COvid19 Network Technology based Responsive Action

The goal of the research project “COVID-19 Network Technology-based Responsive Action” (CONTRA), June 2020 – March 2022, was to contribute to the effectiveness, efficiency, equity, and sustainability of the COVID-19 vaccine distribution inside countries, with a focus on low- and middle-income contexts. The objective was to develop an intuitive web-based decision support system (DSS) to help public health authorities to allocate COVID-19 vaccines to different locations and groups of a country within a short time frame. To reach its objectives, the project followed five steps: (i) stakeholders analysis and COVID-19 supply chain system definition; (ii) identifying Key Performance Indicators (KPI’s) to assess the system, (iii) mathematical modelling, (iv) analyzing different realistic scenarios using the model, and (v) presenting the set of best-in-class allocation alternatives to public health authorities through an open source web-based platform. The outcome of the project can be summarized as follows. First, after studying the problem in a close collaboration with Norwegian public health authorities, it was found that the COVID-19 vaccine distribution inside countries could be investigated in central vs. local allocation levels. While the problem at central level is generalizable to many contexts (high, middle and low income countries) the local level problem found to be significantly context dependent. The central problem deals with distribution of vaccines from central to the regional storage facilities of the country after receiving (and repacking) them at the main entry points (like airports). The local problem covers receiving the vaccines at regional facilities and administering them in vaccine administration points to target groups. Given the focus of the project on low- and middle-income countries, and the access of project consortium members to key informants located mainly in high and higher middle income counties, the researchers targeted the central allocation problem. Second, setting the central level problem as the focus of CONTRA, we found that public health authorities followed achieving five key objectives in vaccine allocation despite limited vaccine supplies: vaccination as fast as possible, vaccinate as many as possible, vaccinate in line with priority and other guidelines set by the coordination body in charge (for instance, the Institute of Public Health in Norway), fair access, and public trust in vaccination. However, public health authorities experienced several challenges to achieve these objectives: extracting doses from vials, dealing with changing dose intervals, prioritizing patients, deciding whether to disposing a vaccine vs. using it, coordinating patients, vaccines and staff, establishing data systems for calling in patients and distributing vaccines, balancing the need for operational staff with the need to train staff. Our findings support identifying ways in which central authorities might promote a harmonic, coordinated national pandemic response while at the same time allowing for distributed improvisation. Third, to support public health authorities to achieve the above-mentioned objectives, specifically fair access to vaccines, the CONTRA team developed a novel mathematical model that could produce equitable allocation strategies within seconds. The model investigates equity through “fair coverage level”, which is calculated based on the size and the importance of multiple regions and eligible population groups. The results of testing the model on actual COVID-19 dataset from Turkey indicated that the factors used for the prioritization of regions may result in significantly different allocation decisions. We also found that the limited capacity of facilities at municipality and regional levels can significantly hamper equity in the whole vaccination system. As such, one recommendation was to develop a system that could incorporate several potential scenarios before or after presenting solution alternatives to public health authorities. Fourth, the CONTRA team developed a decision support system that incorporated the novel mathematical model into an online user-friendly dashboard to support vaccine allocation. Multiple validation meetings were held with representatives of public health authorities worldwide where the system was tested with different scenarios and datasets. The validation outcome revealed that vaccine allocation decision is indeed a complicated task, however, authorities often do not have sufficient time, resources and knowledge to incorporate sophisticated decision support tools. Thereby, allocation decisions have been criticized. Several tests of the developed dashboard at the CONTRA project showed the need for intelligent systems to support vaccine allocation problem in real operations was a realistic assumption. Based on the findings above, principal investigators at the CONTRA project have identified several future research directions.

For academic outcome, one conference paper and two open access articles have been presented and published in prestigious channels: the NOKOBIT 2021 conference, the Journal of Cognitive Engineering and Decision-Making (JCEDM) and the International Journal of Production Research (IJPR). The NOKOBIT paper justifies the roadmap of the decision support development in the CONTRA project in a scientific order and at the same time opens several research directions that are beyond the scope of the CONTRA project for future research. The presentation stimulated interesting Q&As among participants of the session. The JCEDM paper looks at how central authorities can support distributed improvisation, by studying the cognitive accounts of decision-makers involved in local COVID-19 vaccine roll-out in Norway. The study contributes to the literature by identifying ways in which central authorities might promote a harmonic, coordinated national response while at the same time allowing for distributed improvisation. The IJPR paper proposes a novel evidence-based mathematical model for equitable vaccine allocation in central level (from main storage facilities to regions in a given country) that accounts for multiple priority groups (e.g., elderly, healthcare workers), multiple vaccine types, and regional characteristics (e.g., storage capacities, infection risk levels). The study contributes to the literature by developing and validating a model that proposes equitable vaccine allocation alternatives in a very short time generalizable to different contexts and countries. Another article is in progress and will be submitted to a journal identified as level 2 in the Norwegian list of publication channels. Overall, several future research directions have been identified and thereby stated in the published articles and reports of the paper aiming to support scholars continuing the CONTRA project. For the next step of the CONTRA project, three principal investigators (PIs) of the consortium with two new partners from Eastern African countries (EAC) have prepared and submitted the Vac4All project proposal for the call Researcher Project for Scientific Renewal 2022 theme Global Health. The proposal targets the vaccine distribution problem in local levels at lower income countries. Apart from academic presentations and publications, several workshops and meetings have been held over the course of the project were the consortium's PIs discussed the project results with practitioners and policy makers involved in vaccine distribution from Norway, Belgium, and EAC. Participants have been informed about the development of the project and were given the opportunity to test and adapt the developed decision support system for their operations. More specifically, the CONTRA dashboard has been presented to including the U.S. CDC and the African CDC. The latter, African CDC is evaluating how they can incorporate the dashboard in their activities. The dashboard is open access.

Pandemic responders lack a decision support system (DSS) that realizes the characteristics of the vaccine supply chain problem: (a) the DSS must allow balancing different stakeholder objectives related to effectiveness, efficiency, equity, and sustainability; (b) the DSS should satisfy different objectives throughout a planning horizon; (c) the DSS should account for uncertainties in demand and supply; (d) the DSS must address the critical concerns of different stakeholders; and (e) the DSS should consider the dynamic social and technical components of the situation. The CONTRA project aims to develop a DSS for designing an effective, efficient, fair, and sustainable in-country vaccine supply chain while considering uncertainties along with social and technical components. The DSS will be developed based on mathematical modeling and stochastic optimization (for supply uncertainty) as well as machine learning tools (for demand uncertainty) for designing a robust vaccine supply chain. The main focus of the research is on supporting responders in low-income countries to pandemics, such as COVID-19, and their needs to make informed decisions when designing in-country networks for vaccine supply chains. The project will conduct rapid analyses that will provide actionable advice regarding vaccine distribution and delivery to responders in Norway, including crisis managers, and logisticians responding to the COVID-19 epidemic, in response to the on-going COVID-19 outbreak. To accomplish this, the CONTRA project will use the experience of research on developing decision support systems for vaccine supply chains in low-income countries, although the limitations and generalizability of the approach will be considered. The target group of the project is responders and decision-making stakeholders in low-, middle-, and high-income countries required to make decisions for managing vaccine supply chains in response to epidemics and pandemics, and their information needs.