Caitlin Yeaton Rivers edited discussion.tex  over 10 years ago

Commit id: c7d5eac2bfff4ff5460b514b05d129c503db241c

deletions | additions      

       

\section*{Discussion}  Case tree and checkerboard plots can assist epidemiologists with visualizing and analyzing zoonoses with human to human potential. The plots provide valuable insight into the dynamics of an outbreak not available using current visualization tools.   The basic reproductive number is easily estimated visualized  by evaluating the size and shape of each tree in the case tree plot over time. Functions available in epipy can further assist with basic reproduction number calculations.  Clusters of trees that remains short and thin likely have poor human to human transmissibility, whereas trees that become progressively taller and wider as the outbreak progresses may be gaining in transmissibility. Case tree plots also allow quick identification of trees with an unusally large number of branches. This observation is useful for identifying superspreaders, who can play a critical role in accelerating an outbreak, as was the case of SARS. The rate at which new trees pop up is useful for estimating spillover frequency, which in turn is useful for identifiying animal hosts. Numerous trees over a short period of time may suggest a domesticated or agricultural animal host rather than a wild animal. As case tree plots become more common, patterns will likely emerge that indicate certain emergence scenarios, similar to how the shape of epidemic curves can differentiate a point source outbreak from a continuous sort outbreak.  Finally, case tree plots visually represent and accompanying analysis functions allow insight into  case outcomes attributes  by generation and by cluster, which can reveal new insights. For example, a disproportionate number of male index nodes could suggest that expsoure to the disease occurs in a male-dominated setting like hunting. A high case fatality risk among index nodes compared to non-index nodes suggests that the infection is less virulent when transmitted from person to person than when acquired from an animal source. Revealing this patterns is useful for understanding the epidemiology of the disease, and for identifying effective interventions.