Requirements-based (gap analysis) - for the selected topic/problem, determine the requirements your selected user base needs in order to solve each aspect of the issue. Work backward from the desired end-state, identifying all the components. Priority gaps/needs will become evident in this backwards-planning technique.
I believe a major learning point (which seems to be lost much of the time) is that outbreaks do no happen in a vacuum and secondary effects can actually be more detrimental to health and well-being over the long term.
For instance, recent estimates cite >$600 billion dollars in economic damages due to the swine flu outbreak - much of it do to inappropriate and uninformed public actions. I understand that the human health angle is of primary concern - and rightly so - however, very little has been talked about with regard to impact on agricultural, transportation, and economic systems (which are all intertwined).
Egypt preemptively culled all of their pigs - swine flu is a zoonotic disease after all. However, was this merely an excuse to eradicate or relocate the swine population in the country? A secondary effect of the culling and relocation may be seen in the months to come. In Cairo, groups of these pigs are fed refuse collected from the streets, thereby helping to recycle food waste, but also helping to maintain a certain level of sanitation. Additionally, this practice has been a long-standing part of the social structure in the city. As the pigs are culled and/or relocated, what happens to the sanitation system currently in place, the jobs associated with the pig production, and the effect of it's absence on the sanitation of the neighborhood - especially the poorer neighborhoods?
A biosurveillance capability needs to be more of holistic entity, incorporating monitoring of these factors in addition to merely counting and reporting disease signs and symptoms.
Thanks for touching on the biosurveillance subject. The current surveillance situation is a little broader than Notifiable Diseases and depends on the methodology used for the surveillance as well as the end purpose of the surveillance.
Most of what you've discussed would fall into a 'direct reporting' capacity used by the national health infrastructure. Separately, there have been attempts at indirect, Syndromic Surveillance - basically looking for indirect indicators of disease in the population. By and large, this has not proven to be effective. Other systems use a methodology of looking at social/societal disruption associated with an outbreak of disease - such things as cancellation of sporting events, festivals, or school closures. This methodology may not be 'disease specific'; however, it has proven to be fairly sensitive. A downside is that one must understand and take cultural perspectives, expectations, and practices into account.
Another point of debate in the biosurveillance community is that regarding purpose of the surveillance system. Traditional monitoring tends to lean toward a true epidemiological approach - it is slow, but you generally get a high degree of specificity. Alternatively, one can take the 'tip off' approach, effectively trading specificity of detection with much more rapid (and sensitive) indication. You might detect that something is happening and have vague descriptions of disease signs, but not a true clinical description.
Each methodology and purpose has it's strengths and weaknesses, and use of the various systems is not mutually exclusive. The questions become: How is each system's reporting used most effectively? How do we make decisions from the gathered data? How do we effectively communicate our conclusions? How do we integrate our surveillance and response methodologies?