In a future outbreak of pandemic influenza, such as the three pandemics that sickened millions and killed hundreds of thousands of people during the 20th century, supplies of flu vaccine might not be available quickly enough to contain the spread of disease. However, according to research by biostatisticians in Emory University's Rollins School of Public Health, many thousands of deaths could be prevented if antiviral agents were given to the close contacts of those with suspected cases of flu until adequate supplies of vaccine could be manufactured and distributed.
The results of the research by Emory professors of biostatistics Ira Longini, Jr,, PhD, and M. Elizabeth Halloran, MD, DSc, and their colleagues Azhar Nizam, MS, and Yang Yang, BSc, will appear online on March 26, and will be published as a special article in the American Journal of Epidemiology on April 1, 2004.
The Emory scientists used a dynamic stochastic simulation model of an influenza pandemic or bioterrorist attack for an agent similar to influenza A(H2N2), which caused the Asian influenza pandemic of 1957-58 and resulted in approximately 70,000 deaths in the U.S. (A stochastic model includes elements of chance or probability). They determined that if no interventions were used in a similar pandemic, 33 percent of the population would become ill, resulting in a death rate of 0.58 per 1,000 people. If antiviral prophylaxis was given to close contacts of 80% of suspected influenza cases, however, in a strategy that the authors call "targeted antiviral prophylaxis" (TAP), the epidemic could be contained. If TAP were begun within one day of identifying suspected flu cases and used for up to eight weeks, only 2 percent of the population would become ill, and the death rate would be only 0.04 per 1,000 people. The researchers found that eight weeks of TAP would be nearly as effective as vaccinating 80 percent of the population.
"The ability to rely on targeted antiviral therapy in the case of a major public health threat from influenza would be an extremely valuable strategy, due to the characteristics of influenza pandemics," Dr. Longini said. "Epidemics are usually due to strains of flu that acquire genetic variations that differ slightly from earlier strains, while pandemics occur when there is a dramatic shift in the current flu strain. Annual flu vaccines are designed to counter the strain of flu from the previous season, and often there is insufficient time to catch up with making enough vaccine to counter the first wave of a new outbreak. Manufacturing and distributing a vaccine to match a newly identified strain can take six to eight months."
Although the surveillance and containment strategy (i.e., isolation of cases and quarantine of close contacts) was recently used to successfully contain the spread of severe acute respiratory syndrome (SARS), influenza has a much shorter incubation period (1.9 days as opposed to 6.4 days), and has a much broader range of clinical illness, from asymptomatic to primary pneumonia. Thus the surveillance and containment strategy is unlikely to work in containing an influenza epidemic.
"Because surveillance and containment would be an ineffective strategy to contain an influenza epidemic, stockpiling antiviral drugs, along with a well-planned distribution strategy, could be an effective first line of defense while an adequate supply of vaccine is being produced to counter subsequent waves of infection," Dr. Longini explained.