Sources of the Current US Physician Workforce: 1987 – 2000 Graduates

Robert C. Bowman, M.D.

Some medical schools do a much better job of distributing physicians than others. These tables involve graduates of medical schools from 1987 – 2000. The 1987 – 2000 group has a low rate of inactive workforce and is providing the bulk of the physician care in the nation. The data sources include the 2001 and 2004 American Medical Association Masterfile and Robert Graham Center data. Different categories for primary care and family medicine allow comparisons and help avoid the shell game that is played by many medical schools reporting different categories of primary care.

One of the primary goals of physician workforce is to address the patients in most need of access in the nation. The front line specialties of primary care, mental health, Geriatrics, and women’s health need special consideration. There is a combined category designated “essential workforce” that includes these front line specialties.

Another consideration is the international schools. Those nearest the US tend to have more US citizens compared to those from more distant schools. The international schools have been divided into North American and Distant International for comparisons.

One of the first considerations is graduating physicians who will distribute to all areas of the nation. Only family medicine  distributes in  the same pattern as the US population

Internal medicine physicians or the combination of medicine pediatrics is not able to rival the distribution of family medicine. Studies also connect higher ratios of family physicians with increased quality of care and decreased health care costs.

The 1997 – 2003 FP group includes board certified residency grads who completed an allopathic accredited program. This group involves mostly medical students who graduated from 1994 – 2000 and this was the denominator used for a “match rate.” This group was impacted by the managed care era, with a 30-50 % higher choice of FP compared to 1987-1993 medical school graduates.

The 1987 – 2000 All Inclusive group includes family medicine, general practice, aerospace medicine, public health, osteopathic manipulative medicine, family medicine geriatrics and sports medicine, those with other specialty as a primary specialty and family medicine as their second specialty and those with family medicine combined with psychiatry or internal medicine.

The 1987 – 2000 FP GP group includes only those with family medicine or general practice specified as their primary specialty.

The 1987 – 2000 rural physician group includes all physicians of all specialty types who have located in RUCA codes 4 – 10.5, not including those with over 30% commuting to another town for work (.1 codes).

The 1987 – 2000 rural family physician group includes FP GP physicians who have located in RUCA codes 4 – 10.5, not including those with over 30% commuting to another town for work.

Location data is not yet complete for the last few years of graduates who continue with longer specialty training and those with delayed entry into the US from international schools. The inclusion of the longer duration from 1987 – 2000 data minimizes this error and also presents a more balanced view since workforce varied greatly during the managed care era. Data is more complete on primary care graduates with shorter duration of training.

Allopathic private schools and distant international schools graduate the lowest percentages of family physicians and rural physicians. They also have the lowest rates of essential physician workforce and office-based primary care. There are 3 allopathic private schools that provide substantial diversity through graduation of black physicians. These schools also have a higher rate of primary care, family medicine, and essential workforce. Outside of these three schools, allopathic private schools have no advantage in equity for minorities or women. They do graduate more physician researchers, but this is concentrated in the top schools. There are better methods of graduating physician researchers which will be discussed in later chapters. Attempting to improve the graduation rate of physician researchers by increasing admissions of those with speeded intellect, as noted by MCAT scores, adds little to physician researchers compared to distorting the workforce away from primary care and family medicine.

Essential Workforce involves 1987 – 2000 graduates in family medicine, general practice, geriatrics, office-based primary care, all obstetrics and gynecology, public health, and all psychiatry. These are essential specialties to continue to maintain the health quality and access indicators in the nation. They also continue to be the specialties in most consistent shortage in the nation.

All inclusive primary care includes 1987 – 2000 graduates who have designated internal medicine, pediatrics, medicine-pediatrics, family medicine, or general practice as their primary specialty.

Office-based primary care includes 1987 – 2000 graduates who are office-based in primary physician activity in the above all inclusive primary care cohort.

The FP GP group from the previous table is included for comparison.

The Research group includes those who have designated their primary activity as medical research. Because of delays in the time that it takes physicians to report their research careers, data was used for 1987-1996 graduates to give a higher and more realistic representation. Only 28 medical schools had over 1% of graduates as medical researchers. In the allopathic private group, 10 schools graduated half of the medical researchers in the entire group. The allopathic public and private schools graduating the most researchers also graduated the least family physicians.

Osteopathic graduates listed unknown specialty data in 12% of cases compared to 1% for allopathic and 6% for international.

Allopathic Public Medical Schools

In terms of numbers, allopathic public schools provide the largest portion of the workforce. Allopathic private and distant international schools provide the least essential workforce, primary care, family medicine, and rural physicians of all types. Allopathic private schools graduate more researchers, although the difference in numbers is not great compared to other schools. Duluth has been the standout among all allopathic schools, consistently graduating 50 % into family medicine, 22% into rural family medicine, and 30 % into all rural locations. Only a small class size of 60 keeps it from leading all lists other than research.

Osteopathic Medical Schools

Osteopathic schools provide the highest concentration of essential, family medicine, and primary care workforce. In the last 5 years of graduates, three osteopathic schools (Des Moines, Western, Philadelphia) have led the nation in total number of family medicine and general practice physicians. Osteopathic schools in Kirksville and Kansas City also are top ten contributors. The other top 10 FP GP contributors are Ross, American University, Indiana, Iowa, and Illinois. Iowa has one of the smallest class sizes of the top ten group. Ross led all medical schools with 90 allopathic FP graduates in 2003 and has continued to expand class sizes over the past few years. Among the osteopathic schools West Virginia leads in rural family medicine percentage at 20%. Again, only small class size prevents West Virginia osteopathic from leading the nation in more categories. West Virginia is also one of the youngest osteopathic schools.

North American International Medical Schools

North American International graduates are similar to allopathic public school graduates in distribution with a slight edge in family medicine and essential workforce. This suggests a similar composition of students admitted. Indeed some schools have as many as 70 % US citizens who have often attempted allopathic schools prior to international. When comparing schools such as Ross, St.George’s, and American University; the rural, primary care and family medicine numbers are even higher than the averages for allopathic public schools. Increased choice of family medicine involves consideration of admissions. Schools that admit a more diverse group of students, ones with a wider range of MCAT scores and birth origins, are more likely to graduate family physicians. Schools who concentrate efforts on students born in the most urban areas of the nation or students born in foreign countries or those from higher income origins are not likely to graduate the physicians most needed in the nation. Lower MCAT scores reflect admissions of a wider range of individuals with a fair component of those with reflect lower income levels. Many of these have had obstacles in education, family, and neighborhood.  

Studies continue to link lower socioeconomic level to higher choice of family medicine (Cooter). Students who will become primary care physicians also have lower MCAT scores. Schools with lower MCAT averages graduate more family physicians, primary care physicians, and rural physicians and have a higher proportion of internal medicine graduates choosing office-based practice (Bowman distribution).

Allopathic schools have had a more volatile relationship with family medicine. The birth of family medicine was followed by increased popularity of FP choice in the 1970s, followed by a slump in the  1980s. The peak years for FP choice involved the managed care years of 1994 – 1999 and have been followed by a crash in numbers and percentages in 2000 - 2004. International and osteopathic schools have been relatively resistant to such influences. Allopathic schools with increased numbers of urban born students have been the most vulnerable to changes in specialty choice. Urban born students had 50% increased choice of family medicine and primary care during managed care. International schools have been impacted by increased US entry requirements and visa activities. The international students entering family medicine are a smaller percentage, but have not fluctuated as much.

The international schools, both North American and Distant, have a wider gap between primary care and office-based primary care, a gap of 20 and 28 % respectively. This reflects more designating internal medicine, but choosing non-office-based careers. This 20 % gap compares to 10 – 13 % for US schools. US allopathic schools with a higher proportion of internal medicine physicians choosing office-based practice tend to be older, more rural in birth origin, and tend to come from schools with lower average MCAT scores. Medical school type and medical student characteristics are related to the primary care workforce that is available to the nation.

Categorization of medical schools into schools more or less likely to graduate family physicians can allow further analysis of the factors related to physician distribution.

Medical Schools Categorized By Choice of Family Medicine

The relationships of class size, urban birth, MCAT, older students, and rural birth are more apparent from this table. Across the group, the schools with the smallest class sizes have the highest percentages of family medicine. This is particularly dramatic for the allopathic schools that provide the most total physicians for the US workforce. The allopathic private and public schools graduating the least family physicians have the least diverse class compositions in age and birth origins and the highest MCAT scores. They also graduate the most physician researchers. There are no differences in ethnicity or gender for any of the allopathic groups, other than the three predominantly black medical schools.

Expansion Considerations

As the nation considers expanding medical school numbers and class sizes, there are many considerations. A primary concern should be expanding Essential Workforce in geriatrics, primary care, family medicine, mental health, and women’s health. Any medical school can provide physicians who will choose specialties and locate in urban areas. Only a select few have demonstrated the ability to graduate essential physicians and locate physicians in rural areas.

Medical schools have expressed concerns about the ability to maintain the current training capacity of students (Millenium Conference). For this reason and also as the nation enters an era of rural physician shortage, medical schools need to consider training models such as long term preceptorships. Such models have also demonstrated the ability to give superior training, support to rural physicians, and increased numbers of physicians for family medicine and rural locations. The combination of Duluth’s admissions plus the University of Minnesota Rural Physician Associate Program has been particularly effective. Accelerated family medicine residency programs are the only model that has demonstrated the ability to take the predominant type of student admitted from the most urban origins and graduate 50% into rural practice, even without rural training or rural bias in origin or practice selection. This is a model that needs to be replicated. Instead it has been terminated.

Research Emphasis and Side Effects

It is no secret that allopathic medical schools and deans are obsessed with research. The federal incentives in this area are enormous and the dollars invested in National Institutes of Health (NIH) have continued to expand. The funding for primary care or from state sources pales in comparison to research and graduate medical education funding support for graduating more specialists. Schools also have a primary focus on medical student passage of USMLE 1 boards. The design of the Medical College Admission Test reflects this preoccupation. Schools are well aware of the high correlation of MCAT with USMLE 1 and also connect higher MCAT with the graduation of additional physician researchers. Schools are preoccupied with moving up the NIH rank list, but most have little real opportunity to gain ground. The concentration of dollars and emphasis is also obvious in the outcomes. The top 10 schools graduate over half of the researchers in the allopathic private group.

Medical schools appear to be unaware of the connection between higher MCAT and decreased graduation rates for primary care, office-based primary care, family medicine, and rural practice. They appear to be  unaware of the decline in admissions of  students who are “different” such as older, rural, and instate. They are not working with education and colleges in the kind of nationwide approach that would result in better medical students and better distribution of physicians. There are also indications that improved distribution of education resources can result in more physician researchers, the one area where regression studies for on family medicine and researcher careers align. Students born in the west and Midwest parts of the nation with the best distribution of education resources and the best attitudes toward family medicine choose research and family medicine at high rates. Students born in the south attending medical school rank average in family medicine and poor in research graduation rates. Students born in the east rank average in research but poor in family medicine.

The preoccupation with research has side effects that medical schools and the nation must understand. It is the rare school that has been able to do both, and these rare schools have specific admissions efforts and emphasis for primary care and rural practice.

It appears that medical schools of all types other than the “elite” schools have the distortion of MCAT, board score, and research obsession, but do not appear to share in the benefits of National Instituted of Health dollars.

The Road Less Traveled

Some students are forced to take a difficult pathway to become a physician. They are older, from rural areas, from less urban areas, lower in income levels, and not advantaged in educational access. The numbers who attempt to continue their dream and access international schools may be only a fraction of those who hoped to become physicians. This is one of the real tragedies for education and medical education in the United States. The major obstacles that those who could become physicians face include:

1. Medical school preoccupation with research and graduation of specialists with emphasis on speeded intellect and test taking ability out of proportion to emphasis on people emphasis and service orientation
2. Inability of medical school admissions to understand and adjust for the varied characteristics of students and their potential for meeting the needs of the nation
3. Lack of accountability in workforce similar to the career impacts of the managed care era
4. Failure to address spiraling tuition costs
5. Failure to address liability costs, especially for primary care, mental health, and women’s health. The low income specialties are particularly impacted when students see the unlimited potential for overhead costs eclipsing already low reimbursement rates.

Physician Workforce Studies

Reimbursement and Physician Distribution

www.ruralmedicaleducation.org