Delays in isolating cases by even one or two days, however, could hamper control of an epidemic, according to research by Ira M. Longini, Jr., PhD and M. Elizabeth Halloran, MD, Sc, Emory professors of biostatistics and their colleagues Azhar Nizam and Yang Yang, Dr. Longini will present the research at the AAAS Annual Meeting in Seattle on February 14.
Although mass vaccination during or before a smallpox attack would result in fewer cases and deaths than would surveillance and containment, the researchers concluded that an increased rate of vaccine-related serious illness and death would offset the slightly lower number of smallpox illnesses and deaths, especially since no one knows when or where an intentional release of smallpox would take place. The scientists also concluded that even if a particular smallpox virus were altered to cause more serious, or hemorrhagic cases than the ordinary virus, surveillance and containment would still be effective, although more deaths would occur.
In an earlier study (Science, Nov. 15 2002), the researchers concluded that targeted vaccination of the close contacts of infected individuals during a smallpox outbreak following a small attack could rival the effectiveness of mass vaccination, given a sufficiently high level of immunity within the population. However, as of May, 2003, fewer than 35,000 people had been newly vaccinated against smallpox, and many states had temporarily halted their smallpox vaccination programs. Although routine smallpox vaccinations were stopped in 1972, recent studies have shown that previously vaccinated individuals retain substantial immunity.
In the current study, the Emory investigators constructed a model that simulated the spread of smallpox in a hypothetical community of typical contacts, using a population of the approximate age and household distribution of the U.S., following a larger smallpox attack deliberately introduced through aerosolized virus. The model assumed that people interact in structured subpopulations primarily within known contact groups, including their own household, schools or daycare centers, their neighborhood and their community. The model was based on a network of these structured subpopulations of people connected through adult workplaces, high schools, and a hospital to form a population of 48,000 people.
The investigators tracked the smallpox epidemic over time in terms of disease symptoms and recognition, infectiousness and behavior of infected individuals. They divided smallpox cases into three categories of severity: ordinary smallpox, modified smallpox and hemorrhagic smallpox. If individuals in the model had never been vaccinated, they assumed that 95 percent would develop ordinary smallpox and 5 percent would develop hemorrhagic smallpox. They also assumed that for people over age 32 who were vaccinated before 1971, 10 percent would be fully protected, 30 percent would develop less severe disease and 60 percent would develop non-modified smallpox (95 percent ordinary smallpox and 5 percent hemorrhagic smallpox). They also divided the appearance of smallpox cases into a period before recognition and a period after recognition of the first case.
For people receiving a new smallpox vaccination before they were infected, the investigators assumed a complete response. For those receiving a vaccination within four days of infection, they assumed that 90 percent would not develop disease and that 10 percent would develop modified smallpox. For those vaccinated between 5 and 7 days of infection, they assumed that 60 percent would develop smallpox, 38 percent ordinary smallpox, and 2 percent hemorrhagic smallpox.
For their interventions, the researchers used either surveillance and containment or mass vaccination either before an epidemic or 20 days following identification of the first case. In the surveillance and containment strategy, also known as targeted or ring vaccination, after the first case is recognized, all hospital workers who deal with smallpox would be immediately vaccinated, recognized smallpox cases would be placed in hospital isolation rooms, and their close contacts would be vaccinated and kept under observation.
"We found that any delay in individuals withdrawing to their homes or to the hospital after they become ill would have much more of an adverse effect than would failing to vaccinate a few of the cases during the first period before smallpox is recognized in the community," Dr. Longini said. He added that "nevertheless, containment and surveillance would be effective in containing a moderately large smallpox attack."
"This demonstrates the necessity for preparedness in our public health system. A rapid and efficient means of detecting cases and isolating them in case of an attack would be an essential component of the strategy of surveillance and containment," Dr. Halloran said.