Since avian influenza H5N1 became a potential public health threat in 2003, public health agencies around the globe have been diligently planning for the next influenza pandemic. While the concerted response to H1N1 reflects this careful preparation, several expected and unexpeted events, including its apparent North American origin, the rapid overburdening of U.S. laboratory capacity, non-uniform testing and treatment policies among U.S. states, and delays in production of a viable vaccine, all reinforce the need for a dynamic and quantitative playbook for pandemic mitigation using pharmaceutical countermeasures.
By adapting an established algorithm to optimize disease mitigation policies, we advance from the traditional candidate strategy approach to rapid and systematic analysis of numerous policy options. This is just one of many possible optimization methods suitable for this purpose ,,,
. Our choice of UCT was based on the insight that, with some careful modeling, disease intervention strategies can be nicely mapped onto policy trees and that this approach can be coupled to any stochastic epidemic model. This approach has performed successfully on large policy trees 
and has favorable convergence properties 
. In particular, it is guaranteed to converge on the optimal policy eventually, unlike simulated annealing and genetic algorithms. We do not claim that the optimized antiviral strategies (Fig. 4) are necessarily the best possible. Rather, we simply argue that in a reasonable amount of time, the model converged on policies that are predicted to perform equal to or better than obvious candidate policies, and that the resulting policies make intuitive sense.
Unexpectedly, our analysis suggests that the optimal distribution schedule for the U.S. national antiviral stockpile may be quite simple. A monthly distribution of 5 million regimens to states consistently matches or outperforms other policy options, regardless of the level of uptake or rate of misuse (Fig. 3a). Slight variations on this policy, for example, regular distributions of 1M or 10M courses are predicted to perform significantly worse at intermediate levels of uptake. Since there can be many optimal policies, the results of the optimization algorithm do not necessarily have the same simple structure as the 5M strategy. Our optimization allowed for the possibility of distributions proportional to prevalence, although such actions are not consistent with the current strategic national stockpile policy and would likely be both politically and logistically difficult. Notably, the results suggest that prevalence-based distributions are not expected to enhance the impact of antivirals.
The expected impact of this simple 5 million course monthly distribution schedule is highly sensitive to the rate of antiviral treatment (
). From an ongoing study of H1N1 antiviral uptake in Milwaukee, preliminary estimates of the fraction of reported cases receiving treatment within 48 hours of developing symptoms are less than 20 % 
. This suggests that we are likely in the range where all strategies perform equally poorly and are predicted to minimally mitigate transmission. Although treatment beyond 48 hours may not alter clinical course, there is some evidence that it may lead to a more rapid drop-off in viral shedding thus reducing transmission 
. Thus public health measures to increase the usage of antivirals for H1N1 have the potential to slow transmission prior to the availability of H1N1 vaccine, but the impact of such measures will critically depend on the Strategic National Stockpile distribution schedule. Although antivirals may not reduce transmission at current levels of uptake, they can significantly reduce morbidity and mortality associated with H1N1 when used to treat potentially severe cases.
Our analysis did not consider the threat of antiviral resistance. Currently circulating strains of seasonal influenza have acquired resistance to oseltamivir 
and there is evidence that H1N1 has exhibited this resistance as well 
. We also did not incorporate the use of antivirals for prophylaxis, the future availability of vaccines, simultaneous use of NPI’s like school closures, or the option of targeting the stockpile towards particular demographic groups, all of which are likely important and may influence the optimal policy.
From rapid genetic sequence analysis to automated syndromic surveillance systems, public health emergency response is rapidly improving in technical capabilities both in the U.S. and worldwide; the rapid response to and characterization of the novel pandemic influenza A (H1N1) virus is a testament to this. However, planning the policies of public health response to such identified and emergent threats remains a highly non-quantitative endeavor. We present here a policy optimization approach that is highly modular and can be easily adapted to address multiple additional issues. Our hope is that the quantitative methods will assist clinical experts in developing effective policies to mitigate H1N1 using a combined arsenal of vaccines, antivirals and NPI’s. Specifically, a very similar analysis can be used at the international level to optimize global allocation of the WHO's limited antiviral stockpile to resource-poor countries. One can substitute any stochastic model of disease transmission, at any scale, for our national-scale, U.S. H1N1 model. In addition, while the optimization algorithm is particularly well suited for time-based interventions, any well-behaved policy space can be used 
. The approach should thereby facilitate a more comprehensive consideration of pandemic policy options, and will perhaps confirm the efficacy of the current policy or suggest more strategic options for the future 
. The approach should thereby facilitate a more comprehensive consideration of pandemic policy options, and will perhaps confirm the efficacy of the current policy or suggest more strategic options for the future.