We’ve been told by COGME no less that there’s an impending physician shortage, and today I reviewed a whole bunch of material for a hospital that looked like it was true.  However, to put a spike in this balloon, an article published online in Health Affairs by Jonathan Wiener showed that even with extra recruitment over the 1990’s, the large prepaid group practices like Kaiser and HealthPartners still managed to serve their populations with far fewer physicians per 1,000 patients than already exist in the US, let alone the number that will be practicing in 20 years time. And all this with physicians allegedly working shorter hours and seeing fewer patients than in the wider FFS world.

I’ll update my thoughts about this later (It’s late and I just got off the plane!) BUT go give the article a read.

UPDATE. OK I’m off my 3 hours early morning call and can spend a moment extending this post, especially as it has a passing relevance to the project I’m working on.  Here’s the argument:

Way back in the early 1990s those forecasting the physician workforce made some assumptions that the US would move closer to gatekeeper/managed care model.  This type of a model assumed a split between specialists and generalists that’s close to 50-50, and in many countries it’s closer to 20-80.  The model also assumed that there would be fewer surgeries and procedures as the model unfurled.  At that point consultants wondering around with bed days per enrollee of Medicare managed care plans in southern California became a familiar feature of the hospital boardroom (yes, I admit I was one of them).  Essentially if you played out that model nation-wide we had about 75% too many hospitals and a few too few generalists and 50% too many specialists. That future never happened for a variety of reasons, mostly connected with the death of managed care. However we did see a reduction in the number of residency slots, including some teaching hospitals being paid to not train residents.

Instead we started seeing a rash of new procedures and technologies, especially in the unmanaged Medicare population, and the newly unmanaged HMO-lite population. Meanwhile there was a rash of hospital consolidation and bed reduction in the 1990s (although only about 10%). Then the prognositcators started to notice the impending arrival of the baby boom, the leading edge of which hits 60 next year and Medicare in 2010.  So we can do more things to more people and will have an increasing number of people to do them to. They also noticed that medical school applications had fallen (although not the number of those in med school, just fewer candidates for each place) and some surveys showed that most physicians wouldn’t recommend medicine as a career to a new student.  Yet we didn’t fundamentally change our system in the last two decade. So about a couple of years ago you started seeing articles like this one warning that we had a shortage coming and that we needed more doctors, and in fact late last year COGME recommended that we increase the level of residency slots 15%.

Weiner’s article simply points out that we can give appropriate care to a given population with a physician-to-population ratio that is 22-37% below the current national rate.  How do the bigger PGPs do it?  Not apparently by working their doctors harder–in fact they probably work fewer hours. Not by adding more primary care doctors–over the years primary care doc numbers in these groups grew less slowly than those of specialists, although the share of primary care docs remains higher than in the overall physician population. In fact specialty position growth in these PGPs exceeded that of the national average. Instead they use more physician extenders (nurse practitioners and physician assistants) for between 17% and 25% of primary care providers–as opposed to the 10% they represent in the overall primary care provider population. Kaiser in particular uses specialty care nurse practitioners–their growth was 16% annually in the 1990s. They also use more preventative care and disease management programs, probably work their procedural specialists harder (this is certainly the case abroad), and probably do less surgery

Meanwhile Solucient projects strong demand for cardiologists, GI and orthopedics docs while the folks at the Advisory Board (who’ve been known to extend a chart line well beyond breaking point in their time–anyone remember their forecast of 90% capitation by 2005?) believe that there’ll be a shortfall in specialist hours of between 35% for intensivists and up to 70% for cardiologists by 2030. They also have a neat chart in their recent report on physicians which correlates GDP per head in the US with MDs per thousand — in other words the richer we get the more money we want to spend on doctors? (Well that’s one interpretation!)

So how will this play out?  One thing to remember is that thanks to the expansion of med schools in the 1960s and 1970s we are still pumping out docs out of residency programs at about the rate of 20,000 a year, with only about 8,000 a year retiring, and that growth will continue until about 2015.  So the number of active non-federal doctors per 100,000 population, which is about 225 now, will peak at 235 in 2010 and only fall back to 230 in 2020. In a chart which includes NPs and PAs, Wiener shows that while the US now has a total of 230 MDs, DOs, NPs and PAs per 100,000, the big PGPs get by with 145-175. So although the rest of US health care lives in a different world than Kaiser and Group Health, anyone wanting more money for medical school and residency places is going to have to make a pretty convincing argument that they’re really needed–especially with $500 billion deficits out as far as the eye can see.  So, as it takes 8-10 years for a policy change to show up as the first "additional" doc in practice, I believe we’ll work with what we’ve got at least out until 2030 and probably beyond.

Medicare will inevitably have to slowly change its payment arrangementss to reflect this–although that’ll be a touch battle. Private plans are already working on similar ideas, such as pay for performance, and the folks at Leapfrog and IOM are also pushing for changes in the model of care delivery. So slowly over time expect the obvious:

–More use of phsyician extenders, such as other clinical professionals. 
–More and better use of technology to make physicians more efficient and patients better at self-care
–Innovative patient-centered practices that get around the "broken chassis" of the 8 minute office visit, and require less physician intervention
–Longer waits (eventually) for the real hard-core sub-specialists, higher salaries for those guys and more struggles between hospitals for the revenue generating superstars.
–Concominant rationing of the really expensive stuff. Don’t worry–you wouldn’t be able to afford it by then anyway!