How can OR help stop the spread of the swine pandemic?
For those of you who weren’t at the talk this morning by Mark Ehlen, Chief Economist at Sandia National Laboratories, you missed a very interesting discussion on applications of OR in Homeland Security. Of particular interest to those of us who have been traveling to and from this conference, is the concern about the spread of the swine disease and whether or not we might be subject to getting it. So, it was reassuring to know that although the disease has already spread to 7 or 8 states in the US, Homeland security is actively involved in curbing the spread. Dr. Ehlen discussed the economic implications for closing borders to try and stop the spread from Mexico, which might be our first reaction. However, analyses show that this would have only bought the US maybe a couple of weeks before it ultimately did land in our backyard with very severe economic implications, which would only compound our current economic crisis. The approach is that we know the disease will be in our backyard even if we close borders due to global travel as a way of life and business, so how do we preposition vaccines to minimize the time it takes to get these vaccines to people to minimize mortality. There are many opportunities for OR practitioners and academics in the area of homeland security, and I hope these kinds of talks will also be a part of future INFORMS programs.
Several INFORMS members who attended the conference are modeling pandemic and preparedness: At GA Tech, Eva Lee and Pinar Keskinocak (who collaborates with Julie Swann on humanitarian logistics) and John Fowler, whose Decision Theater at ASU has been examining best ways of reducing the disease’s spread, particularly in Arizona. Another INFORMS member, Sheldon Jacobson, has been modeling pediatric vaccines and the fight to avert pandemic.
Is anyone interested in nonparametric survival analysis of A(H1N1)? Statistically sufficient data are available from the WHO text updates, http://www.who.int/csr/disease/swineflu/updates/en/index.html. Survival analysis could show actionable features:
->Half of the SARS fatalities were due to preventable? secondary pulmonary infection after ~4 weeks
->Country comparisons could show virulence and effectiveness of treatments and other measures
->Multivariate analysis (network tomography) to estimate joint travel time and survival time distribution for comparisons, between countries and across epidemics.
The results are nonparametric estimates of the age-specific survivor function, P[X > t], and the actuarial death rate. The random variable X is the time from reported or confirmed A(H1N1) infection to death according to the WHO. The scale is daily, from the WHO updates.
The survival analysis results will be similar to http://www.fieldreliability.com/SARSStat.ppt and http://www.fieldreliability.com/BFSummry.xls.