recorded since the start of the COVID-19 pandemic, countries are searching rapidly for solutions to address the toll that COVID-19 will wage on their populations. A key component of a comprehensive strategy to control the spread of COVID-19 includes effective contact tracing and quarantine of confirmed cases. The information that local contact tracing teams gather should be entered into a platform including data on source case, lab data, and information on contact status over time in such a way that the correct epidemiological scenario is presented. The Global Outbreak Alert and Response Network (GOARN) has developed Go.Data to manage such case-contact relationships and track the follow-ups of contacts over time. 

Through a global expert consultation on COVID-19 contact tracing hosted by WHO in June 2020 which reviewed the use of Go.Data along with a range of other software tools for contact tracing and outbreak investigation, many countries and public health stakeholders identified the lack of interoperability across tools as a major gap in effective response activities. 

Developing a mechanism for existing data systems to exchange important information is thus been stated as a major priority for WHO in the next phase of the COVID-response.

The proposed project will create a centralized hub-and-spoke interoperability layer to allow Go.Data to interact with a range of other software tools used for COVID-19 case and contact surveillance. This is highly desirable over the current approach of developing bi-directional solutions for each specific software that approaches WHO for data exchange with the Go.Data tool (now used in over 45 countries)

It is not the first time the interoperability of public health digital systems has been addressed. A good example is the FHIR is a standard for health care data exchange, published by HL7. Beyond Go.Data, this solution proposes an open source Minimum Viable Product for pandemics interoperability need to get the exact information that would make the control of the pandemic possible, defining a common data model domain so that information can be translated to and from every system. 

We propose the translation for some of the most popular systems, to demonstrate the effectiveness of this approach (for example, Go.Data, DHIS2, EWARS, OpenMRS) and creation of an API standalone server able to write and read from this data model, to ease the implementation for other systems.

This project will greatly improve the workflow of public health professionals globally who are currently collecting and monitoring case and contact data in order to better respond to the COVID-19 pandemic in their setting. It will also impact Ministries of Health and other public health institutes by reducing fragmentation in the health information system and allowing for seamless data exchange across existing platforms.

In addition, national and local decision-makers will be direct beneficiaries by being better equipped to make evidence-based policy decisions on response activities and resource allocation with more timely and high quality epidemiological data

Finally, the general public, all of whom are affected by COVID-19,more robust and targeted public health policies and programming with the potential to save lives and prevent spread.

Pandemics are one of the most difficult problems our world can face. Politics, economy and global solidarity are challenged with a short timeline to react. We have seen it with COVID-19. Lockdowns vs economic depression, vaccines, test and treatments for all vs guaranteeing own population health security...the list of difficult decisions is long. Taking these decisions without the appropriate data, and without crossing that data with all the available one from the rest of the countries makes each second with a better knowledge of the situation massively valuable. Interoperability between systems is means more time for knowledge building.

For over 2 years, EyeSeeTea has vast experience with public health information systems and how they can be maintained most effectively. EST have had extensive and long-term collaborations with several departments at WHO that register their epidemiological data using DHIS2 in a platform called WHO Integrated Data Platform (WIDP). We also hold partnerships with the Polytecnic University of Catalonia (UPC) that are designing the concept of a data lake of information where different information systems could coexist. It’s inside this solid team that the idea of interoperability minimum viable product became a need. We first worked on solutions that could make the same systems with different metadata interoperable, and then moved to the most complicated scenario, using different systems. With COVID-19, the interoperability is essential. Different one-to-one interoperability projects have been addressed in an ad-hoc fashion necessary given the time restrictions in a quickly escalating public health crisis. The Global Outbreak Alert and Response Network (GOARN) has developed Go.Data, a software specifically designed to manage case-contact relationships tracking the follow-ups of contacts. It became a high priority to collaborate with WHO/GOARN and Go.Data project team to establish stronger interoperability solutions between Go.Data and other COVID-19 platforms.

In EyeSeeTea we are passionate about global goods, open standards and open-source software. Being a company, we are conscious that working only with open source and making our work immediately available for the public is not common, but is strongly needed when working in challenging scenarios. We are a social-committed enterprise, and our goal is to improve the health processes with what we know best: Information Technologies. We have been working with health systems for years now, and this experience has led us through the heterogeneous world that we live in. With a company board made of NGO-related and research people, we love technical challenges and are very proud of providing not only effective solutions, but well-designed from a software architectural point of view. The essence of this problem, the concept that we would like to develop has many things in common with the architectural approaches we use in our solutions. Discovering that our architectural approaches for the IT solutions could be extrapolated to contribute to the mitigation of COVID-19 has been extremely encouraging for us. 

Several people and organizations will benefit from this in the future, and we will have definitely left a good durable footprint in this world.

As a consolidated software development company, with experience in interoperability between systems, particularly for the epidemiological domain, our background is exactly what this project would need, having in the team all technical skills that this project will need. That said, our strength is not only our sole strength, but particularly the powerful team that we integrate with our partners. In the last year EST has been working on the DHIS2 interoperability issues enabling smooth communication and metadata/data transfer among different DHIS2 instances. Polytechnic University of Catalonia (UPC) with its Database Technologies and Information Management (DTIM) group is working on innovative research topics related but not limited to databases and Big Data management. DTIM has been working the last 5 years on the WISCENTD project together with the Neglected Tropical Disease (NTD) department at WHO, with the goal of integrating data about NTDs from a variety of data sources and enable integrated analysis of NTD data to the epidemiologists and statisticians at WHO. Recently, UPC has been granted an emergency project to help developing countries in Africa in fighting against COVID-19 by providing interoperability between outbreak information systems and countries’ national health information systems. UPC is actively collaborating with GOARN and WHO team in this COVID-19 interoperability project. As the last but not least team mates, GOARN with all their experience managing the COVID-19 data during this pandemic provide a valuable knowledge impossible to replace. There is no need to elaborate on WHO's knowledge and experience in the pandemics area.

2 years ago a big organisation we collaborated with initiated the process to block our participation in an international congress about an open-source framework that was supposed to drastically reduce the cost of Android development for epidemiological data with DHIS2, making it affordable for NGOs with little software development budgets. The origin was a misunderstanding in the goal of our presentation. They were wrongly advised to think our presentation would reveal key aspects of their business model. The event organizers received its pressure to block our presentation. 

To avoid our disparition in an ocean of attorneys, we negotiated a presentation with them that could guarantee our presence in the event (sharing all our materials), while we decided to artificially reduce our prices in order to engage new clients in the event, reducing the high level of dependency with that organization. Due to the high risk of that approach, the strategy was agreed with all people in the company. From that conference, we returned back to our headquarters having 2 new clients, that quickly replaced most of our activity with the big organization, introducing more diversification to our clients base, and opening to us new horizons that we are yet exploring.

In EyeSeeTea we strongly believe in team strength. The leader is always somebody who takes the responsibility of having the big picture in mind, making the right questions, explaining as clear and transparent as possible the situation to the team, acting as speaker of the collective intelligence the team represents.

A situation where leadership plays an important role is when the coordination between different teams, actors and developers, is challenging. 

WIDP is a DHIS2-based information system in WHO, shared by 6 different WHO groups to register, analyze and represent countries' data. EyeSeeTea coordinates the IT team implementing the maintenance, which is also integrated by other internal staff from WHO NTDs department and the UPC. In addition, the infrastructure where WIDP is hosted is managed by the IT department of WHO, which means that EyeSeeTea doesn’t have permissions on some of the basic infrastructure levels. The challenge in this coordination has to do with dealing both with 6 different units and other external people in the IT team working on a platform hosted outside our full control, but keeping a high level of stability. We are proud of the progress made, joining together those efforts and interests in a complicated environment.

EyeSeeTea is a small for-profit company base in Spain

Polytechnic University of Catalonia (UPC) is a public university in Spain. 

This project will work directly with focal points on the WHO/GOARN Go.Data team, based at WHO HQ in Geneva.EST and UPC are working collaboratively for last 2 years on the development and maintenance of the WHO Integrated Data Platform (WIDP) for WHO.
UPC and WHO/GOARNare collaboratively working on a project to enabling interoperability of Go.Data and other national health information systems in under-developed and developing countries in Africa.

At this moment there are specific interoperability projects in the market, implementing the compatibility one-to-one (e.g. the interoperability that UPC is currently building between DHIS2 and Go.Data). Until now there is not such a technology that makes possible to intercommunicate and exchange data between all systems at once. What makes this project innovative is the approach, trying to concentrate on a minimum set of information. Up to now, all the interoperability efforts, including FHIR specification, had tried to define it mapping all that could be mapped, representing all that could be represented and, hence, creating a generic solution. What we propose here is a pandemic-centered solution, which will only understand about a certain subset of concepts whose transmission save lives. The innovation of this, is that even when mapping the data models of 2 different servers might sometimes even be impossible, because they are not thought to solve the same problems, If you just want to track cases, deaths, contacts and their relationships then the mapping problem becomes much easier to solve.

WIDP is currently in use to collect country data and facilitate data use across the following domains:

• Malaria epidemiology and entomology

• Hepatitis

• Neglected tropical diseases (buruli ulcer, chagas, dengue, dracunculiasis, echinococcosis, yaws, foodborne trematodiases, human african trypanosomiasis, leishmaniasis, leprosy, lymphatic filariasis, onchocerciasis, rabies, schistosomiasis, soil-transmitted helminthiasis, taeniasis/cysticercosis, trachoma, chromoblastomycosis and other deep mycoses, scabies, snakebite envenoming)

• Emergency, trauma and acute care 

• Health workforce 

All these impacted population could easily be responding to any pandemic just by having an specific DHIS2 package for it, as it exist right now for COVID-19 (https://www.dhis2.org/covid-19). We could estimate the number of people by counting the population of the reporting countries to any of those health areas. 

Go.Data tool is being used in: 46 countries 

In 5 years, interoperability with most of the systems in use all around the world could be already working, so virtually everyone in the world could be covered in one way or another by the impact of this project.

Next year Go.Data + Digital data collection tools interoperability (such as CommCare). 

The idea for the next year is to build compatibility with Go.Data and Digital data collection tools, writing this code in these systems programming languages (JS + Python) and defining the common domain of information that every system would need to translate to. Only those entities of information that are absolutely necessary from an epidemiological point of view will be defined. During this time we wil also create the website and define the common data model that will represent the domain.

Next five years: EWARS + OpenMRS + Intermediate API RESTful server + Clear documentation on how to build interoperability by using any other programming language. EWARS (Early Warning, Alert and Response System) is designed by WHO to respond to emergencies. Pandemics information can also be extracted from other emergencies, so integrating this system can be crucial, as well as OpenMRS, that has a more facility-centered approach, where we can also extract very valuable information about pandemics. In order to make other systems able to contribute and be contributed by this interoperability an intermediate standalone server with no graphic interface but a RESTful API will be built. The idea of this API server is to expose an implementation of the proposed common data model and an easy way to modify it via a RESTful API, so a quick integration could be very easily built with no matter which system.

The barriers to build this interoperability are mainly the following ones:

• Financial barriers: Even when following a hub-approach instead of a one-to-one approach for the interoperability means distributing the effort so every system can take care of their translation to the common data model, building all this interoperability layer and defining the essential data model implies some specific work that needs to be funded.

• Political / market barriers: Facilitating the interoperability will help everyone and all servers. Limiting the transmitted information to the minimum useful for a pandemic control will also avoid systems to see the interoperability as a way for smaller servers to use their information and gain a better position in their market, but obviously those who has a bigger part of the market risk always to see any communication with other systems as a threat. Explaining clearly the purpose of this interoperability for pandemic layer is crucial for it to be accepted and integrated in all systems, and the biggest actors must play the game for it to work.

The strategy to overcome these barriers comes from the global strategy for designing the solution and vice versa. Making this project a global good, that would be guided by an international organization like WHO avoid the suspicion of any company willing to push this for their profit, which will make it easier for others to adopt the solution. This will probably reduce the size of the market barrier. Additionally, the main actors of the digital health must be targeted and involved in a steering committee that might

We collaborate with some WHO departments already in this WHO Integrated Data Platform, as the maintainers of their digital platform to register epidemiological data, developing software to complete the platform functionality, modelling metadata for the platform and in general keeping the platform working. In this maintenance we partner with UPC, which also are working on other projects aiming to facilitate the data entry/analysis with a platform-independent approach.

This close collaboration, in COVID-19 times, has lead some members of this team to integrate the WHO GOARN network in order to help in COVID-19 response, where many of the main problems dealing with non-interoperable data arose. In the framework of this situation, the WIDP maintenance team has done short contributions to GOARN by creating scripts to translate certain metadata objects from DHIS2 and the global-boundaries WHO initiative to Go.Data systems. 

Lastly, UPC has a long collaboration with Neglected Tropical Diseases (NTD) program at WHO in developing data-driven solutions for the collection and analysis of NTD data. Both UPC and EST have had successful R&D collaboration with the Doctors Without Borders (MSF) team in Barcelona. Moreover, UPC have had successful collaboration with other organizations and companies from the healthcare domain like Probitas (Grifols). 

In summary project organizations already cultivate a strong network of partners in the healthcare and medical area which will be crucial for the development and future sustainability of the project. 

The business model of this idea is to be an initiative maintained by the international public health sector, represented by WHO. All platforms are interested in contributing to pandemic responses like COVID-19. Making the barrier of that contribution as lower as possible will encourage the companies/public institutions behind those systems to adapt them for the common domain we are proposing. As an open-source data structure model, it won’t need very big investments after the initial development is finished.

The PATH for the financial sustainability would be:

1. To find the funding to finish the pandemic-centered data domain design between epidemiologists and computer scientists

2. Implement interoperability with the main actors where the data is sitting

3. Engage those actors via a steering committee for pandemic-data compatibility

4. Offer the compatibility services to WHO as a resource of their property and look for its maintenance to be included in its budget

UPC’s Center for Cooperation Development - grant - 16.581,60$ (COVID-19 interoperability project for UPC)

We are looking for funds in the form of grants to fill the initial gap of research, design and software implementation, to take the project to the point that WHO could take the relay. The funding we are looking for would imply the following working packages, activities and estimations of budget:

• Working Package 1: Research. Leaders of the WP1 UPC (~ $80k)

  • Post-doc research activity on pandemic-centered domain definition and theoretical framework 

• Working Package 2: Data model design. Leaders of the WP2 GOARN (~$25k)

  • Expert group creation expenses (meetings-related expenses)

  • Contribution and collective design between epidemiologists and IT staff 

• Working Package 3: Development. Leaders of the WP3 EyeSeeTea (~$195k)

  • Project website creation and maintenance for 5 years

  • Implementation of interoperability layers for Go.Data, DHIS2, EWARS, OpenMRS

  • Implementation of standalone domain translation RESTful API

  • Transfer to WHO documentation (videos, technical documentation, other resources)

For 2020 the needs would be as follows:

• WP1: Research: ~$8.5k

  • Post-doc research for 3 months

• WP2: Data model design: ~$8k

  • Meetings and collaborations between epidemiologists and IT staff in order to propose the epidemiological concepts to be covered by the model

• WP3: Dev: ~$8.5k

  • Project website creation and dissemination of initial results.

In order to overcome the financial barrier we are identifying the Elevate Prize as a very good opportunity

The interoperability means to exchange very important data between systems and countries, which might imply to resolv many legal and regulatory issues. We would like to receive advise on this part.

Likewise, in the media and exposure side, we would be interested in receiving some guidance, in order to reach all the organization developing the targeted systems

Our main partnership to envisage is WHO as organization, as well as ministries of health or organisations appointed by ministries of health. The project already counts on WHO departments and networks like GOARN, but WHO as the full organization would be the key partner to make the project sustainable in the 5 years+ future.