Distribution issues and doctor shortages

The following is a guest post by Allan Joseph, a medical student at the Warren Alpert Medical School of Brown University and TIE research assistant. You can follow Allan on Twitter: @allanmjoseph. This post is part of a series on the Institute of Medicine’s report on graduate medical education (GME), which you can follow with the tag IOM on GME.

In my first post on GME, I considered the evidence for and against an aggregate doctor shortage, ignoring distinctions in the types and locations of physicians. I concluded that the evidence suggests we might be able to cover any forecast shortages with improvements in productivity, but that it’s still reasonable to worry about it in the long term. But that’s not the whole story — those distinctions on type and location really do matter.

Let’s start with the types of physicians. Most people who have heard about the idea of a doctor shortage have heard about in the context of a primary care shortage. And there’s some reasons to be concerned, or at least investigate further. Almost one-fourth of primary care physicians are “near retirement” (over age 55), while medical students display historically low interest in primary care in large part because it pays far less than specialty careers — millions of dollars less over the course of a career. If primary-care physicians leave the workforce and there aren’t enough medical students to replace them, then it’s easy to forecast that supply won’t meet demand for primary care.

But if our look into the aggregate numbers taught us anything, it’s that we have to consider the forces of technology and delegation in improving productivity when forecasting supply. And it’s in primary care where these forces are most applicable. Technological advances could be important for care coordination and practice efficiency, especially when it comes to new practice arrangements or scheduling protocols. In-office equipment such as ultrasounds might allow for quicker diagnosis. Telemedicine could allow primary-care doctors to work more efficiently as well. (Of course, that’s all pending evidence in favor and reimbursement for these items by insurers, and assuming that new best-practices and treatments don’t slow physicians down We’re outlining the optimistic scenario here.

But delegation is the driving factor here: some estimates suggest over half of chronic care could be delegated from physicians to other clinicians. Even if that’s an upper bound, it suggests we really could be delegating a large number of primary-care duties from physicians to other health professions, as long as those services are provided at an equivalent quality. This 2010 review article would suggest that outcomes and patient satisfaction for patients treated by nurse practitioners (NPs) — even in randomizedtrials.

Yet delegation from primary care physicians to other clinicians is incredibly controversial, so though I want to keep this relatively short, let’s discuss why that’s so. I mentioned this in my first post on the doc shortage, but your view of delegation in primary care depends on your perception of the services offered by health professionals other than physicians, primarily physician assistants (PAs) and NPs, though pharmacists and other professionals are sometimes involved. It’s the difference between a shortage of primary care physicians and a shortage of primary care services.

If you think the primary care services to be delegated can be delivered by PAs, NPs, and others at an equivalent quality to those delivered by physicians, you probably think there might be a shortage of primary care doctors but not of primary care services, which are what really matter. If you think there’s some fundamental difference — e.g., that there’s a greater chance of non-physician providers missing a rare diagnosis, or that patients will dislike being seen by them — then you probably think delegation isn’t such a good idea because it doesn’t truly address the shortage of primary care services.

Primary care isn’t the only field where we see shortages. There are some specialty fields that are projected to see shortages in the future, and here’s where things get a little dicey. Technology might still play a role in increasing productivity here: new treatments or surgical techniques might reduce time spent treating, following up, or managing side effects. Yet it’s also possible that technology for specialty cases could harm productivity (think new surgical techniques that take longer but have better results, or health IT changes that slow workflows)Again, productivity-enhancing technology is an optimistic scenario.

More importantly, delegation is far less powerful in specialized fields. There are far fewer specialty-physician tasks that can be appropriately delegated to an advanced-practice nurse, physician assistant, or other clinician. (That’s not to say there aren’t some — certified registered-nurse anesthetists are a good example.) You couldn’t even delegate those tasks to many other physicians without years of additional training. So if there is a forecast shortage in certain specialties, that’s probably much more serious than a shortage in primary care. We simply can’t increase supply as easily.

Geography is perhaps the most pressing distributional issue right now. Certain areas of the country are facing severe shortages right now, in a problem that’s been recognized for years. On a per-capita basis, there are more than three times as many doctors in metropolitan areas as in rural areas. That’s problematic, especially since very rural areas are about as sick as so-called “inner city” areas.

Physicians’ choice of where to practice is influenced by a host of factors, including family choices such as opportunities for spouses and children. The types of amenities physicians want are often found in urban and suburban communities, where relatively wealthy professionals in general like to live; primary care shortages are concentrated in rural areas.

The easy answer to geographic shortages is simply to increase reimbursement in areas where physicians are in short supply. Many (if not most) doctors would move if the pay difference was big enough — if a primary-care doctor could tack a zero on the end of his or her paycheck by moving from Chicago to Southern Illinois, it might well be a choice worth making. After all, with enough money, you can send your kids to boarding school and fly out to visit them every weekend and your spouse is less worried about finding a job in a rural area. But the sort of public investment we’re talking about doesn’t seem to be on the table.

That’s not to say that the problem is intractable. Physicians tend to stay near where they trained. Medical students who were born rurally, expressed an interest in serving underserved populations, or were exposed to rural training programs during medical school are much more likely to choose to practice in rural or other underserved locations. The federal government also funds well-known loan-forgiveness programs such as the National Health Service Corps (NHSC) that pay physicians’ (often substantial) loans for practicing in medically-underserved areas. Thoughresearchsuggests doctors participating in such programs tend to leave their placement sites quite quickly after they fulfill their obligations, they also move to practices that cater to underserved areas.

Armed with this research, it’s possible to envision a strategy for increasing the proportion of doctors who choose rural or underserved medicine: select applicants who will want to practice in those areas, and expose them to those areas throughout medical education. That strategy doesn’t necessarily guarantee that every place with a shortage will receive more doctors, but it’s a start.

It’s clear that geographic distribution of physicians is a problem, and the IOM report that prompted this series on GME thinks so, too. It’s also problem that requires more attention than it gets. If you’ve got research that you think might be informative, do send it my way.

I’ll keep posting on GME in the days to come. Stay tuned!


Hidden information below


Email Address*