An article published in the Early Edition of the Proceedings
of the National Academy of Sciences in the United States of America
suggests that we can reduce the likelihood of a pandemic influenza
outbreak in the United States by quickly implementing social-distancing
alongside antiviral treatment and prophylaxis (preventive measures)
until a vaccine becomes accessible.
The study was conducted by three teams of researchers in the US and England who worked closely with federal officials. The teams and an informatics group - part of the Models of Infectious Disease Agent Study (MIDAS) Network, funded by the National Institute of General Medical Sciences (NIGMS) - set out to study several intervention combinations to aid the planning process for a national pandemic.
The authors, led by M. Elizabeth Halloran, M.D., D.Sc. and Ira M. Longini Jr., Ph.D., (Fred Hutchinson Cancer Research Center and professors of biostatistics at the University of Washington), investigated the natural course of infectious diseases by using sophisticated mathematical and statistical models. Longini notes that to make sure the results were robust, the federal government wanted three groups to be working on the same problem. The data, says Longini, "would be used to inform national pandemic planning," and thus they got the highest level of input.
Since a flu vaccine was not available when the researchers began the study, they focused on the effectiveness of using antiviral and social-distancing interventions (e.g. closing schools) together in order to prevent an influenza pandemic. According to earlier studies, models have shown that an available vaccine, even if low-efficacy, would be helpful in reducing the speed of a pandemic.
Halloran notes: "The good news was that all three of the disease-modeling groups involved in the study found that an outbreak of pandemic flu similar to the pandemic of 1918 could be mitigated if these measures were implemented quickly."
One research team consisted of Halloran, Longini, a computer scientist Shufu Xu, and others at the Los Alamos National Laboratories. Researchers at Imperial College in London and the University of Pittsburgh comprised a second group, and a third group included researchers at Virginia Bioinformatics Institute at Virginia Tech in Blacksburg, Va.
Each research team used separate but similar computer models. They calculated how influenza would disperse in a city of about 8.6 million people, such as Chicago, IL. The models assumed that residents of the virtual community interacted in households, schools, workplaces, and the community, just as they usually do. The models all assumed attack-rate patterns that were analogous to US flu pandemics that have occurred in the past.
To take into account real-world unpredictability and several other characteristics of the disease and the population, Halloran and colleagues use stochastic modeling when predicting the spread of influenza. The first step in building pandemic models requires researchers to come up with various assumptions about the ways in which people interact and how the virus spreads. They can then introduce intervention strategies to the model in order to test how effective they are.
Two categories of intervention were assessed:
- Medical intervention: surveillance used to identify cases, and antiviral agents are used to treat patients and prevent the disease among close contacts
- Non-pharmaceutical: social distancing - closing schools, voluntary quarantine, restricting travel restrictions
The three computer simulations predict about 47 to 60 percent of the population will have symptomatic influenza if there is no intervention.
In the least-stringent scenario, interventions were implemented after 1 percent of the population had developed symptomatic influenza, schools were closed, 60 percent of clinical influenza patients received antivirals and their contacts received prophylaxis, 30 percent complied with quarantine, and 60 percent complied with social-distancing measures.
The three computer-modeling groups predict an 83 to 94 percent reduction in cases of influenza using combined intervention strategies at a lower transmissibility of the virus, even in the least-stringent scenario.
Longini maintains that the researchers "ran this simulation with the assumption that the pandemic was as virulent and lethal as the 1918 pandemic." He adds: "Even when modeling the situation of pandemic flu, with a modest compliance range in social-distancing measures, and modest ability to identify and treat and prophylax with antivirals, the interventions were similarly - though not identically - effective in all three models."
The researchers believe that the policy implications of these findings are noteworthy. "If one could achieve these levels of compliance, ascertainment and social distancing, then there is a possibility of considerably mitigating a pandemic until a vaccine was available." They caution, though, that the levels of disease ascertainment and compliance that were entered into the models may not be realistic.
"These models, which are built from the best available data and with the best tools, contribute greatly to our understanding of how a pandemic could spread and what measures might protect the public's health," said Jeremy M. Berg, Ph.D., director of the NIGMS, which supports the MIDAS program. "But they are not our only resource; field work and experimental studies remain critical and will enhance the quality and reliability of these and other models," conclude the authors.
Modeling targeted layered containment of an influenza pandemic in the United States
M. Elizabeth Halloran, Neil M. Ferguson, Stephen Eubank, Ira M. Longini, Jr., Derek A. T. Cummings, Bryan Lewis, Shufu Xu, Christophe Fraser, Anil Vullikanti, Timothy C. Germann, Diane Wagener, Richard Beckman, Kai Kadau, Chris Barrett, Catherine A. Macken, Donald S. Burke and Philip Cooley
Proceedings of the National Academy of Sciences of the USA. (2008).
Click Here to View Abstract
Written by: Peter M Crosta