Value of Information Analysis in Ebola Management

Open Access
Mummah, Riley O
Area of Honors:
Bachelor of Science
Document Type:
Thesis Supervisors:
  • Ottar N Bjornstad, Thesis Supervisor
  • Katriona Shea, Honors Advisor
  • Ebola
  • Value of Information
  • management
  • next generation methods
Over fifty mathematical models have been published to project the number of cases and the success of epidemic intervention strategies in the 2014 West African Ebola outbreak. Structural and parametric uncertainties exist between and within models, making it challenging to quantitatively rank the utility of intervention tools. Without a formal method of comparison, policymakers are forced to subjectively decide which models should dictate decision-making. Value of information (VOI) analysis, commonly used in economics and resource management, quantifies the extent to which decision-making is hampered by such uncertainties, thus providing guidance as to how to prioritize future information gathering. We focused on one of the most established models and performed a two scenario analysis to explore the uncertainties about hospital and funeral transmission and to demonstrate the utility of value of information analysis to address parametric uncertainties. We implemented a stochastic simulator using the Gillespie exact algorithm on a six compartment SEIHFR model introduced by Legrand et al (2007). We found that decreasing community transmission is universally most effective at minimizing case count and mortality. Public health policy in the 2014 Ebola outbreak should likely have targeted community transmission more intensely through awareness campaigns, distribution of household protective kits, and by encouraging self-quarantine.