I think you're misrepresenting "Dunning Kruger." It only predicts that the subset of your respondents already known to be low ability in the domain of health care economics will end up inflating their assessment of their own ability in the same domain. (Perhaps also inverted for high ability respondents.)

You're claiming to predict that-- for particular branch of response types-- all respondents will be low ability. There are a lot of ways I would characterize that claim, but none of them would be "Dunning Kruger."

In fact, my gut tells me that some significant number of flame wars I've read over the years were due to this confused heuristic.