"The specialties that hospitals offer, and that residents choose, play an important role in determining the direction of medicine. Give a man a hammer and everything's a nail; make him a surgeon, and everyone needs an operation.
The problem is that because it takes so long to train new doctors—10 years after medical school in some fields—the models used to suss out future needs don't work well. They can account easily for future population and fairly well for the diseases people will have, but they can only make wild guesses about what technologies and treatments will be available to treat those diseases. The pediatrics my father learned during his residency in the late 1960s is different from the pediatrics he practiced in the 1980s and '90s—when there were new vaccines, new treatment guidelines, and even new diseases. When a government committee issued an influential report in 1980 about the supply of physicians, cardiac stents (the small mesh tubes that keep Vice President Dick Cheney's heart beating) hadn't even been invented. So, how could the committee guess how many invasive cardiologists would be needed today?"