W e consider a sequential decision problem in national health policy: the weekly deployments of limited influenza vaccine doses, as they become available, to the various geographical regions of the United States during a pandemic event.As with the 2009 H1N1 pandemic flu, we assume that the progression of flu infection varies from region to region, with some regions starting their flu waves weeks before others. We also assume that vaccine doses only become available after flu waves have already started in some regions. Whereas the traditional deployment of vaccines is in direct proportion to resident population, without regard to flu status in any region, we show that a new policy that dynamically considers the status of the various regional flu waves can dramatically reduce the incidence of flu infection over the entire country. The method uses mathematical models of flu spread and requires capturing real-time data on flu incidence.

Our analysis suggests that current Centers for Disease Control and Prevention policy of allocating flu vaccine over time in direct proportion to states' populations may not be best in terms of averting nationally the maximum possible number of infections.

No abstract

W e present a policy-oriented summary of our six-year "service-systems-focused" research into pandemic influenza. We cover three topics: (1) R 0 , the basic reproductive number for the flu; (2) NPIs, non-pharmaceutical inventions to reduce the chance of becoming infected; and (3) flu vaccine allocations. We use a service-systems framing and mathematical modeling approach incorporating theories and data on the spread and control of influenza. We examine how behavioral actions and governmental policies, thoughtfully derived, can minimize influenza's societal impact. There is widespread misinterpretation that R 0 is a numerical constant of a given virus. We argue that it is not, but rather that its value is largely determined by local conditions and actions, many under our individual and collective control. This control is, in the absence of vaccine, intelligent use of NPIs-highly effective in reducing the spread of influenza. Our vaccine analysis relies on government data depicting flu-like cases and vaccines administered during the 2009 H1N1 outbreak. During that outbreak, barely half of all states received allotments of vaccine in time to protect any citizens. The method of vaccine deployment-in proportion to census population-ignored the temporally uneven flu wave progression across the United States.

No abstract