Agent-Based Predictive Modelling for Humanitarian Epidemiological Response

  • Institution Country: U.S.A.
  • Implementation Country: Democratic Republic of the Congo
  • Sector: Life Saving Information
  • Funding Stage: Proof of Concept

Virtual Response Simulator: Agent-Based Predictive Modelling for Humanitarian Epidemiological Response

By President and Fellows of Harvard College through the Harvard T.H. Chan School of Public Health

Models of infectious disease outbreaks are essential to optimize health crisis response, especially in complex and conflict-affected environments. This informs pharmaceutical distribution, immunization campaigns, medical care deployment, security provisions, and more. While data-driven approaches are standard, they remain largely reactive, focusing on what happened and not taking advantage of newer, algorithmic methods that enable responders to instead ask what could happen and do so safely, cheaply, and effectively.

 

The Visual Response Simulator (ViRS) at the Harvard Humanitarian Initiative offers the means to virtually respond to infectious disease outbreaks and forecast the effects of epidemiological intervention strategies, with current focus on Ebola in the Democratic Republic of the Congo. Based on data from the 2013-14 West African Ebola Crisis, ViRS can rapidly project contingency plan scenarios and inform operational decision support at the speed, complexity, and contextual specificity required. Current forecast methods fail to fully inform operations on the ground as they often are too general and don’t account for the interdependent conflict variables. The most vulnerable, and hardest-to-reach populations in the northeastern Democratic Republic of the Congo are not only victims of the spread of Ebola but the ongoing conflict there that too often makes vaccination, treatment, and contact tracing impossible in these dangerous environments. The virtual response that modelling provides gives responders the means to conduct rapid, multi-level contingency planning that rigorously accounts for linkages between violent incidents and health impacts, bridging the gaps between health and conflict data for solutions that realistically address threat and logistical difficulty so that even the most vulnerable and hardest to reach are equally considered.