Family Medicine Standards and Constants

Robert C. Bowman, M.D.  February 15, 2007

For 1975 - 1999 Class Years for those born in the United States of America and becoming physicians:

• For normally distributed populations about 1 in 100,000 born in the population becomes a family physician each year.
• 25% Maximum choice of family medicine - those with more are working to have more
• 7% Minimum choice of family medicine - those with less are working to have less
• Increasing rural or lower income admissions will increase physicians to rural or lower income areas
• A plain demographic factor increase such as non-specific rural or lower income admissions will likely not increase family physicians, since the increases will predominantly involve those not choosing family medicine.
• Increases in family medicine by admission must involve specific attention to the individual characteristics related to family medicine such as service orientation, overcoming obstacles, breadth of scope rather than narrow concentration.
• Family medicine choice results in two or more times distribution to all locations outside of major medical centers.
• Family medicine is the only form of primary care that can demonstrate such facilitation when controlling for birth origins such as rural or lower income birth.
• Primary Care Retention is unique and specific to family medicine and has been so for decades. It is the only dependable source of primary care, with or without health policy.

There are at least 5 important ramifications of these constants.

1. Allopathic medical schools or states graduating greater than 25% have managed to modify admissions or other factors in ways that concentrate those likely to become family physicians in the school, state, or county.
2. Family physicians somehow manage to resist the obstacles to admission to a greater degree than other types of physicians.
3. Medical schools graduating no family physicians or fewer than 7% have modified admissions and training in ways able to overcome even the incredibly strong ability of family physicians to become family physicians.
4. One strategy for admissions is to increase rural origin or lower income origin medical students. This is a fine strategy for increasing physician distribution, but not family medicine choice. Family medicine remains about 1 per 100,000. With broader admissions more are admitted who do not choose family medicine. Distribution of specialists is likely to improve given the broader range of origins, but limited by training and health policy.
5. The strategies used by successful schools and programs involve selecting students based on the individual student characteristics related to family medicine, not just rural or lower income origins. Service orientation, obstacles overcome, previous career experiences, broad versus narrow focus, and older age are just a few factors considered. This is a largely unstudied area, sad because it is also most relevant to physician quality for all physicians.
6. Health policy can modify the 1 per 100,000 constant somewhat. The constant does require the existence of family medicine and reasonable levels of support for primary care and for health care outside of major medical centers. Before 1965 and perhaps in the current era, predictions become unpredictable.

One per 100,000 for Recent Physicians 1987 - 1996 Class Years

Out of a normally distributed population of 100,000 each year, about 6 – 7 medical students will gain admission. Within this group about 1 out of this 100,000 will choose family medicine each year. For populations with the lowest probabilities of admission in the nation related to the greatest barriers of income, education, and social distance, this 1 family physician will be joined by 3 or 4 others not in family medicine for a ratio of 25% for choice of family medicine. The single family physician will be joined by 8 to 11 other physicians per 100,000 in the highest status populations with 6 – 9% choice of family medicine.

Overall admissions probabilities and admissions of those not choosing family medicine are greatest for those most closely associated with concentrations of income, education, professionals, and people.

County Level Admissions Table

Admissions of those found in family medicine share the significant correlations with education and professional factors, but the correlation magnitudes are greatly reduced. In admissions of those who become family physicians, income and population density are no longer significant. High school graduation rates retain more of the original strength.

Ratios of college to high school graduates or a measure of breadth of education (high school graduation greater) outcome rather than depth (college or above greater) have even higher correlations with family medicine as do other measures of concentration such as Gini indexes of income, highest income quintile divided by lowest income quintile, highest educated to least educated, or ratios of professionals to lower ranking occupations.

Broad distributions of education, income, and health resources favor family medicine. Increased concentrations favor increased specialists. When admissions levels are lowest, family medicine admissions are also among the lowest. However the proportion choosing family medicine is greatest.

Maximum 25% family medicine proportion by origins

When comparing family physician admissions to total admissions for lower income, rural origins, or various low status demographic factors, the maximum choice of family medicine is about 25%.

The environments with the lowest rates of admission of medical students who will choose family medicine are the most inequitable in the nation, including the lowest income counties, the 84 counties with poverty levels over 30%, lower income counties adjacent to major metro areas (urban influence 1993 codes 4 and 6, counties designated as commuting counties), and combinations of the most rural and lowest income counties (whole county primary care shortage areas). Only the most extreme barriers decrease admissions of family physicians to levels below 0.8 per 100,000. The environments that are the origins of the highest concentrations of family physicians at 1.2 – 1.5 per 100,000 include major university towns and 200 counties with the highest levels of family physicians in the nation with a ratio of 60 serving a population of 100,000 people.

Minimum 7% family medicine proportion by origins

Again with normal distributions of population, in the most urban or highest income populations about 7% or 1 in 14 will be found in family medicine. With origins associated with stacked concentrations of professionals such as in the New York City or Washington DC areas or in small towns with major universities or concentrations of the most highly educated in the nation (Huntsville AL, Los Alamos NM) the family medicine proportion arising from these origins can be as low as 5%. Allopathic medical schools graduating fewer family physicians have biased selections or training in ways that prevent choice of family medicine. Medical schools graduating no family physicians or less than 5% have altered the distributions.

The counties with lowest total admission rates also had some of the lowest rates of admission of those who became family physicians. Despite this, these are counties with higher ratios of family physicians of those who did gain admission. County types with lower total admissions include counties adjacent to larger urban areas and many of the same counties characterized as “commuting” counties, counties with high poverty or primary care physician shortages, counties with older populations, or states without public medical schools

Some states admit greater levels of physicians with broader distributions of education and income, a true “No Child Left Behind” approach. These are also largely Midwestern states that lead in rural high school graduation rates and have significant rural populations. Fewer lower and middle income children are excluded from opportunity.

Extremes of Lowest Admission

Lowest Admissions per 100,000 Population in the Nation, Relative Preservation of admissions of those who will become family physicians.

 Numbers refer to Urban Influence Codes (1993, Parker, Ghelfi)

National Standards for Distributional Workforce for the 1987 – 1999 US MD Grads

National averages for rural workforce at 11% and underserved workforce at 5.4% serve as standards. Those found in rural areas at greater than national levels have rural origins or have chosen family medicine. Those found in underserved areas at greater than national levels have underserved origins or have chosen family medicine. Older age at medical school admission or graduation is related to greater levels of underserved and of rural distribution, additive beyond birth origins or choice of family medicine.

Increased Probability of Admission Means Lower Probability of Distribution

Those distributing to rural areas at greater levels have lower probability of medical school admission. Those distributing to underserved areas at greater levels have lower probability of admission. Those with lower probability of admission are most geographically and socially distant from those found in major medical centers. Those distributing at the highest levels share origins with those in most need of health care. Those most likely to choose family physicians and US MD Grads older at medical school admission also share lower probability of admission and higher levels of distribution.

Progressive Increases in Rural Distribution By Origin, Older Age, and Family Medicine Training

Rural distribution is 5 – 8% for urban and highest income US MD Grads, specialist physicians, or graduates of top MCAT medical schools. Rural distribution improves to 10% for generalist primary care internal medicine and pediatrics. About 11% is the national mean for rural workforce for US MD Grads. Older age graduates and high access dependent general specialties (general surgery, orthopedics) have 15 - 17% rural distribution. Physician assistants also have 17% rural location. (PA) Urban born family physicians have had a consistent 20% rural location. Rural location for the first family medicine graduates began at 30% and settled into a range of 22 – 26% for the last 30 class years. The mean for all US MD Grad family physicians is 24.4%. Distribution increases to 23% for older urban born family physicians, 43% for rural born FPs, and nears 50% for older rural born FPs, accelerated FPs, and Native American FPs. The range for rural graduate training in family medicine begins with 45% for programs in urban and less rural areas and extends to 70 – 90% levels for programs with predominantly rural location and focus. The physicians found in more and more rural areas in 2005 are increasing combinations of family medicine, older graduates, lower income graduates, and rural origin US MD Grads from the entire 1987 – 1999 class year period.

Progressive Increases in Underserved Distributions By Origin, Older Age, and Family Medicine Training

The national norms for underserved location are 5.4%. Underserved areas are not in major medical center or military zip codes. Underserved areas are zip code practice locations with 20% of the population or nearest population in poverty or share a common zip code with a Community Health Center, National Health Service Corps site, or whole county shortage area. Underserved distributions are 3 – 4% for urban, highest income, youngest graduates, foreign born US MD Grads, secondary or tertiary physicians, or graduates of top MCAT medical schools. Primary care internal medicine and pediatrics levels are 5 – 6% with underserved contributions limited to urban underserved locations. Family medicine levels begin at 10% for higher status types and average 12% for all family physicians. Family physicians have top levels of urban underserved distribution and dominate rural underserved distribution. Underserved distributions are 18% for accelerated family medicine graduates, increase to 20 – 25% for Black or Mexican American family physicians, and increase to 30% or more for Native American family physicians and inner city family medicine residency programs. Diverse socioeconomic origins, geographic origins, age, and family medicine choice combine for the highest levels of underserved distribution. Older age at graduation, birth in a lower income county, and male gender each add a few percentage points to underserved distribution.

Family Medicine Choice Contributes More To Distribution Than Origins

It is important to note that both family medicine training and specific family medicine training programs (rural or underserved) are associated with even greater levels of distribution compared to birth origins influences. No other form of primary care has demonstrated this facilitation effect. The effect may have been present in physician assistants before the last ten years of losses in rural, underserved, and primary care location. Family medicine accomplishes facilitated distribution by remaining 99% active, 98% in family medicine, 90% in office primary care, and 50% outside of major medical centers. This is what allows the specialty to remain 24% rural and 12% underserved for decades of class years.

Logistic regression equations involving the 189,000 US MD Grads (93%) with a birth origin and practice location in the 50 states (also not military connected) reveal the facilitating effect of family medicine career choice. The odds ratios for rural locations and rural underserved locations extend to 3.5 times for choice of family medicine and this contribution is greater than rural origin, lower income origin, or older graduate contributions at 1.5 to 2 times. The odds ratios for urban underserved locations are as low as 2 times for family medicine, but still exceed birth origins and age contributions. Distribution to predominantly Black rural counties is doubled by family medicine choice, but factors connected to Black race do contribute at 3 to 7 times, specifically birth in a predominantly Black nation (3 times), graduating from a Historically Black medical school (4 times) or birth in a predominantly Black county in a rural area (7 times). Family medicine retained the doubling effect or tripling effect with fully loaded equations or even overloaded equations with redundant variables.

Combinations Concentrating on Family Medicine

Duluth medical school selects for family medicine and rural practice, trains with rural family physicians at the highest levels, and has the highest levels of family medicine and rural choice consistently. Duluth Plus RPAPOsteopathic schools have similar emphasis and also admit older graduates. The effects of a focus on family medicine and older graduates is best illustrated by at least 12 specific training programs that were once spread across the nation. Over 132 accelerated family medicine residency graduates completed 3 years of medical school and 3 years of residency between 1997 and 2003. This is a group not hampered by selection bias or training bias. Accelerated graduates were not selected by program directors for rural or for underserved practice or trained for these careers (except one rural track with outcomes no different than the other programs). There are 3 unique characteristics of accelerated graduates: early commitment to family medicine, readiness for a challenging curricula, and maturity. Their maturity is reflected specifically in their older age range. With this comes significant life experience and other careers prior to medicine. They were 56% older than 29 years at medical school graduation (compared to 28% for FP and 22% for all allopathic graduates). This was a training program with the lowest costs, with the shortest duration, with top level academics, and with optimal career and location outcomes. Accelerated Family Medicine Training Programs

Optimal Physician Career Outcomes

Rather than have various outcomes compete or overlap in confusing ways, a new field can be created in the database to assess all optimal outcomes. An optimal outcome graduate was found in 2005 in a small or isolated rural location (not micropolitan), in a full time academic career, in a rural or urban underserved location, or in a military career. Top levels of optimal location at 40% or more were shared by family physicians who were Native American, those born in whole county shortage areas, medical schools selecting most specifically for family medicine, and accelerated graduates. Family physicians as a group average 24% optimal outcomes and other types of physicians have 18% or less levels of optimal location or career choice. It is hard to find a group that does not contribute to optimal outcomes in family medicine. Some groups in family medicine contribute more to small or isolated rural location, some contribute more in urban underserved areas, females do more teaching, and males do more military, rural, and underserved.

Competitive Health Policy Limits Distribution

Specific health policy efforts have limited maximal distribution to rural and to underserved areas. Black male, Hispanic male, and rural male family physicians with the highest levels of optimal distributional outcomes are also found in military careers at 2 to 3 times the 4% family medicine average. The older ages of these distributional types suggests marriage, family, and previous barriers of income and education. The military support package is clearly the best for medical students in these circumstances. Black female, Hispanic female, and Native American female family physicians are found in academic careers also at 2 or 3 times higher levels compared to the 4% in teaching careers. Better support packages for rural and underserved location similar to the excellent military support package could shift the most distributional types of family physicians to the most needed distributional locations. The point made is not which is the “right” career, but which career is most desired by a particular physician, since a career choice that is preferred is likely to lead to retention and less need for support programs in the future.

During supportive health policy periods family medicine training dispersed to more distributional locations such as inner city, Community Health Center, and a variety of rural locations.

Barriers related to health policy involving graduate medical education could be overcome and training location for medical students and residents could provide additional rural and underserved workforce. Accreditation guidelines could be restructured to disperse all medical education to a greater degree in the hope of better distribution of specialists and primary care physicians. In many states, medical education already provides significant contributions again with the major limitation being health policy involving primary care and support of care outside of major medical centers.

Health Policy and Impact on Family Medicine Choice and Physician Distribution

Five Periods of Health Policy and Physician Career Choice  When categorizing physicians and family physicians by class year, there have been two periods of time associated with maximum physician distribution and choice of family medicine. These are 2005 Masterfile location and career choices. During the class years of 1965 – 1978, US MD Grads increased in primary care, family medicine, rural location, and underserved location. Primary care numbers quadrupled, a result of a doubling of primary care numbers and a doubling of medical school class size at the same period of time. Rural percentages increased with family medicine the major contributor. (*Note: the current 30% increase in class size without significant increase in choice of family medicine will not improve rural or underserved distributions, especially during the current health policy.) During the period from 1990 – 1995 Medicaid doubled, managed care forced choice away from hospital careers to primary care, primary care reimbursement reforms and increases were significant and predictable, and subspecialty reimbursement declines were just as predictable. The United States peaked family medicine choice for the 1995 – 1997 class years. This group of US MD Grads has maintained maximal rural locations, maximal primary care levels, and maximal underserved location, again a result of family medicine increases. The levels of US MD Grads found in rural locations by class year stabilized for the first time in years. Since this period of time the cost of delivering health care has dramatically increased, the cost of college and medical education has dramatically increased, the nation and the nation’s medical schools have decreased emphasis on primary care and family medicine, the nation has reduced distributions of support for lower and middle income areas of the nation (areas outside of major medical centers involving family physicians to a greater degree), and a variety of dramatic reports have contributed to declines in choice of primary care, family medicine, and rural location. Internal medicine graduate retention in primary care is at record low levels along with family medicine choice in US MD Grads. Physician assistant and nurse practitioner activity levels and primary care levels continue to decline. Underserved location for US MD Grads is as yet unchanged but the impact of loss of obligated personnel is yet to be seen.

The United States has entered another chaos period similar to the time period before 1965 when the base of support provided by health policies was insufficient for primary care, rural, and underserved locations. Five Periods of Health Policy and Physician Career Choice


Robert C. Bowman, M.D.
UNMC Department of Family Medicine Director of Rural Health Education and Research
983075 Nebraska Medical Center
Omaha, NE 68198-3075
(402) 559-8873 or fax at -8118
Email: [email protected] 
 

Restoring America by restoring Young Professionals to Underserved Areas

Family Medicine Central: National Comparisons of Workforce

Distribution of Physicians - Medical schools with distribution levels listed, family medicine is the major contributor

Head to Head: Physician Assistants in 2000 Compared to Family Physicians in State and National Location

Instate Retention of Family Physicians: Dependable Primary Care Workforce Retained Within States

Physician Workforce Studies

www.ruralmedicaleducation.org