Medicine is fractal

Couple of papers on generalist versus specialism crossed my desk. One in the BMJ (Munro et al, page 3 in the Careers section; and a debate pair of articles in Clinical medicine, by Firth and Fuller, Clin Medicine,(2014) 14;(4) p354 and 357. The articles in clinical medicine are well worth reading, although very focussed on the nightmare that is general medicine / acute takes (sadly, the online version of the RCP journal lags behind the paper version: a bit like the RCP). From a narrow dermatological perspective, much of what has been suggested in the Shape of Training review is flimsy, and parochial, and dominated by what I think is an out of date UK perspective on medical practice. The UK model is no longer an example for much of the world. There are however some general points worth thinking about whenever people trying and fit medics into a binary divide of ‘specialist or generalist’.
Medicine is fractal, we can move from being a primary care doctor who sees all age groups, through to primary care doctors who see (say) just children. Once you think about specialists, there are degrees of generalism. A dermatologist may see rashes and tumours, or be a surgical dermatologist and just see tumours. You may be a paediatric dermatologist, and not see adults, and a general paediatric dermatologist will refer patients with rare disorders such as xeroderma pigmenta to other sub-specialists, or to those with expertise in genodermatoses. If there is room for only one doctor to look after a bunch of scientists ion the Antarctic, we can think hard about what skills are needed, but they are not the same skills as a generalist primary care doctor in Morningside. When you have a lens, you see more detail; when you have a light microscopy you see yet more; when you have a scanning electron microscope, you see even more. The richness and knowledge needed of natural history, and therapy, is fractal.
The second point is that continued expertise is not just a feature of certification or historical accreditation, but of continued exposure. You can educate GPs all you like about melanoma (if I were a GP, I would find the terminology patronising, but I am repeating what others say), but if your exposure to true cases is ~ 1 every 10 years (such as for melanoma), there are experiential limits on competence. I do not know the shape of the function of competence against exposure, but I do not believe it to be linear. Part-time working also has to be considered.
Specialism comes with transaction costs if mis-referral happens, and those with blinkers do not see the error of their ways; or fails when the signal to noise ratio is too low to allow segmentation of the target group. Conversely, in anything but a small population, it would be bizarre to think that medicine would be immune to the advantages of specialism that, as much as anything else, have led to the scientific revolution and the modern industrial state.