Robert C. Bowman, M.D.     rcbowman@atsu.edu

Medical students, medical schools, and career choices can be divided into types that distribute graduates where most needed and types that concentrate graduates into 4% of the land area in top concentrations of physicians, people, and health resources. The types of physicians that are most likely to gain admission are associated with concentrations birth to admission (origin factors) and are found in most needed health access careers at the lowest levels. Those less likely to gain admission are more normal and are less likely to be associated with exclusive concentrations such as concentrations (parent income, most urban origins, concentrations of physicians, concentrations of highest educated, most socially organized) A major theme confirmed over and over is "those with higher probability of medical school admission, have lower probability of most needed health access careers (rural, underserved, primary care, family medicine) while those with lower probability of admission have higher probability of most needed health access when they do get through the substantial birth to admission and admission barriers facing over 65% of American children. Those with higher probability of admission (2 to 10 times) have origins associated with concentrations and especially combinations of concentration (people, income, professionals, physicians) and have half the probability of distribution (0.5 odds ratios) to rural locations, to underserved locations, and to locations with 65% of the US population that only has 20% of physicians. These studies have the highest validity as they are logistic regression studies that include controls for exclusive versus more normal origins, exclusive versus more normal training, exclusive versus more normal (family practice) career choice, and exclusive (younger) as compared to more normal (older age at graduation)  as noted in cross section studies of complete populations of physicians using 12 or more consecutive class years of graduates. Studies can compare medical students to census data to generate probability of admission and changes in medical school admission and census data can be used to project future admission. Each of the populations represented have consistent career choice and practice location that can also be used for future workforce estimates.

 Changes in Numbers of Medical Students Admitted

	Medical School Grads 1994-2000	Medical School Grads 2004-2010	Projected Grads 2014-2020	Change in Numbers 1994 - 2020	Medical School Grads 1994-2000	Medical School Grads 2014-2020
Asian Indian	7785	8788	12241	57.2%	7.0%	7.5%
Chinese	4325	4882	6855	58.5%	3.9%	4.2%
All Asian Students	25580	28876	39171	53.1%	23.0%	24.0%
All Urban Born*	102319	115505	151789	48.3%	92.0%	93.0%
US Urban Born	83412	94162	122411	46.8%	75.0%	75.0%
Foreign Born*	18907	21343	31011	64.0%	17.0%	19.0%
Foreign, Not Asian	8897	10044	13057	46.8%	8.0%	8.0%
Asian Born	10009	11299	14689	46.8%	9.0%	9.0%
US Allopathic Total	111216	125549	163214	46.8%	100.0%	100.0%
White	71178	80351	101193	42.2%	64.0%	62.0%
All Rural Born*	10009	11299	11425	14.1%	9.0%	7.0%
African American	7451	8412	11425	53.3%	6.7%	7.0%
Native American	667	753	816	22.3%	0.6%	0.5%
Low Income Rural*	2780	3139	3264	17.4%	2.5%	2.0%
All Hispanic	6117	6905	9793	60.1%	5.5%	6.0%
Mexican American	3336	3766	5549	66.3%	3.0%	3.4%
By Income Quintile
Top Quintile Income	74515	85373	115882	55.5%	66.8%	71.0%
2nd Quintile Income	18073	19460	22850	26.4%	16.2%	14.0%
3rd Quintile Income	10810	11802	14689	35.9%	9.7%	9.0%
4th Quintile Income	6784	7282	8161	20.3%	6.1%	5.0%
Bottom Quintile	1435	1632	1632	13.8%	1.3%	1.0%

Increases in admission are also found for those with higher probability of admission. Those with average or lower probability of admission are being replaced. This includes American lower and middle income children and those from families in the United States prior to 1980. Newer entry families and children with the most urban, highest income, most professional family origins increasingly dominate admission and have origins also more likely to be in 4% of the land area with top concentrations of practicing physicians.  Massive changes in admission have involved 1997 to current matriculants in allopathic US and osteopathic US schools indicating a much more narrow range of physician origins than ever before in parent income, most urban origin, and multiple demographic factors related to concentrations.

 

Admission Ratios

	1994-2000		2004-2010		2014-2020
	% of US Population 18-24 By Group	Ratio Medical Student % to Pop %	% of US Population 18-24 By Group	Ratio Medical Student % to Pop %	% of US Population 18-24 By Group	Ratio Medical Student % / Pop %
Asian Indian	0.3%	20.90	0.4%	17.50	0.5%	15.00
Chinese	1.2%	3.24	1.2%	3.24	1.3%	3.23
All Asian Students	5.1%	3.18	5.3%	4.26	5.5%	4.36
All Urban Born*	77.0%	1.13	77.0%	1.19	79.0%	1.18
US Urban Born	69.7%	1.04	69.7%	1.08	70.0%	1.07
Foreign Born*	8.3%	1.71	8.3%	2.05	10.0%	1.90
Foreign Born Not Asian	7.5%	0.98	7.5%	1.07	7.5%	1.07
Asian Born	2.5%	3.07	2.5%	3.60	2.5%	3.60
US Allopathic Total	100.0%	1.00	100.0%	1.00	100.0%	1.00
White	67.0%	0.97	64.0%	1.00	62.0%	1.00
All Rural Born*	23.0%	0.57	23.0%	0.39	21.0%	0.33
African American	14.0%	0.51	14.8%	0.45	15.5%	0.45
Native American	1.7%	0.41	1.7%	0.35	1.7%	0.29
Low Income Rural*	9.9%	0.30	9.9%	0.25	9.0%	0.22
All Hispanic	15.0%	0.32	17.0%	0.32	18.0%	0.33
Mexican American	9.0%	0.26	11.0%	0.27	12.0%	0.28
Born Med School County	35.0%	2.00	37.0%	1.95	39.0%	1.90
Not Born MS County	65.0%	0.46	63.0%	0.44	61.0%	0.43
By Income Quintile
Top Quintile Income	20.0%	3.00	20.0%	3.35	20.0%	3.55
2nd Quintile Income	20.0%	1.00	20.0%	0.81	20.0%	0.70
3rd Quintile Income	20.0%	0.53	20.0%	0.49	20.0%	0.45
4th Quintile Income	20.0%	0.34	20.0%	0.31	20.0%	0.25
Bottom Quintile	20.0%	0.14	20.0%	0.06	20.0%	0.05

 

With rapidly changing populations such as Asian Indian populations (immigration, 3rd highest US fertility rate), census figures may underestimate population. Other studies estimate about 10 times greater probability of admission. Native American changes also complicate ratio generation. Wide variations are seen in Hispanic populations as some are more closely associated with top concentrations as in Hispanic foreign born US populations entering for higher education or professional careers.

 

Admission levels follow one other pattern. In the populations admitted at lower levels associated with lower and middle income origins as compared to the most exclusive origins, female medical students gain admission at significantly higher levels. Male African American and male rural born medical students have half the admission probability of their female counterparts. This is a natural experiment that also indicates the birth to admission barriers facing males as noted in studies of education and higher education (but beginning at or before birth). See Education Barriers in the US The divisions in admission seem to be magnified for males who are lower and middle income in origin or those most normal.

 

There is one interruption of this lower and middle income pattern. Hispanic female medical students lag slightly behind male admissions. Cultural considerations are also a reason for lower female higher education and medical school admission; however studies of multiple types of populations indicate another reason. Using a population based count at the county level, there is a lower threshold of 3 - 4 admitted per 100,000 per medical school class year. This is seen in lowest income rural origins as well as counties with top percentages of minorities and lower income. One perspective would be that despite numerous barriers, the United States is unable to suppress admission below this threshold level. The cultural reason is popular and has numerous anecdotal stories of Hispanic females limited in higher education options. The population based interpretation has statistical backing in the populations left behind. Also females have been assumed to have lower rural practice location rates in a number of studies. However in about a dozen states, the states where the rural origin admissions have been female greater than male, higher percentages of female family physicians are found in rural locations. This illustrates the need to control studies for experiential place origin factors as well as experiential place in age, training, and career choice.

 

The birth origin county method has about 3 - 4 admitted per 100,000 per class year. This is about one-third of the US average of 8 to 9 admitted for the 1990s compared to 1970 birth origins. The top range is 14 - 20 admissions per 100,000 per class year in the most urban, highest income, counties with medical schools and top concentrations of physicians. Exceptions confirm the theme of concentrations. Counties with concentrations of people, higher education, or medical students/physicians that are rural counties join the counties with top admission and also graduates with these origins have lower probability of most needed health access.

 

	% of Medical Students 1974-1980	% of Medical Students 1984-1990	% of Medical Students 1994 - 2000	% of Medical Students 2004-2010	% of Medical Students 2014-2020	Change in Per Cent 2000 - 2020
Asian Indian	<1%	<2%	6.5%	7.0%	7.5%	15.7%
Chinese	<1%	<2%	3.9%	3.9%	4.2%	8.0%
All Asian Students	2.4%	6.8%	16.2%	23.0%	24.0%	48.1%
All Urban Born*	80.0%	84.0%	87.0%	92.0%	93.0%	6.9%
US Urban Born	75.0%	77.4%	72.8%	75.0%	75.0%	3.0%
Foreign Born*	5.0%	6.6%	14.2%	17.0%	19.0%	33.8%
Foreign Not Asian*	3.8%	3.8%	7.3%	8.0%	8.0%	9.2%
Asian Born	1.2%	2.8%	7.7%	9.0%	9.0%	17.4%
US Allopathic Total	100.0%	100.0%	100.0%	100.0%	100.0%	0.0%
White	85.0%	81.0%	65.3%	64.0%	62.0%	-5.0%
All Rural Born*	20.0%	16.0%	13.0%	9.0%	7.0%	-46.2%
African American	6.2%	5.8%	7.1%	6.7%	7.0%	-1.0%
Native American	0.3%	0.4%	0.7%	0.6%	0.5%	-27.9%
Low Income Rural*	3.5%	3.2%	2.9%	2.5%	2.0%	-32.0%
All Hispanic	2.0%	4.0%	4.8%	5.5%	6.0%	26.1%
Mexican American	1.3%	1.7%	2.3%	3.0%	3.4%	47.9%
Born Med Sch County	64.0%	67.0%	70.0%	72.0%	74.0%	5.7%
Not Born MS County	36.0%	33.0%	30.0%	28.0%	26.0%	-13.3%
By Income Quintile
Top Quintile Income	58.0%	59.0%	60.0%	67.0%	71.0%	18.3%
2nd Quintile Income	20.0%	20.0%	20.0%	16.3%	14.0%	-30.0%
3rd Quintile Income	10.0%	11.0%	10.5%	9.7%	9.0%	-14.3%
4th Quintile Income	8.0%	7.0%	6.7%	6.1%	5.0%	-25.4%
Bottom Quintile	4.0%	3.0%	2.8%	1.3%	1.0%	-64.3%

 

Admission breaks down into 5 different levels.

• 300% admissions - The top level of admissions involves students born and raised in concentrations of people, education (major university in a rural area), professionals (education or federal facility, corporation research headquarters), income, and health facilities (medical school, major medical center). High ratios are also found in the rural counties with medical schools, major medical centers, or major universities. About 6 medical schools have rural locations.A few of the most rural counties from North Dakota to Minnesota to Iowa to Nebraska to Kansas manage to admit top levels of medical students.
• 150% admissions - The second level of admissions remains above average and the students involved have some but not all of the factors involved in top level admissions. Many are again in locations in the nation with broader distributions of education and income.
• 80% - 120% admissions - Average admissions is possible for larger rural areas, micropolitan areas, and those with education and income characteristics similar to urban areas. Rural population growth has often been adjacent to steadily expanding metropolitan areas. Rural females and black females have admissions probability at the lower end of this range. Certain Midwestern states with the best distributions of income (lowest Gini Index, lowest quintile income ratios) have rural admission at the national average for admission (ND, NE, KS, IA, MN, WI). This is a level twice the rural average for the nation. There are indications of declining rural admissions even in these states, especially when crossing from majority rural to majority urban. See Changing Rural Admissions (includes graphics 1940 and beyond)  Certain foreign born populations (not Asian, unless lower income Asian) have average to below average admission.
• 40 – 80% admissions - Lower than average admissions levels are the usual when the grouping is all rural students, all African American students, rural origin males, and African American males. Admissions are not improving for this group over time. Although the levels shown are averages, higher levels of admission follow higher status as in all groups. Also at the lowest status levels, few medical students gain admission. Social organization levels are lower for this group, but not at the lowest levels. Barriers of education, income, geographic distance, social distance, and culture are present. Distributions of resources from federal and state resources are lower and local ability to generate funding for education and opportunity are also lower.
• Below 40% probability of admissions – Combinations of barriers of income, education, race, ethnicity, culture, language, and other factors are likely in this group. The bottom 20% of the nation in income only manages about 2% of medical school admissions. Probability of admission is most commonly misinterpreted as a barrier impacting the immediate step before medical school admission - college. This again is a narrow view. The life experiences that shape students away are far before. The Stanford studies by Barr in Academic Medicine did confirm a leaky pipeline with African American and Hispanic freshmen students actually accomplishing their goal of applying to medical school whereas 100 - 110% of white and Asian students continuing in the path to medical school and applying. Similar findings are seen for lower and middle income compared to top status in a number of studies. Rural, Mexican American, African American, and other populations represent concentrations of lower and middle income origins and when they do, they also represent lower probability of admission, greater probability of first generation to college, greater choice of family medicine, and greater probability of most needed health access careers. Stanford is not alone in selectivity away from lower and middle income. The bottom 25% of the nation in income only manages 3% of college positions in the top 146 colleges, an important stepping stone to medical school admission. This step is even more important as US medical schools are moving to an ever narrowing smaller group of exclusive colleges for admission. There is no end to the movement toward more exclusive in so many dimensions and standardized testing is involved directly or indirectly. Fewer poor children are even taking college admission exams as noted by SAT exam data. College preparation is lowest in these children. Private schools and schools in top property value school districts clearly have greater college preparation, higher percentages entering college, higher percentages entering the more selective colleges, and better performance on standardized tests related to all of these factors as well as their parent influences. Performance on the first standardized tests in second grade reflect differences regarding verbal and other areas shaped by higher status or professional parents. Studies have illustrated these divisions for decades although few have paid attention. In more recent years the nation's top educators have warned the US and US leadership multiple times without response. Only 5 - 8% of Native American and African American children have a 50% probability of making an A or B on their first college math or science course, important for all professional careers. Difficulties can be tracked back to birth and early childhood. Short of an entire focus from child development to early education emphasizing age 0 – 8, the nation will not make much progress in any lower or middle income population.  The children who work their way through such environments to higher education and medicine have the highest levels of primary care, family medicine, rural, and underserved choice in the nation. They are also more likely to become the teachers, nurses, public servants and all who protect, defend, stabilize, and serve the nation where most needed. They are the human infrastructure that determines an efficient or effective nation or one that fails. See Infrastructure America

 

Origins, Admissions Deficits of Physicians, Admissions, and Practice Locations

 

Basically the same forces shape admission and shape physician distribution, not surprisingly in the same magnitudes. The same forces also shape family practice choice and therefore sustained primary care levels, or not.

 

Various pieces of data are missing due to lack of data availability. Parent income data would be outstanding, but is poorly collected and not even used in studies that demand its use, such as debt and tuition studies of career choice. MCAT scores of individual physicians would work since they so closely reflect combinations of concentration, but these are not available. Every school, type of student, or population associated with higher concentrations has higher MCAT scores and greater probability of admission to more exclusive schools, greater probablity of more exclusive careers, and greater probability of concentration in practice locations away from 65% of the American people.

 

Overall the pattern can be filled in. The consistent theme is concentration. Those attempting admission that are associated with higher concentrations or combinations of concentration are admitted at higher probability. They are also less likely when admitted to be found in underserved, primary care, rural, and family practice careers.

 

Graduates associated with lesser concentrations are less likely to gain admission but when admitted are more likely to be found in most needed health access careers.

 

Birth Origins	Medical Student From	Ratio of  Admis-sion	US Pop 2000	Ratio of Distri-bution	% of Physi-cians	Rural Location	All Under- served	FPGP Choice
Major Med Center Location	70%	1.5 – 3.0	33.7%	2.18	73.4%	10%	2-5%	2 – 12%
Top Quintile	70%	3.5	20%
Medical School County	69%	1.5	50%	1.2	60.2%	7.7%	4.5%	12.3%
Foreign US MD Grads	15%	1.5	10%			5.3%	4.9%	10.5%
Asian*	20 - 23%	5 – 5.5	4.2%			6%	5%	10%
Asian Indian*	6.5%	10.7	<0.5%					2.2%
DC Area  MSA	3.34%	1.25	2.67%	1.45	3.9%	8.0%	4.4%	11.7%
NYC Area MSA	14.92%	2.2	6.80%	1.1	7.4%	4.7%	3.2%	8.5%
Major Med Center Dominant County	0.5%	2.16	0.2%	2.2	0.5%	13.8%	4.4%	17%
White	65%	0.94	69.0%					15%
No Med School in County	32.3%	0.60	54.0%	0.74	39.8%	15.2%	6.8%	18.9%
Urban Underserved	4–6%	0.2 – 0.6	10.7%	0.28	3.0%	6-10%	15-20%	14–20%
African American*	7.1%	0.48	14.8%
Historically Black MS						7.9%	10.3%	18.4%
Mexican American*	2.3%	0.22	10.4%
Bottom Quartile  Income County	7.13%	0.62	11.5%	0.76	8.7%	12.1%	8.2%	17.3%
Urban Whole County PC Shortage	0.28%	0.27	1.04%	0.20	0.21%	15.0%	10.1%	20.0%
All Rural Born	10%	0.5	20.0%	0.49	9.8%	46%	8%	23.6%
Rural Underserved	<1%	0.25	7.1%	0.38	2.7%	6-10%	14–18%	20-29%
Rural Whole County Shortage	1.32%	0.39	3.36%	0.26	0.86%	23.3%	14.1%	24.9%
Bottom Quartile Income County	8.2%	0.56	14.4%	0.48	7.0%	24.1%	9.8%	24.4%
Rural Commuting County	0.49%	0.20	2.42%	0.18	0.45%	22.1%	10.4%	21.0%
Predominantly Black	0.39%	0.65	0.61%	1.0	0.63%	11.0%	14.8%	20.4%
Predominantly Hispanic	0.15%	0.51	0.29%	0.66	0.19%	8.4%	17.4%	29.2%
Predominantly Native	0.10%	0.48	0.20%	0.15	0.03%	100.0%	68.2%	27.3%
Isolated Rural	1%	0.25	4.2%	0.26	1.1%	18-24%	14%	28.0%
Isolated Underserved	< 0.5	0.1 - .2	2.0%	0.2	0.4%	28%	16-20%	48.0%

* The seven class years of 1994 - 2000 US MD Grads were compared to the seven year age group of census population for 18 – 24 year olds (those closest to medical school age).  Other than these, the comparisons include 1987 – 1999 medical school graduates

 

The consistent relationship between probability of admission and probability of distribution suggests that the same factors are involved in both. Nations and states that do poorly in the birth to admissions process are likely to do poorly in distribution of physicians.

 

Health access recovery begins with lower and middle income children that become the physicians with 2 or 3 times greater return to lower and middle income populations in need of health care. Health access recovery continues with medical schools that admit such children, train for health access, and graduate the most family physicians. Family practice choice doubles underserved choice above birth origin, age at graduation, and training. Family practice choice triples rural choice above origins, age at graduation, and training. A focus on family medicine in birth to admission, admission, training, career choice, and policy restores health access. Exclusive origins, exclusive training, exclusive career choice, and exclusive policy all work to concentrate physicians in 4% of the land area in top concentrations.

 

The World of Rural Medical Education	New Web Site - Basic Health Access
Physician Workforce Studies	Infrastructure America Website
Birth To Admission: Education Aspects	Culture Health and Access

 

www.ruralmedicaleducation.org