Synthetic Information Systems for Better Informing Public Health Policymakers

点击翻译按钮获取中文摘要

DESCRIPTION Our hypothesis is that synthetic information systems built by integrating relevant mathematical models can provide timely, comprehensive situational awareness and course-of-action analysis that policymakers can and will use to inform their response to infectious disease outbreaks. By synthetic information systems we mean software tools that synthesize diverse, seemingly incommensurate data, models, and causal hypotheses into plausible and justifiable pictures of a specific population and locality that support analysis of demographically and/or geographically targeted interventions. By comprehensive, we mean the tools include constraints and consequences due to behavior, sociology, logistics, and economics as well as health sciences. By provide and inform we mean that, rather than define studies and publish prescriptive policy guidance ourselves, we will create tools that allow analysts and other end users to explore policy and implementation options themselves. We will evaluate this hypothesis by tailoring to epidemiology our synthetic information technologies developed in a variety of decision-informatics contexts. We will extend these methods to address specific lessons learned during our efforts to engage policymakers in the 2009 influenza pandemic. Specific aims: 1. Create a synthetic information set tailored to infectious disease epidemiology that provides users distributional estimates of the health, social, and financial consequences of outbreaks and interventions in any target subpopulation. This includes designing and implementing a well-defined language for specifying outbreak and intervention scenarios flexibly, sophisticated models and simulations of disease spread, and methods for analyzing the resulting information. 2. Develop integrated dynamical models for individuals' behaviors relevant to the spread of disease and opinions (e.g. prevalence elasticity and sociological theories of complex contagion). 3. Compare the rankings of interventions given by compartmental and individual-based models. The comparison will trace differences in outcomes to specific differences between the models. 4. Conduct a comprehensive investigation of community-based, non-pharmaceutical interventions in an influenza outbreak. In the course of achieving these aims, we will introduce a formal mathematical treatment of multi- perspective, multi-theory, coupled network dynamical processes into epidemiology and epidemiological modeling.