New Medical Schools: Serving Concentration or Distribution?

Robert C. Bowman, M.D.

The nation's medical schools have moved two different directions in the past decades.

Medical schools that admit and train differently graduate family physicians, primary care physicians, and physicians who serve lower and middle income populations. These can be called distributional medical schools. Another perspective would be that these are schools that

1. Admit more students of lower and middle income origins

2. Train students in normal health care venues (not concentrations of patients and rare conditions)

3. graduate more physicians who are found outside of concentrations of physicians

The national average is 75% of physicians found in zip codes with 75 or more physicians, zip codes with 400 - 1500 physicians per 100,000 (average is 300 for the nation), zip codes with 150 - 330 primary care physicians per 100,000 (sufficient primary care is considered to be 100). These are locations with 4% of the land area and 35% of the population and over 90% of GME, NIH, medical students, residents, clinical funding, and other sources of health care funding. Internal medicine and pediatric generalists have 70% found in concentrations along with 75 - 92% of subspecialists. Top ranking MCAT schools have 76 - 87% of graduates found in concentrations. Concentrations in admissions, training, and policy concentrate optimally. Combinations of concentrations are the rule in concentrational school admissions. The medical students represent concentrations of parent income and most urban origins that have concentrations of top scores. They were born, raised, educated, and trained in top concentrations of physicians, professionals, health and education resources, property values, and opportunity.

Distributional medical schools have balanced distributions nearing 50% inside and 50% outside of concentrations. Family physicians from distributional medical schools have 60% of graduates outside and this matches up well to the 65% of the population outside of distributions. Osteopathic family physicians have 60% distribution outside. Distributional medical schools are health access vehicles with admission, training, mission, and family practice contributions optimizing distributional and primary care outcomes. Distributional medical schools lead in primary care contributions with the most family physicians lasting 30 years on average in primary care. Distributional medical schools graduate higher ratios of family physicians compared to other forms of primary care. In the distributional medical school graduates, those choosing internal medicine and pediatric residency programs also are more likely to remain in primary care.

About 23% of physicians (subtracting military and international) are found in zip codes with less than 75 physicians. These are locations with 80 - 150 physicians per 100,000 ranging from half served urban (150 per 100,000 compared to the national average of 300) to one-fourth served urban (80). Rural half served (120) and third served (100 per 100,000) areas are in between these extremes. The terminology describes their relationship to national average physician per population concentrations. All of the outside locations have 30 - 60 primary care physicians per 100,000 or levels insufficient for health access spanning 65% of the nation's population and 96% of the land area.

The best distributional osteopathic schools and Duluth distribute 55% or more outside with top levels of primary care, health access, women's health, care for the elderly, and rural specialists.

It is not possible to promise researchers and specialists and health access. Distribution at the current time requires a fully coordinated effort at multiple levels across birth to admission, admission, training, and policy.

Physicians found in distributional locations or outside of concentrations (less than 75 physicians at a zip code) have different characteristics from those with top status, most urban origins, professional or physician parents, and proximity to medical schools and physician concentrations.

• Those born outside of concentrations have 2 - 3 times odds ratios of greater distribution outside.
• Those choosing family practice have 2 - 3 times probability of distribution outside of concentrations.
• Osteopathic or Historically Black or distributional allopathic schools also multiply distribution by 1.3 or more times.
• Older graduates have 1.3 times greater distribution.

Standardized test scores and older age are a secondary measure of outside origins. Older age and lower scores reflect differences of income, education, and parents as well as different probability of distribution. Those who are rural, African American, Native, or Hispanic do have higher levels of distribution outside, but distribution can be limited in higher status children in each of these origins. Higher status, most professional children have common ground regardless of origins and have limited distribution unless training and policy are modified. Lower and middle income white and Asian physicians are also "outside" in origin and in distribution tendency. Of course the challenge is admission since only 25% of admission positions are available to the 75% of the population of lower and middle income origins. Over 60% of admission goes to the to 20% and another 20% goes to the second quintile leaving 60% of the population to compete for 20% of positions. Those with 3 to 1 probability of admission have the lowest distribution levels and set the national average. Those more normal have above average distribution. Those who have lower, lower middle, and middle income origins have top levels of distribution at 2 to 3 times distribution outside.

Distribution outside can be urban locations outside of concentrations or rural locations outside of concentrations. Outside is not just rural or minority or different. Outside is rural and urban, all races and ethnicities, and anyone except extremes of concentration.

The Myth of Why Doctors Won't Go

Higher probability of distribution does not mean that all rural born or all inner city origin medical students return to rural or inner city sites. Training, policy, and career choices are far too powerful for this. Origins are also at the beginning of life years before the career and practice location decisions. Life experiences just before career and location decisions are the most powerful, particularly residency locations. The best odds ratios involving origins are 2 to 1 or 3 to 1 regarding a better distribution.

For rural origin about 10% of the nation's medical students are rural origin and 25% of rural origin physicians are found in rural locations. This compares to 90% admitted from urban origins supplying 75% of rural physicians. Now those who point to the fact that 75% of rural origin physicians are found in urban locations are correct, but the odds ratios are 2.5 times or 25% rural compared to 10% of physicians in rural workforce on average.

Odds ratios multiply. Those with odds ratios stacked in their favor (outside origins, FP choice) are the most likely to distribute. For example urban underserved location (One-Fourth Served) is the practice location of about 4% of total physicians. African American family physicians have 18 - 20% target location rates. The doubling of African American and the doubling of family practice facilitates distribution. Of course most African American family physicians (80%) choose other locations. From the observations, it would appear that few distribute. This is not an uncommon statement to hear. This is not an informed statement. Those who are outside in origins do have a consistent 2 - 3 times odds ratios of distribution outside. Those who choose family practice have 2 - 3 times odds ratios of distribution outside. Older age multiplies distribution by 1.3 times. Distributional training contributes 1.3 times. Without a clear understanding of concentrations and distributions, it is possible to miss the truth of distribution. The truth of the matter is that admission, training, career choice, and policy influences so favor concentrations that anything different improves distribution.

The Myth of Distributional Impossibility

Distributional medical schools such as osteopathic schools and Historically Black have been successful in distributing physicians to lower and middle income America for over 100 years by admitting more normal distributions of students (often older), training in more normal settings (not top concentrations), and focusing on the physicians most needed for basic health access. Allopathic public schools created under the 1970s template of primary care reform also accomplished distribution using the same principles.

The Challenge of the Present

Osteopathic medical school positions will double from 2004 to 2017. Not all of the osteopathic schools have embraced distribution. The best distributional osteopathic models include the initial Kirksville design, the osteopathic public schools, and schools with serious efforts to admit and train outside.

Allopathic medical schools indicate their steady movement away from distribution with declines in family practice, rural, and underserved choices. New schools would seem to have little incentive to choose distribution. The new school at UT El Paso has embraced a border mission but this is a difficult task as it involves embracing lower and middle income, Spanish speaking medical students interested in family practice, primary care, and serving the underserved. 

The distributional medical schools face great challenges maintaining their missions under current education, health, and market forces policies of the nation.  

A new challenge is that medical students who are really interested in distribution can end up in medical schools that have less emphasis in this area.

This takes a great deal of advising, maturation, and careful consideration.

Many if not most medical students do not know where they intend to be. For those that intend primary care or underserved careers, the choice of a typical school under current policy is a choice away from these careers. Even a choice of internal medicine is a choice away from primary care since IM residency graduates remain in primary care less than 25% of the time with more declines to come. IM and PD choices of primary care are more difficult as 70% are located in saturations of primary care. This of course is a huge problem for primary care salaries. Admission of medical students not likely to consider outside distribution makes these saturations even more likely. All forms of primary care willing to leave concentrations can generate more salary and support, but many graduates born, raised, and trained for 30 years in concentrations are unwilling to leave, and will accept lower salaries. There appears to be no effort to change this.

The past 100 years has involved a focus on centralization, urbanization, professionalization, and physicianization in American medical education and health care. The end result is concentrations of physicians, professionals, health care resources, nurse practitioners, physician assistants, and even primary care with 65% of the population left behind in one or more areas. There are consequences for lack of health access support and a solid dependable primary care foundation.

The challenge for newer schools is which way to go - Concentrational or Distributional

Newer schools can choose the current concentrational model with specialization and research and urban focus and extremes of admission and training to complement this focus or they can choose the distributional model with different admission and training and outcomes. This is not a judgment on choice or consequence. It is a plea to be consistent for the sake of students. The nation needs distributional oriented students to go to distributional schools. There are too few to be wasted and policy does not allow distribution to be created from concentration, as in the 1970s and 1990s. Distributional students need specific birth to admission preparation and student support services. They also need functional primary care environments to reinforce their training and career decision.

Medical schools cannot do both concentration and distribution. Admission and training are too complex and too costly to hope to include both directions. Policy has shaped this focus. The 1970s schools were shaped to primary care with significant federal and state support and primary care requirements. The funding was provided so the requirements were less burdensome. The nation quadrupled primary care graduates in a decade. The current expansion will be lucky to maintain the same level of primary care graduates per year despite expansion.

The Myth of the Dual Mission School

There are those who point to the WWAMI system as a way to do both. The WWAMI system is an example of the complexity involved. WWAMI concentrations are readily evident in graduate outcomes. WWAMI distributions and improvements in health access are limited by the same parameters of admission, training, and policy. WWAMI outcomes are as good as distributional selections (in Wyoming, Idaho, Alaska, Montana) and distributional training (away from concentrations and back in target states). This means as little time as possible in Seattle among top physician and population and professional concentrations. Limitations in selections and training defeat this model. In recent years more incentives and obligations are planned to reinforce returns. Lack of family practice choice in recent years severely limits returns to states with 35 - 60% of their physicians family physicians. When a state is forced to take every applicant and cannot choose those most likely to return or those most likely to choose family practice, the effort is crippled from the start. This could also be called the myth of the rural admissions focus since some states try to take all rural candidates but find that taking all rural candidates also means taking those that are not suitable for primary care or rural location. 

This highlights the national role for birth to admission, admission, and policy support. The nation must also cooperate to maximize the impact of distributional training. Graduating more internal medicine or pediatric residents or more nurse practitioner or physician assistant graduates to improve health access is not a viable policy with poor primary care support. The current desertion of primary care policy severely cripples current and future health access. The policy deficits also mean fewer choosing the remaining permanent form of primary care. Loss of hundreds of family practice graduates compared to peak levels represent decades of primary care lost for each family practice graduate lost over the past decade. Primary care is a one way street with no return once departed.

There are models that illustrate coordinated admissions, training, and policy.

Duluth gets 50% family practice choice and top distribution and top production of primary care years because

1. the state invests in lower and middle income children such that there are many to select from during admissions

2. the school has been consistent in mission, admission, and training

3. the school trains half of graduates away from medical school settings for the third year

4. advisors know that it is mostly a waste to send those not fitting the mold of background or interest in rural practice and family practice, those sent tend to fit the mission, students in the state know of the school and its mission

Of course the state could have perhaps increased the support for Duluth, unchanged for decades. If you think that lower tuition costs help Duluth, think again as the tuition is at the $40,000 level common ground with the most costly schools.

New schools must be clear in mission and admission and training.

New Schemes for Jump-Start

The University of Central Florida has captured Orlando funding such that the first 40 students have free medical education. Schools that offer full tuition are going to face a severe challenge weeding out all of the stories. Students and their parents will do incredible things for such a generous offer. If the school has a clear idea of the students that they want and the training that they want, then it is easier.

Schools with distributional focus require more time and effort in a number of areas. Such a school is very difficult to establish and maintain unless college health professional advisors have a clear understanding of the school and the mission of the school.

There is a need to match up students with their desired training to fit national needs. The nation has few interested in improving health access. Those who end up in a typical school can be shaped away from health access careers. While it is a great honor to be selected to a top school or to have funding needs met, the requirements may involve attending a school that has the lowest probability of primary care and distribution. Schools with top alumni funding have been able to capture the students that they desired for research or the higher scoring students that fit their profiles as well as diversity indicators. This is of course not considered fair by medical students investing signficant time and effort to develop research or preadmission pipelines, but such are current policies and practices. Of course if a student with top probability of distribution is influenced into a concentrational school, the probability of distribution is diminished. 

Medical students with interests in research or subspecialty careers sent to distributional schools by advisors or those who had no other options for admission may also find the environment difficult.

I applaud those who have managed to capture full tuition for charter classes. I envy their efforts and success.

I hope that our school can also match this level of student support but my hope is that the hoped for support package matches our admission and training and mission. We do not have this level of support, but we can dream. With 100% funding for the medical students who commit to 10 years of underserved primary care, admission, training, and mission all fit together for maximal health access outcomes. Students unclear of their intentions or those desiring low cost medical education are steered elsewhere by the commitment. Nations and states and foundations and associations should be willing to fund such a model. So far this is still a dream.

Is the nation's more or less likely to meet health access goals when National Health Service Corps positions go to students trained in the top concentrational medical schools?

Why do we send medical students with distributional origins and distributional training to military careers since the military support package is the best for those who are older with families and primary care career choice?

Education, opportunity, admission, and training environments guided by market forces and current policies make concentration easy to obtain.

Birth to admission, admission, training, career choice, and policy effects must be carefully crafted to obtain health access in the current environment.

Each extra primary care year from each remaining primary care practitioner is important.

Those born, raised, educated, and trained in concentrations may not understand this - but they can be taught.