Integrating Exposure and Epidemiologic Models to Select Nonpharmaceutical Interventions for Influenza

Abstract

PP-29-113 Background/Aims: Traditionally, the effectiveness of hygiene interventions—-hand washing, respirator use, and cough etiquette—-have been explored through mechanistic exposure models within a quantitative risk assessment framework; whereas the effectiveness of social distancing interventions has been explored using epidemiologic models. We present an approach, inspired by Larson (2007), in which the transition from Susceptible to Infected group in the SIR epidemiologic model is based on infection probabilities determined by exposure/risk models, rather than a reproductive number. Methods: Monte Carlo simulation of a Markov model of influenza exposure and infection risk provides estimates of infection probabilities for selected contact scenarios. Hygiene interventions are included in the exposure/risk model, altering the infection probabilities. The infection probabilities define the transition from Susceptible to Infection in the standardized incidence ratios epidemiologic model. For illustration purposes, we use 2 population strata, with high and low contact rates, respectively; and model the impact of social distancing interventions by modifying the number of people in each strata, and their contact rates. Given repeated model simulation, the epidemic dynamics resulting from different interventions and compliance rates, the can be compared to select the intervention(s) which minimize costs and maximize health risks, given uncertainties about influenza exposure and infection risk. Results: This new methodological approach will be illustrated using a simple scenario. Conclusion: The advantages of this approach are the direct comparison of hygiene and social distancing interventions, and explicit treatment of influenza as an environmentally mediated disease, for which the risk of infection is highly dependent upon the type and magnitude of contact between susceptible and infectious persons.