Monday, February 14, 2011

Does Building a Primary Care Workforce Start with Medical School Admissions Committees?

Between COGME’s 20th report recommending expansion of the nation’s primary care workforce (see 1/28/11 post) and the health care rhetoric and legislation coming out of Washington over the past few years, it seems that those in health care industry and policy are in full agreement that a primary care shortage exists in the United States and that the number of primary care physicians we produce in the coming years will have a significant impact on both cost of and access to quality care.

But how do we as a nation produce more primary care physicians? And is setting a numerical goal (i.e. 40% of physicians will be primary care physicians by 2020) enough?

Assuming we can improve reimbursement for primary care physicians (both male and female – see 2/3/11 post on gender-based physician salary gaps) such that family medicine becomes a more desirable financial option for those students who see reimbursement as a barrier to a primary care career, we will still have to address the disparity in access to primary care that occurs between geographical boundaries in the U.S.

In 1970, the federal government created the National Health Service Corps (NHSC) in an attempt to combat the changes in access to care that had begun over the prior two decades as rural physicians retired or moved to cities, where medical practices offered higher salaries and job opportunities for spouses. With the passage of the act that created the NHSC (and its subsequent amendments), the government recognized that proportion of patients living in areas with a population to practitioner ratio of 2,000:1 as “underserved.” These geographical areas were termed Health Professional Service Areas, or HPSAs.

Currently, the Health Resources and Services Administration (HRSA) website states that

“As of September 30, 2009, there are 6,204 Primary Care HPSAs with 65 million people living in them. It would take 16,643 practitioners to meet their need for primary care providers.”

Although over the past two years President Obama has reversed a trend in dwindling HRSA funding and expanded HRSA health workforce programs (including doubling NHSC funding to $300 million in the fiscal year 2010 budget and increasing that number by an additional $27 million for 2011 along with increasing funding for community health centers), we have to realize that more federal funding alone isn’t the panacea for curing the problem of geographic health workforce disparities.

A study published in the November 2010 issue of the American Journal of Public Health analyzed interview responses from primary care physicians in Los Angeles County, California concerning their reasons for practicing in their particular geographic location1. Only 24% (5 of 21) of interviewees practicing in underserved areas in the county chose their practice location because of loan repayment obligations, which seems to indicate that federal funding is a minor player in the eventual retainment of primary care physicians in such areas. Compare that to the 19% (4 of 21) of physicians in nonunderserved areas serving in their respective areas due to loan repayment obligations, and one starts to wonder whether loan repayment money really has an impact at all.

Digging deeper into the numbers of the study reveals that those physicians working in underserved areas were most likely to do so because of mission-based values (a “sense of responsibility or commitment to a particular community, a defined patient population, or a moral obligation”) and self-identity (including “language, personal, family, cultural, socioeconomic, and geographic backgrounds”). A 2003 study of a rural physician workforce in Florida provides similar insight, noting that physicians from rural backgrounds and physicians who were exposed to rural experiences in medical school and residency were more likely to practice in underserved rural areas2. To further the connection between training and practice, a 2009 study in the Annals of Family Medicine looked at training residents in community health centers (CHCs) and found that family medicine residents who trained at CHCs were four times more likely than their colleagues to go on to practice at CHCs3.

In addition, among primary care specialties in the L.A. study, the vast majority of physicians practicing in underserved areas were family physicians, whereas internal medicine and family physicians shared an equal percentage of the total in the nonunderserved cohort.

So what does all this mean? It would appear that prior experience in an underserved community and a sense of responsibility to that community is truly the major motivator in where a primary care doc practices. It means that medical schools and residencies need to partner with CHCs to encourage graduates to work with the kinds of populations that frequent them. Even more importantly, it confirms what we should have known all along: the best medical school candidates are those who already enter with mission-based values and who self-identify with those communities that most require family physicians.

Easier said than done, but we can’t hope to rely on loan repayment opportunities and training alone to funnel more students into family medicine. Opportunities to train with an underserved population best stimulate those who already want to work in that setting. It’s up to medical school admissions committees to select medical students who fit that bill.

The Robert Graham Center (RGC), with private foundation money, is already producing objective data on how well medical schools and training centers are fulfilling their “social mission” (that is to say, how well they are meeting the health care needs of the public)4. The RGC uses geographical information systems mapping tools to display these results visually. The only limitation of this data is that it is elicited from the American Medical Association Physician Masterfile which, although fairly comprehensive, still is prone to a small amount of error.

Knowing that, I propose a more effective use of federal money to improve the primary care services in the United States: create a national database of registered physicians and their specialties and locations of practice; then, instead of tying the majority of a medical school’s funding to the amount of research it produces, tie a large portion to the percentage of each school’s graduates who end up working as primary care physicians in underserved communities.

It would be a carrot approach, much like President Obama’s “Race to the Top Fund” for improving primary school education, and perhaps it would create an incentive (to compete with incentives from research funding and alumni donations) for admissions committees to take a closer look at the values of their applicants.


  1. Odom Walker K, Ryan G, Ramey R, Nunez FL, Beltran R, Splawn RG, Brown AF. Recruiting and retaining primary care physicians in urban underserved communities: the importance of having a mission to serve. Am J Public Health. 2010 Nov;100(11):2168-75.
  2. Brooks RG, Mardon R, Clawson A. The rural physician workforce in Florida: a survey of US- and foreign-born primary care physicians. J Rural Health. 2003 Fall;19(4):484-91.
  3. Morris CG, Chen FM. Training Residents in Community Health Centers: Facilitators and Barriers. Annals of Family Medicine. 2009 7:488-494.
  4. Bein, Barbara. Robert Graham Center to Study 'Social Mission' of Teaching Hospitals, Health Centers. AAFP News Now. 27 July 2010.


  1. Good post.
    I think the recommendation is to have 40% of graduates entering PC by 2020; you couldn't have 40% of doctors be PC even if every class between now and then wa 100%! It takes 30 years to replace the workforce. So, we need to shoot for much higher percents of graduates than 40%.
    Clearly, the characteristics of students admitted are the main determinants of doctors graduated. There is plenty of research on what characteristics of students make them likely to do FM/PC or work in a rural or underserved area. Some are demographic: from a rural area; older; underrepresented minority group; lower socioeconomic status. Others are experiential: volunteer work (Teach for America, Peace Corps, Americorps, Vista, or at least carrying the rape crisis beeper in college.) Stated intentions on an essay do not predict this. In fact, even the "experiential" group, while they may enter PC and practice in underserved areas, are unlikely to practice in rural areas (actually, of course NHSC and IHS are exceptions). So the only issue is: will medical schools completely change their admissions criteria for the bulk of the class, not a small group?
    See recent article by Stephen Smith in NEJM A Recipe for Medical Schools to Produce Primary Care Physicians,(Feb 10 2011).
    The issue for me is Smith is calling for admissions criteria for the "newer" medical schools; I think they need to be for ALL medical schools.
    Also, of course the social missions issue was addressed by Mullan and his group from George Washington (Annals of Internal Medicine, Jun 2010).
    I will be commenting on this later this week on my blog, Medicine and Social Justice (

  2. How about step1 score? Does that have a positive predictive value?

  3. IMO, trends in desire for specialty and limitation of spots in those specialties drive competitiveness. USMLE scores are an easy starting point for Residency Directors/Screening Committees. If there was a sudden influx of applicants for a specialty, regardless of the field, there needs to be objective data to utilize - USMLE scores happen to be an objective data point between applicants. Also, for what it's worth, recent studies have shown no correlation between MCAT, USMLE/COMLEX 1/2/3 Scores.

  4. MCATs correlate with grades in preclinical work. Not with clinical performance. And to a certain extent, correlate with characteristics probably not desirable in most physicians (less interpersonal skill, empathy). Good discussion on the last two posts (Feb 4 and Feb 8) on John Delzell's blog "Education in Medicine",
    He reviews a lot of literature on this, including Gough (not only the 1978 article he cites, Gough HG. Some Predictive Implications of Premedical Scientific Competence and Preferences. J Med Educ 1978; 53: 291-300.
    but even this from 1963, !) and Robert Sade,, both definitely worth reading.

  5. Great post.

    I plan to quote you in a post I am writing about the National Health Service Corps on

    Thank you for this
    Maggie Mahnar

  6. This is a really good read for me. Must agree that you are one of the coolest blogger I ever saw. Thanks for posting this useful information. This was just what I was on looking for. I’ll come back to this blog for sure! I bookmarked this blog a while ago because of the useful content and I am never being disappointed. Keep up the good work.

    Laith Salma