Ethnicity, Gender, Admissions, and Distribution of Physicians

1994 - 2000 Class Years	Probability of Admission for those of Medical School Age					FP Choice	Family Medicine Grads 1997 - 2003		Parent Factors
United States Allopathic Graduates	One out of ___ males are medical students	One out of ___ females are medical students	US Age 18-24 (Census)	Medical Students 1994-2000 (AAMC)	Male/ Female (AAMC)	(FP Grads in Masterfile compared to AAMC Class Yrs)	Rural Choice in FP Graduates	Under-served Choice in FP Grads	Approximate Money Income Levels 2003 (Census)	Parent Income Level of Accepted (AAMC)	Parent Income Level of Applicant (AAMC)	MCAT all applicants 1996 (AAMC)	MCAT Accepted 1996 (AAMC)
Asian Indian	18.8	20.4	159236	8136	0.53	2%	15%	13%	over $55,000	100000		9.70
Chinese	41.2	44.6	208868	4882	0.53	6%	6%	6%	$55,000	80000		10.30
All Asian	48.9	53.0	1034000	20340	0.53	7%	11%	6-13%	$55,000	90000	80000	9.70	10.6
Vietnamese	65.7	71.2	97196	1424	0.53	22%	5%	10%	30 - 40k	42500		9.10
All Urban Born	183.9	176.7	19691600	109228	0.5	13%	21%	10-13%	higher	higher
US MD Grads	206.9	198.8	25466000	125549	0.5	18%	24%	5.4%
White	200.6	226.3	17413000	81973	0.54	14%	26%	10-13%	$48,000	90000	80000	9.50	10.3
Native American	270.2	240.6	222000	871	0.48	9%	47%	27-30%	$33,000	60000	55000	8.10	8.9
All Rural Born	462.7	321.9	6218400	16321	0.42	22%	30%	14-20%	in lower	lower
Black	589.6	305.0	3593000	8880	0.35	13%	11%	19-22%	$33,000	55385	50000	6.90	8.4
All Hispanic	488.4	597.2	3204000	5975	0.56	12%	12%	14-30%	$33,000
MexicanAmerican	967.4	1222.1	3110600	2887	0.57	19%	13%	21-24%	$33,000	50000	48000	8.00	9.1
Other Hispanic	931.7	1139.3	2400000	2346	0.56	7%	11%	14-15%	$33,000	70000	60000	8.60	9.7

Admissions varies with densities of income, education, and professional occupation of parent; proximity to medical school, and additional barriers. Population density, education, and income levels of the county of birth of medical students predict well over 50% of the variance in admissions. These socioeconomic measures also influence test scores such as SAT/ACT and MCAT. Higher socioeconomics means low choice of distributional careers and poor retention for practice in the state where students attended medical school. Family medicine is a more likely choice in all but the top 30% in status and in allopathic public schools, FPs are 44% more likely to be retained in their state for practice (1987 - 2000 medical school graduates compared to 2005 locations in the Masterfile). Data on ethnicity was not shared other than in family medicine but family physicians are a reliable two times as likely to be found in rural or in underserved areas compared to the same types of medical students across similar demographic origins or medical schools.

Origins help determine career choice and practice location. Those with major medical center origins such as birth in a city or a county with a medical school are most likely to end up in major medical centers. Those born in urban areas tend to return and those born in rural areas have greater rural distribution. Distribution levels are greatly facilitated by family medicine choice. Facilitating Physician Distribution

Those with origins beyond major medical centers return to practices away from major medical centers. Rural origins improve rural location. White and Native graduates have the only significant percentages of rural origins.

Underserved is determined by zip code locations and can also be categorized by rural or urban. Underserved coding included:

Equal proportions of designated and not designated sites were involved and both had the same number of people in poverty, physicians, and ratios of people, poverty, and physicians. The zip code determinations were smoothed for consistency with adjacent zip codes (this only involved 128 physicians out of 200,000). Zip codes with NHSC or CHC zip code categories as the sole determinant were rejected

Other locations were military, international, urban served, rural served, urban major medical center, and rural major medical center

Background: Certain types of students are more likely to distribute to careers and locations in great need. Student ethnicity, gender, income, and age are considered regarding medical school admissions, choice of specialty, and practice characteristics.

Methods: This study compared medical students graduating from 1994 - 2000 to census data to generate ratios of admission by ethnicity and gender. The medical students were also compared by choice of career and location, age distribution, and various birth origin characteristics.

Results: The average medical student was admitted at a ratio of 1 medical student out of 200 total citizens or residents in the United States age 18 to 24 years old. Asian Indian citizens or residents were admitted at the highest ratio of 1 out of 20, Native Americans 1 out of 280, rural born 1 out of 360, black female 1 out of 300, black male 1 out of 600, lower income rural born 1 out of 760, Mexican American male 1 in 1000, and Mexican American female 1 in 1300. Rural born, Vietnamese, Mexican American, and low income origin students were most likely to choose family medicine, rural locations, and office-based primary care in poverty locations.

Discussion: Admissions levels and physician distribution varied across population density, income level, education level, age, and MCAT scores. Those of lower socioeconomic level had the lowest levels of admissions but greater choice of distributional careers.

The composition of medical students in the United States has been changing in recent decades. There has been a slow decline in white students, males, rural born students, and those born in the same state as their medical school. Each of these groups is more likely to choose family medicine and rural practice. The most dramatic increases involve students with parent income levels of over $100,000 which increased 74% or from 3765 to 6560 between the 1997 and 2001 matriculating classes. This was a period dominated by news headlines noting the reversals of affirmative action. Visits to universities by affirmative action reversal activists were followed by threats of lawsuit or actual lawsuits.

Asian students have increased to 3500 out of the 16000 matriculants or 22%. Asian students represent a group with the highest levels of parent income, education, professional degree, (MIM) and urban origin (census, Bowman birth) Asian and urban born students represent the youngest medical students. (Bowman birth) The nation has also entered a new era with parity in gender composition. Studies have noted that females do not choose rural practice as often as males. (ellsworth)

Medical students who are born in counties with lower levels of income, education, and population density (Bowman birth) are more likely to choose family medicine. Students from lower income origins are more likely to choose family medicine (Cooter, Bowman birth) and they also have higher attrition rates (Cooter). Senior students interested in rural practice have the greatest family medicine choice at 68% and they tend to be older, married, from rural high schools, and have greatest levels of interest in serving the underserved at 60% levels. Rural interested seniors are not likely to be Asian, Black, or Hispanic. (bowman rural interested) Those are most different in ethnicity and income are also more likely to have plans to serve the underserved at 40% compared to 10% for average sutdents. (AAMC MIM)

Comparing gender, ethnicity, and career choices can give important information about future workforce.

The data sources that were compared included the number of medical students from the seven years of classes of 1994 – 2000 by ethnicity and gender (AAMC MIM, AAMC warehouse), seven years of U.S. citizens and residents by ethnicity and metro/nonmetro location who were age 18 – 24 years (Census 1990), parent income levels (AAMC MIM), income by ethnicity (Census 2000), family medicine graduates by gender and ethnicity, and graduate practice locations (Masterfile, Robert Graham Center). The Masterfile has ethnicity data on 14686 of 18230 or 80.5% of family physicians who attended U.S. allopathic medical schools. Rural practice location was determined by Rural Urban Commuting Area coding using the practice zip code selected by OfficeMax software. Poverty designation included a zip code that was associated with the National Health Service Corps designation, a Community or Migrant Health Center, or a population with over 20% in poverty.

Those most likely to gain admission have higher income levels, higher MCAT scores, and the most urban origins.

Table of Allopathic US Medical Student Admissions by ethnicity, income, rural vs urban

Vietnamese students are an exception. Despite the lowest parent income and education levels, Vietnamese students are admitted at levels greater than average and equal to Asians of much higher income levels. Vietnamese make choices of family medicine at higher levels consistent with other lower income level students.

Choice of rural location and rural poverty location is related to rural origins. When comparing the percentage of rural born students from a school to the percentage of graduates choosing rural locations, there is a 0.86 to 0.92 positive correlation for typical medical schools.

Medical Schools By Choice of Family Medicine for 1987 – 2000 Graduating Classes

Studies of birth origins by county also confirm a linear relationship between county of birth income levels and choice of family medicine (Birth Origins and FP Choice).

Ethnicity appears to play a role in rural practice location for family medicine graduates. Rural birth origins are a major reason but older graduates also choose rural locations.

White family physicians are basically the only ethnicity with significant rural origin numbers and provide rural family physicians in the highest numbers. Native American family physicians choose rural practice in the highest percentage, but less than 10 per year go on to choose family medicine. Native males from rural areas may have admission ratios as low or lower than rural white males and black males.

Regardless of ethnicity, family physicians born in rural areas have approximately the same 40 - 50% choice of rural location; however there were only 118 black family physicians born in rural areas and less than a third of this number for other ethnicities. Admissions from the constraints of rural birth plus lower income level plus diverse origin may represent the greatest challenges to any potential medical student and maybe to a young professional of any type. Such students add great versatility to physician workforce since they can an do choose any specialty and location, especially when choosing family medicine.

It is important to understand that rural practice distribution is limited for all other medical specialties by the availability of patients, facilities, and call coverage issues. Family medicine distribution is essentially the same as the distribution of population.

Even for family medicine there is a slight decline in choice of the most isolated rural communities compared to population distribution. The various medical school types also impact the distribution of family physicians. Again this is a matter of the types of students that they admit. This involves older coding based on 20% poverty levels and not adjusting for zip codes with 75 or more physicians.

There are groups with much better outcomes. Inner city programs graduate 60 – 70% into urban poverty practices. These programs include concentrations of Black and Mexican American female family physicians born in urban areas. Rural training tracks have 70 – 90% rural location choice, but tend to have rural training bias, rural birth, and older age. The 150 accelerated family medicine graduates in this group led in total underserved distribution with 55% of graduates in rural locations, 25% in urban poverty locations, and 10% in teaching positions and do not have birth origin or training location bias. Accelerated graduates did have selection based on maturity levels and dedication to family medicine. (Accelerated Family Medicine Training Programs).

The students most likely to distribute to the most needed locations as physicians are the least likely to be admitted to medical school. These students have origins in counties and populations with lower income levels, lower population density, and lower MCAT scores. They are the least likely to gain admission to the prestigious colleges and medical schools and specialties.

The changing composition of physicians in the United States predicts worsening distribution to rural and lower income areas of the nation. Students born in lower income counties and populations have the greatest choice of distributional careers but are admitted at lower and lower levels. This means less choice of family medicine, fewer rural physicians, and fewer likely to choose office-based primary care in poverty locations in future years.

The male advantage in rural practice choice is countered by a female advantage in urban poverty location. Female senior students actually have more interest in rural practice but they are less likely to be married or engaged. (Rural interested) Female career choice may be more sensitive to graduate location and spouse issues. The gender differences for rural practice choice are not universal. Females in the most distributional programs in the nation choose rural locations in the same pattern as males, including the Physician Shortage Area Program at Jefferson, the potent combination of Duluth admissions and Rural Physician Associate Program training, and the maturity-based selections and training of the accelerated family medicine residency programs. Unfortunately the accelerated model has been terminated nationwide, Duluth was forced to merge, PSAP has been refused much deserved state lines of funding, and these rural medical education programs have had 30 – 50% reductions in the number of students signing on in recent years.

Gender comparisons offer important clues regarding physician distribution. In some states where access to education in rural areas lags the most, there are family medicine programs where females are more likely to become rural family physicians. Upon closer examination, the female graduates were born in rural areas and the males were born in urban locations. In these states the males in rural parts of the state access college or medical school poorly. Across the nation rural males and black males are much less likely to gain admission compared to females.

Service-orientation is an important component of primary care choice (Madison Service Orientation) the expression of service orientation may vary by ethnicity and gender and longitude. Those from eastern inner city poverty locations tend toward internal medicine and pediatrics while those in the Midwest and west tend to choose family medicine. There are few that can escape chronic poverty. Those that do have devoted parents, elite test taking ability, and participate in elite programs. Those with elite scores are treated much the same as those who are privileged and are less likely to choose family medicine. There is a difference for Vietnamese, rural populations, and Mexican Americans. These populations are not enmeshed in extremes of poverty lasting generations. There are more who are the first or second generation in their family to attend college. Those most involved in chronic poverty in the US have low achievement scores, disrupted family structures, low high school graduation levels, and the lowest college participation rates. Chronic poverty also breeds hopelessness and low expectations of professional careers that may kill off important service-oriented careers as teachers, nurses, public servants, and family physicians long before college. Studies indicate major problems with education funding, distribution of education (Funding Gap) and college access (Mortensen) for those in chronic poverty. It is not a surprise that those delayed entry by barriers of education and income are older and tend to choose family medicine, rural practice, and poverty practice.

Medical School Admissions By Income - see what increasing income levels of admitted allopathic students are likely to do to the FP match in just a few years.

A special acknowledgement goes to health professional advisor Dan Marien who died recently.

Cooter R, Erdmann JB, Gonnella JS, Callahan CA, Hojat M, Xu G, Evaluation & the Health Professions, Economic Diversity In Medical Education / September 2004 http://ehp.sagepub.com/cgi/reprint/27/3/252

Madison, Donald L Medical School Admission and Generalist Physicians A Study of the Class of 1985, Academic Medicine Vol 69 Number 10 October 1994 p 825 - 831

Martini JM, Veloski JJ, Barzansky B. Medical school and student characteristics that influence choosing a generalist career. JAMA . 1994;272:661 -8 Lower mcat

McGaghie WC Assessing Readiness for Medical Education Evolution of the MCAT JAMA September 4, 2002 Vol 288 No 9

McGaghie, W. C. (1990). Qualitative variables in medical school admission. Academic Medicine, 65, 145-149.

Medical Marketing Service I. AMA Physician MasterFile: Medical Marketing Service, Inc; 2000 and 2004, access to 2004 data via Robert Graham Center for collaborative work

Rabinowitz HK, Diamond JJ, Markham FW, Paynter NP, Critical factors for designing programs to increase the supply and retention of rural primary care physicians, JAMA 2001; 286: 1041-1048

Seyan K, Greenhalgh T, Dorling D The standardised admission ratio for measuring widening participation in medical schools: analysis of UK medical school admissions by ethnicity, socioeconomic status, and sex, British Medical Journal 2004;328:1545-1546 (26 June), http://bmj.bmjjournals.com/cgi/content/full/328/7455/1545 .

Xu, G, et al. (1997). The relationship between the race/ethnicity of generalist physicians and their care for underserved populations. American Journal of Public Health, 87(5), 817-822

Wheat, JR, Brandon, JE, Carter, LR, Leeper, JD, & Jackson, JR. (2003). Premedical Education: The Contribution of Small Local Colleges. Journal of Rural Health 19(2), 181-189.

Family Physicians from Allopathic Schools 1997 – 2003 Residency Grads	Major Medical Center Zip	Birth in Medical School County or City	Core Metro Born (area over 1 million in 1993)	Birth in a Rural Zip (RUCA)	Non-Metro Born
Allopathic National Data
Totals for Non FP	71%	70%	40%	8%	8%
Totals for FP	43%	59%	38%	22%	21%
Family Medicine Data for All Types
All 1997 – 2003 Grads	34%	59%	44%	22%	21%
White F	35%	58%	43%	24%	21%
White M	30%	54%	34%	28%	28%
Black F	39%	72%	61%	10%	9%
Black M	37%	64%	58%	12%	9%
Mexican American F	40%	59%	65%	9%	6%
Mexican American M	27%	78%	59%	15%	13%
Asian F	43%	78%	70%	8%	7%
Asian M	37%	80%	71%	8%	8%
Puerto Rican F	44%	100%	63%
Puerto Rican M	41%	75%	29%
Other Hispanic F	46%	67%	65%	13%	12%
Other Hispanic M	33%	70%	59%	9%	11%
Native American F	21%	27%	33%	47%	41%
Native American M	20%	33%	31%	36%	31%
Other F	40%	63%	52%	18%	16%
Other M	34%	59%	35%	25%	26%

Divisions by Quartiles/Most or Least FP Choice (n)	Medical School Grads 2004	% Born Urban Inf 1	Foreign Born	Born Urban Inf 3-9	% FP Likely	% Age 30 and older	MCAT 2000 Average
Allo Priv Least FP (11)	112.5	63.1%	14.4%	4.2%	19.3%	14.3%	10.9
Allo Priv Next Least (11)	141.0	60.9%	14.4%	4.7%	21.9%	16.2%	10.5
Allo Priv Next Most (11)	120.1	56.6%	13.1%	4.8%	20.9%	19.6%	9.6
Allo Priv Most FP (8+3)	115.2	51.9%	15.8%	8.7%	30.7%	23.2%	9.4
Trad. Black (3)	75.4	53–82%	4-10 %	7–14 %	14-38%	21-33

Allo Pub Least FP (20)	139.1	56.0%	13.1%	5.5%	25.3%	17.3%	9.8
Allo Pub Next Least (21)	146.0	46.7%	11.4%	10.5%	40.7%	21.4%	9.5
Allo Pub Next Most (20)	127.0	41.7%	9.9%	16.4%	47.5%	25.6%	9.6
Allo Pub Most FP (20)	92.0	36.9%	7.7%	22.7%	54.1%	29.2%	9.3

Osteo Least FP (9)	161.3	47.5%	8.3%	4.4%	22.4%	33.6%	8.4
Osteo Most FP (8)	133.0	33.3%	7.3%	13.6%	39.8%	42.9%	8.2

Allopathic Medical School Family Medicine Graduates	Rural Birth Origins			Rural Practice Location
	Female	Male	All	Female	Male	All
White	17.7%	20.0%	19.0%	23.7%	28.3%	26.3%
Black	10.8%	9.6%	10.4%	9.9%	11.7%	10.5%
Mexican American	7.0%	10.6%	9.0%	9.3%	14.6%	12.3%
Asian	3.8%	1.9%	2.8%	7.7%	7.5%	7.6%
Other Hispanic	8.1%	4.6%	6.0%	13.2%	8.3%	10.3%
Native American	56.7%	36.4%	44.6%	46.7%	36.4%	40.5%
Unknown/Other	12.5%	14.6%	13.5%	17.0%	24.8%	20.8%
All Graduates	14.9%	17.0%	16.0%	19.7%	24.7%	22.4%

Medical School Graduation Year	US Population 1998	1987 - 2000 FPGP	1987 - 2000 Office Based Primary Care Not FPGP	1987 - 2000 All Physicians Not FPGP
Urban/Urban Focused	77.6%	73.9%	89.7%	89.4%
Large Rural	9.3%	10.5%	5.9%	6.3%
Medium Rural	6.9%	9.7%	2.8%	2.6%
Isolated Rural	6.1%	4.7%	0.8%	0.9%

Source	Urban FP in Poverty	Rural FP	Rural FP in Poverty	Military	All Poverty	All Rural, Urban Poverty, Military	Total FP
Allopathic Private	876	643	154	275	1030	1783	4196
Allopathic Private	20.9%	15.3%	3.7%	6.6%	24.5%	42.5%	1
Allopathic Public	2343	3182	763	767	3106	6288	12894
Allopathic Public	18.2%	24.7%	5.9%	5.9%	24.1%	48.8%	1
Osteopathic Private	329	499	113	188	442	1016	2172
Osteopathic Private	15.1%	23.0%	5.2%	8.7%	20.3%	46.8%	1
Osteopathic Public	96	182	46	42	142	320	689
Osteopathic Public	13.9%	26.4%	6.7%	6.1%	20.6%	46.4%	1
North American International	269	228	62	17	331	513	1240
North American International	21.7%	18.4%	5.0%	1.4%	26.7%	41.4%	1
Distant International	405	277	70	21	475	703	1787
Distant International	22.7%	15.5%	3.9%	1.2%	26.6%	39.3%	1
Total	4318	5011	1208	1310	5526	10623	22978
Total	18.8%	21.8%	5.3%	5.7%	24.0%	46.2%	1