Community Mitigation Strategies in an Influenza Pandemic
Community Mitigation Strategies in an Influenza Pandemic
Using a networked, agent-based computational model of a stylized community, we evaluated thresholds for rescinding 2 community mitigation strategies after an influenza pandemic. We ended child sequestering or all-community sequestering when illness incidence waned to thresholds of 0, 1, 2, or 3 cases in 7 days in 2 levels of pandemic severity. An unmitigated epidemic or strategy continuation for the epidemic duration served as control scenarios. The 0-case per 7-day rescinding threshold was comparable to the continuation strategy on infection and illness rates but reduced the number of days strategies would be needed by 6% to 32% in mild or severe pandemics. If cases recurred, strategies were resumed at a predefined 10-case trigger, and epidemic recurrence was thwarted. Strategies were most effective when used with high compliance and when combined with stringent rescinding thresholds. The need for strategies implemented for control of an influenza pandemic was reduced, without increasing illness rates.
Community goals during an influenza pandemic include protecting people from illness and maintaining critical societal functions by limiting time away from usual occupations. Vaccine and antiviral medications are standards of influenza prevention, postexposure prophylaxis, and treatment. However, vaccine for a new influenza subtype may not begin to be available for at least 20 weeks after the onset of a pandemic and would be supplied over many months. Antiviral drugs may be in greater supply, but their effectiveness and rapid availability are uncertain. The US government has proposed community mitigation strategies for limiting the harm or managing the pace of an influenza pandemic until vaccine becomes available. These behavioral- and pharmaceutical-based strategies rely on reducing viral transmission and include dismissing schools and public gatherings, voluntary sequestering in the home, staggering work shifts, keeping symptomatic persons isolated, and treating ill persons rapidly with antiviral drugs and providing antiviral prophylaxis for their household contacts. These community mitigation strategies would be applied according to a pandemic severity index (PSI) scaled as categories 1-5. Category 5 would be a 1918-like event (case-fatality rate >2.0%) and category 1 (case-fatality rate ≤0.1%) would be akin to a bad seasonal influenza year. Modeling studies have estimated the effectiveness of mitigation strategies with and without vaccine and antiviral drugs. However, an independent review of pandemic influenza modeling studies raised the question of whether and when community containment strategies might safely be rescinded without reinitiating an epidemic.
An earlier study of this computational model demonstrated that closing schools and curtailing contacts of children and teenagers for the duration of a mild 1957-like epidemic in a stylized community reduced the number of infected persons by >90%. The model was constructed with assumptions that children and teenagers are responsible for influenza transmission in a community because of the frequency and nature of their person-to-person contacts. However, sensitivity analyses showed that permutations of mitigation strategies that included adults were effective at reducing infections in the model population, even for more highly infective 1918-like viral strains or with removal of enhanced children/teenagers' role in transmission. Several studies have shown that combining strategies such as social distancing of adult groups in and outside the workplace and removing symptomatic persons from community contact substantially reduced infections except in epidemics caused by the most infectious viral strains.
The US government's community mitigation guidance recommends rapid initiation of strategies, then up to 4 weeks of school closure for a PSI 2-3 pandemic and 12 weeks for a PSI 4-5 pandemic. However, this guidance fails to address the gap between 12 weeks of mitigation strategies and estimated vaccine availability beginning at 20 weeks, especially if antiviral drugs were of limited effectiveness or availability. A pandemic could recur in the intervening period, and nonpharmaceutical community mitigation strategies with rules for their use would be valuable tools. Although nonpharmaceutical community mitigation measures have been used with apparent success in past pandemics, there are concerns about unintended consequences such as economic losses, interruption of education, and restrictions of personal freedom. The potential impact of community mitigation strategies warrants further study and consideration.
We evaluated effects of rescinding 2 community mitigation strategies for influenza pandemics, seeking a balance of the effect of illness, risk for epidemic recurrence, and minimization of the duration of mitigation strategies. The 2 strategies bracket mitigation measures that might be logically used in a situation in which effective vaccine and antiviral drugs are not available. The strategies are child sequestering, which is included in the US community mitigation guidance of February 2007, and a most conservative measure of all-community sequestering. We instituted strategies early, after 10 cases of mild (1957-like, PSI 1-2) or severe (1918-like, PSI 4-5) pandemic influenza occurred in a stylized community; these strategies were rescinded according to incident cases within a specific period.
Using a networked, agent-based computational model of a stylized community, we evaluated thresholds for rescinding 2 community mitigation strategies after an influenza pandemic. We ended child sequestering or all-community sequestering when illness incidence waned to thresholds of 0, 1, 2, or 3 cases in 7 days in 2 levels of pandemic severity. An unmitigated epidemic or strategy continuation for the epidemic duration served as control scenarios. The 0-case per 7-day rescinding threshold was comparable to the continuation strategy on infection and illness rates but reduced the number of days strategies would be needed by 6% to 32% in mild or severe pandemics. If cases recurred, strategies were resumed at a predefined 10-case trigger, and epidemic recurrence was thwarted. Strategies were most effective when used with high compliance and when combined with stringent rescinding thresholds. The need for strategies implemented for control of an influenza pandemic was reduced, without increasing illness rates.
Community goals during an influenza pandemic include protecting people from illness and maintaining critical societal functions by limiting time away from usual occupations. Vaccine and antiviral medications are standards of influenza prevention, postexposure prophylaxis, and treatment. However, vaccine for a new influenza subtype may not begin to be available for at least 20 weeks after the onset of a pandemic and would be supplied over many months. Antiviral drugs may be in greater supply, but their effectiveness and rapid availability are uncertain. The US government has proposed community mitigation strategies for limiting the harm or managing the pace of an influenza pandemic until vaccine becomes available. These behavioral- and pharmaceutical-based strategies rely on reducing viral transmission and include dismissing schools and public gatherings, voluntary sequestering in the home, staggering work shifts, keeping symptomatic persons isolated, and treating ill persons rapidly with antiviral drugs and providing antiviral prophylaxis for their household contacts. These community mitigation strategies would be applied according to a pandemic severity index (PSI) scaled as categories 1-5. Category 5 would be a 1918-like event (case-fatality rate >2.0%) and category 1 (case-fatality rate ≤0.1%) would be akin to a bad seasonal influenza year. Modeling studies have estimated the effectiveness of mitigation strategies with and without vaccine and antiviral drugs. However, an independent review of pandemic influenza modeling studies raised the question of whether and when community containment strategies might safely be rescinded without reinitiating an epidemic.
An earlier study of this computational model demonstrated that closing schools and curtailing contacts of children and teenagers for the duration of a mild 1957-like epidemic in a stylized community reduced the number of infected persons by >90%. The model was constructed with assumptions that children and teenagers are responsible for influenza transmission in a community because of the frequency and nature of their person-to-person contacts. However, sensitivity analyses showed that permutations of mitigation strategies that included adults were effective at reducing infections in the model population, even for more highly infective 1918-like viral strains or with removal of enhanced children/teenagers' role in transmission. Several studies have shown that combining strategies such as social distancing of adult groups in and outside the workplace and removing symptomatic persons from community contact substantially reduced infections except in epidemics caused by the most infectious viral strains.
The US government's community mitigation guidance recommends rapid initiation of strategies, then up to 4 weeks of school closure for a PSI 2-3 pandemic and 12 weeks for a PSI 4-5 pandemic. However, this guidance fails to address the gap between 12 weeks of mitigation strategies and estimated vaccine availability beginning at 20 weeks, especially if antiviral drugs were of limited effectiveness or availability. A pandemic could recur in the intervening period, and nonpharmaceutical community mitigation strategies with rules for their use would be valuable tools. Although nonpharmaceutical community mitigation measures have been used with apparent success in past pandemics, there are concerns about unintended consequences such as economic losses, interruption of education, and restrictions of personal freedom. The potential impact of community mitigation strategies warrants further study and consideration.
We evaluated effects of rescinding 2 community mitigation strategies for influenza pandemics, seeking a balance of the effect of illness, risk for epidemic recurrence, and minimization of the duration of mitigation strategies. The 2 strategies bracket mitigation measures that might be logically used in a situation in which effective vaccine and antiviral drugs are not available. The strategies are child sequestering, which is included in the US community mitigation guidance of February 2007, and a most conservative measure of all-community sequestering. We instituted strategies early, after 10 cases of mild (1957-like, PSI 1-2) or severe (1918-like, PSI 4-5) pandemic influenza occurred in a stylized community; these strategies were rescinded according to incident cases within a specific period.