SF: Mostly driven by demand, there will be a substantial shortage of physicians by 2020 that cuts across most specialties. The specialties that serve the elderly will show the greatest shortages. That urges us to take a look at the physician supply situation that has evolved and ask some probing questions about demand and physician supply. The demand and supply analysis that we developed shows approximately a 25 percent shortage nationwide of physicians per 100,000 patient population by 2020 and a shortage of between 27 and 43 percent in Pennsylvania specifically, under current supply patterns. Nationally, there’s been about a two to three percent growth in the number of physicians every year, going back about 25 years. We had a similar growth pattern in Pennsylvania until the mid to late 1990s, when the growth effectively leveled off. If you use the national 20-year-average supply trend pattern, Pennsylvania will have a 27 percent shortage by 2020. But if you look at physician supply as having leveled off, and that continues, then the shortage becomes much greater: the 43 percent level.

SF: We looked at some specialties. I have yet to look at regional trends. If you look at specialties like orthopedic surgery, the shortage gets substantially larger – in the range of 40 to 50 percent. If you look at urology, the shortage becomes ghastly – in the range of 80 percent. The reason is that baby boomers are about to come into their 65-year-old age bracket starting at about 2011. It is baby boomer demand that really fuels the shortages. Our current supply pattern is geared to serve the population in its current aging form, with baby boomers as contributors to the economic funding of the system, not as heavy demanders. There are even findings from a recent national ambulatory care survey that show that the demand for medical care by baby boomers, people in the 45-65 year old age group right now, has substantially increased over the past generation, and that baby boomers are already demanding more medical care than we might expect.

SF: Some of them are hitting now, but by 2011 we’re really going to be into new territory in terms of the demographics of the demand side. People between 65 and 80 demand a lot more medical care than any other group: three or more times the rate as younger people. The demographics of the baby boomer phenomena will double, triple and quadruple the 60-80 year age group between 2011 and 2030.

SF: I made the assumption that supply and demand are pretty much in equilibrium right now. There’s not a lot of evidence of physician surpluses anywhere and there is some, but not an overwhelming amount of, evidence of any real shortages. I made the assumption that, over the last 20 years, supply and demand have been in equilibrium. My demand projections are based on demography and people’s use of physicians. I assumed a continuing increase in demand, and there are about ten factors that play into that, including ever-increasing levels of technology, people’s wealth and insurance levels and non-linearity in an elderly population – where demand for physicians more than doubles with a doubling of the number of elderly. These are all things that we can tinker with if we want to change demand pattern. All I’m saying is, if something doesn’t change, this is where we’d end up. Obviously, that’s not where we would end up if we make some policy decisions and make some changes.

SF: I made the assumption that physician supply would continue the way it’s evolved over the last 20 years. That includes the same growth in doctors of osteopathy, the same proportion of foreign medical graduates that comes into residency and stays, same number of medical school slots. I assumed that a two to three percent growth would continue, nationally.

SF: Basically, yes. And that’s a strong assumption because it’s not quite true in all areas. There are a number of specialties that are already showing shortages, but there’s general acceptance of the current level of equilibrium.

SF: Medical care will not be as accessible as it is now. Usually when you have shortages you’ll get a price push: the prices of services increase. I don’t think the system is economically equipped, prepared or able to increase the price of physician and hospital services in response to a shortage. The question then becomes, how will the system adapt to deal with it? There’s a whole range of responses. Will there be pressure to reduce demand by increasing copays or denying care to some groups of people? Will there be pressure to increase the number of medical school slots or foreign medical graduates? Will there be pressure to allow greater use of limited licensed practitioners? How will people react to longer waiting times for physicians, shorter physician visits, or inability to see physicians? Will Medicare decide to provide benefits at 70 instead of 65? Will we have national health insurance? If this evolves as the model suggests, and people have trouble accessing medical care, that will become a political issue in local, state and federal campaigns. Depending on the interventions we determine, there could well be mortality and mortality effects. We’re talking about a 25 percent shortage. That’s huge.

SF: No. I have not done anything empirical on that issue. That would be a good area for research. Future research should also look by specialty and geography in terms of asking where these shortages are likely to hit first and hardest.

SF: A lot of the previous data were for specific specialties and focused on recent trends in Pennsylvania. Data from my previous study showed that there are actual physician departures from Pennsylvania, or lower numbers of physicians per population in high risk specialties. We’ve lost about 1,400 physicians from the state. The current study is more of a macro model, focusing on the aggregate using 20 year average trends. So, I’m not really focusing on some of the problems that have occurred in Pennsylvania recently.

SF: Yes, that was the worst case scenario I talked about. In the best case scenario, if we returned to a two to three percent growth, that still gives us a physician shortage problem.

SF: Not that I know of. Physicians are harder to study, and changing the physician market will be harder to do than dealing with large insurers and large hospitals. The data is difficult to get at. The rewards and the policy level implications for drawing conclusions aren’t as poignant as they are when dealing with HMOs or hospitals.

The Pew Grant has funded some really good liability work and the level of commitment to it is substantial. Unfortunately, I think a fair number of people who come at this from the academic arena don’t quite understand everyday physician practice issues. William Sage has looked hard at liability trends over some period of time. I think Bill has some personal opinions about tort reform that may be warranted from his perspective, but the every day world demands some of them – I think Bill is opposed to caps, for example. The Case Western Reserve University professor who has looked at ethical and philosophical concerns isn’t dealing with the direct impact of liability dollar problems on a physician practice. Randall Bovbjerg advocates a modified no-fault system for medical liability issues. I think he’s right, but the practical world of physician liability isn’t ready for it yet. Professor Michelle Mello at Harvard has put together a survey of Pennsylvania physicians to take a look at the manpower issue. But nobody is really looking at this as a physician economics issue. Liability is only a piece of it. My latest research is saying that, while the things that we have studied are important – liability, reimbursement, payment processes and supply, when you overlay the coming demand issue, the demand problem swamps us all.

There are physician-to-patient ratios that some HMOs use to benchmark adequacy of physician supply by specialty. They are really bare bones minimums to make sure that HMOs have enough doctors, and they are not the level that the population will demand for medical care. So those are undercounts, if anything. Should we be setting better benchmarks? Absolutely. It would be nice to try to do a bottom-up grassroots study of some major specialties and establish a benchmark.

SF: They have issues of today to deal with. They’re worried about the MCARE Fund. There is no state agency or organization to do this kind of research. We talk about transportation, high speed rail transport, factory jobs, airlines, farm production, milk pricing – a whole gamut of issues that we trust our state and private universities to research. One of my visions for Pennsylvania is that there would be a research entity that will look at public policy questions across the spectrum.

SF: The trial lawyers are saying that the number of physicians in Pennsylvania is rising. They use licensed physician numbers. They don’t exclude residents. They double-count residents who moonlight. They include retired physicians. If you just look at licensed physicians, you do not get a decent picture of what’s going on here. I use about five or six sources for physician supply data. The principal one I use counts physicians engaged in patient care, as listed in the Area Resource File of the United States Bureau of Health Professions. I’ve overlaid that with MCARE data, Universal Physician Identifier Number data, and AMA Physician Master File Data. It’s so difficult getting good counts that you try to get information from a variety of sources and try to find areas of agreement.

SF: First of all, the bias is there – I think it’s fair to make that allegation. The scientific response to the criticism would be, "All right, show me a better model, a better projection and let’s discuss it." The data that I’m using is not the Pennsylvania Medical Society’s list of physicians in the state. All of these data are collected by governmental entities for other purposes. It is my spin, if you want to be fair: I’m showing baby boomer demographics and I’m claiming there’s a problem there. Other people can dismiss it. My response is to say, "Look, there’s a coming demographic issue here. Don’t just dismiss the fact that there might be a future problem because I’m the person saying it. Go look long and hard, and show me there’s not going to be a problem here. I’ll listen."