SCIENTIFIC PROGRAMS AND ACTIVITIES

Jonathan Wang (University of Toronto)
A real-time prediction model for aid in hospital-level pandemic influenza planning

In light of the recent H1N1 pandemic in 2009, there has been an increased interest in using mathematical modelling for public health decision making. However, there is very little research into using mathematical modelling to estimate the impact that the pandemic will have on hospital resources. One such model in the literature is FluSurge (http://www.cdc.gov/flu/tools/flusurge). A model developed by the CDC in 2006, it allows for an estimation of the number of incoming patients into a hospital based on historical data and from that estimation, assesses the adequacy of the existing resources in the hospital to meet the demand.
I will be addressing some of the shortcomings of FluSurge as well as proposing a novel model that improves upon the aforementioned shortcomings. Some of these improvements include: integrating real-time hospitalization data to generate a prediction of pandemic characteristics, changing the FluSurge algorithm to utilize epidemiological theory to model hospitalization data, and developing the granularity of the model parameters for the user to manipulate.
The goal of this modelling exercise is to develop a model that balances the accuracy of the algorithm with the simplicity of the design to facilitate integration and use in a real-time hospital setting.

POSTPONED

May 13, 2011
11:30 am
Stewart Library

POSTPONED

Mar. 25, 2011
11:30 am
Stewart Library

David Fisman (University of Toronto)
Cholera in Haiti: Insights from a Simple Gravity Model

Background: Haiti is in the midst of a cholera epidemic. Surveillance data for formulation of models of this epidemic are limited, but such models can aid understanding of epidemic processes and help define control strategies.

Objective: We used a mathematical model to predict the sequence and timing of regional cholera epidemics in Haiti and explore the potential impacts of control strategies.

Design: Compartmental mathematical model allowing person-to-person and water-borne transmission of cholera. We modeled within- and between-region epidemic spread with between-region transmission dependent on population sizes and distance between regional centroids (i.e., a so-called “gravity” model).

Setting: Haiti, 2010-2011.

Data Sources: Haitian hospitalization data, 2009 census data, literature-derived parameter values, and through model calibration.

Measurements: Dates of epidemic onset, hospitalizations.

Results: The plausible range for cholera’s basic reproductive number (R0 , defined as the number of secondary cases per primary case in a totally susceptible population without intervention) was 2.06 to 2.78. Order and timing of regional cholera outbreaks was predicted by our “gravity” model.
Analyses incorporating changes in disease dynamics over time suggest that public health interventions have made a substantial impact on the course of
this epidemic. A limited vaccine supply provided late in the epidemic was
projected to have a modest impact.

Limitations: Simplifying assumptions necessary for modeling; projections based on initial epidemic dynamics inferred from available data.

Conclusions: Notwithstanding limited surveillance from the Haitian cholera epidemic, a model that associates the strength of between-region disease transmission with the size of and distance between populations (analogous to
gravity) closely reproduces reported disease patterns. This model is a tool that planners, policy makers, and medical personnel seeking to manage Haiti’s cholera epidemic could begin to use immediately.

Jan. 21, 2011
12 pm
Room 210

Jane Heffernan (York University)
Mathematical immunology: The effects of the immune system and of immunity on disease progression in-host and in a population

The immune system, how it affects disease progression in-host and how immunity is developed and is used to prevent future infections are not well understood. I will discuss different aspects of these topics, informed by mathematical models, of four different viral infections: HIV, Herpes, viral hepatitis and influenza.

Dec. 2, 2010
12pm
Stewart Library

Eva Wong, MPH (c) 1,2, and Amy Greer, MSc, PhD 1,2

1 Division of Epidemiology, Dalla Lana School of Public Health, University of Toronto
2 Surveillance and Risk Assessment Division, Public Health Agency of Canada

Using Mathematical Models for Public Health Decision-Making: How Models are Contributing to Public Health Planning for Remote and Isolated Communities in Canada and the Renewal of the National Antiviral Stockpile

The 2009 Influenza A (H1N1) pandemic had a relatively mild effect on the general Canadian population. However, a disproportionate burden of illness was observed among vulnerable groups including Aboriginal populations and individuals living in remote and isolated communities. Dynamic models for infectious disease have the ability to reshape the strategic thinking for pandemic planning in the future from the development of planning scenarios to providing scientific advise on specific public health measures such as strategies for antiviral drug stockpiling, vaccine development and prioritisation, as well as non-pharmaceutical intervention measures. We will discuss two examples of the use of mathematical models for public health decision-making. First, we will describe ways that models are contributing to decisions regarding the renewal of the National Antiviral Stockpile. Second, we will discuss ways that agent-based models can help us to better understand the transmission of respiratory infections in remote and isolated communities in Canada and how we can use these models to explore optimal intervention strategies. The research discussed will play an important role in informing policy decisions to mitigate disease outcomes and protect the health of vulnerable populations in the future.