MIDAS Consultation on Pandemic Influenza Meeting Report

Emory Conference Center
October 27-28, 2004
Paul Glezen, chairman


Paul Glezen, M.D. , Chair
Influenza Research Center
Houston, Texas 77030

Robin Bush, Ph.D.
University of California Irvine
Irvine, CA  92697

Richard Compans . M.D.
Emory University
Atlanta, GA

Nancy Cox , Ph.D.
National Center for Infectious Diseases
Centers for Disease Control and Prevention

Marja Esveld
World Health Organization
Geneva, Switzerland
0041 7913004

Kathleen Gensheimer, M.D.
Maine Bureau of Health
Augusta, Maine 04333

Frederick  Hayden, M.D.
Division of Infectious Diseases
University of Virginia
Mark Lipsitch, Ph.D.
Harvard School of Public Health
Boston, MA 02115

Marshall McBean, M.D.
University of Minnesota School of Public Health
Minneapolis, MN

Arnold Monto, M.D.
University of Michigan
Ann Arbor, MI 48109-2029

Peter Palese, Ph.D.
Mount Sinai School of Medicine
New York, NY 10029

MIDAS investigators met with experts (see above roster) on influenza on October 27-28.  This consultation was intended to increase MIDAS participants' knowledge of the virus and disease and to establish working relationships with scientists who can provide guidance to the MIDAS pandemic influenza project.

MIDAS investigators invited experts to help the modeling effort by

  • Identifying important modeling questions;
  • Keeping the models "honest" by evaluating the models' underlying assumptions, determining the appropriate parameters and parameter values, and assessing the model outcomes; and
  • Identifying resources, especially data for estimating parameters and for docking models.

The experts gave a series of presentations and participated in lively and productive discussions on all of these issues.  Topics included

  • Natural History of Influenza
  • Surveillance
  • Characteristics of Pandemic Influenzas
  • Data Collection Systems
  • Population Biology and Evolution of Influenza Virus
  • Intervention by Antivirals and Vaccines
  • Public Health Issues
  • International Issues and Projections

The experts' experience brought out the complexity of influenza information and public health issues, such as understanding the population and evolutionary dynamics of the virus, studies of vaccine and antiviral efficacy, the natural history of the disease, and the difficulty of getting useful data.  The presentations are available on the MIDAS Portal at www.epimodels.org Link to external Web site.

MIDAS researchers and influenza experts agreed to continue to collaborate and to work toward a better understanding of influenza outbreaks by sharing ideas, results, and data.

After two days of informative and intense discussion, the consultation ended on the afternoon of October 28.

Immediately following, the MIDAS researchers met to consider how to make use of this knowledge.  The following researchers were present:

Don Burke
Phil Cooley
Derek Cummings
Jamie Cuticchia
Irene Anne Eckstrand
Stephen Eubank
Neil Ferguson
Elizabeth Halloran
Suzanne Lea
Ira Longini
Mike Riggs
Diane Wagener

Action Items:

Derek Cummings
Collect Thailand influenza data, including census, mortality, hospitalizations, etc.

Irene Eckstrand
Set up MIDAS Influenza Collaborative Group

Neil Ferguson
Find out if and when the G8 flu parameter meeting will be held
Send contact information for Mallick Peiris to Irene

Suzanne Lea
Build bibliography relevant to influenza modeling

Ira Longini
Invite small group to set model parameters

Consider how to become involved in WHO, possibly by sending an intern to work in the organization and present MIDAS
Immediately start developing model for H5 influenza outbreak in Asia

Model for Asia Outbreak of H5 with Human-Human Transmission

The Network agreed to develop two models for an outbreak in Asia, one based on Ira Longini's influenza model and the other based on Neil Ferguson's model.  The specific modeling question is

What intervention strategies can contain or eliminate an outbreak of avian influenza in Asia?

Model Characteristics
The model will be based on a rural and small town region containing 500,000 people.  Don Burke and Neil Ferguson have demographic data and will be able to collect some additional information quickly.  Once the data are in hand, the social structure will be determined.

Data Needed

  • Population density
  • Distribution and size of schools
  • Distribution and size of health centers
  • Age structure
  • Probability of travel outside locality
  • Age specific attack rate (estimate from U.S. data)
  • Household secondary attack rate (estimate from U.S. data)


  • Transmission probabilities
  • Time of antiviral introduction
  • Amount of available antiviral (approximately 2 million doses)
  • Time of detection
  • Travel incidence


  • Number of illnesses
  • Number of deaths
  • Time to extinction of outbreak
  • Likelihood of escape


  • Complete data collection and model parameters by December 10, 2004
  • Complete model runs and summarize results by January 10, 2005

Parameter identification
It is important to estimate model parameters as accurately as possible.  We agreed to ask two or three experts to help with this task, either at a face to face meeting or by conference call.  Ira Longini will contact Fred Hayden and Arnold Monto and possibly Mallick Pieros about participating.

MIDAS Influenza Collaborative Group
A productive collaboration among influenza experts and modelers is fundamental to the project's success.  We agreed to set up a collaborative group to

  • Ensure accuracy and quality of models
  • Focus the modeling questions appropriately
  • Assist MIDAS in working effectively with policymakers
  • Help MIDAS present model results to the public

Irene Eckstrand will identify members and launch the group.


Dr. Miron Livny presented information about the Condor computing project at the University of Wisconsin and its applicability to MIDAS.  Condor is a specialized distributed computing system for compute-intensive jobs. Condor provides a job queueing mechanism, scheduling policy, priority scheme, resource monitoring, and resource management.  We discussed several issues such as security, availability of computers, and the ability to network existing computers.  Ira Longini's group will explore how well Condor works on the existing Emory models.  Information on Condor is available at http://www.cs.wisc.edu/condor/ Link to external Web site.