The Framingham study is a longitudinal study used to assess heart risk in the human population, using a large and long-term sample from a single community. That community is supposed to represent the rest of the human population (or at least the U.S.). Of course, there are inherent problems with that, but statistics and sample size, length of study, and medical rigor help to make this single study translatable to the rest of the population.

No study is perfect. So let's assume that you are asked to develop a similar study on heart attack and stroke, but for the fire service in the U.S. How would you go about picking your population to work with? Be very specific here - what department, agency, or area would you look at? How would you control for people leaving the fire service and moving away? How would you make sure you have good representation within the fire service? Is there any issues that you may not be able to address? For instance, working just with one agency may yield firefighters who work mostly in a very urban and large city. Using a rural fire department might help get a broader representation, along with some WUI/wildland, but may lack significant access to urban, high-rise, or even various traffic collision responses. How do you solve this problem... or can you?