Robert C. Bowman, M.D.
There is often much discussion regarding the ability of family medicine and rural practice to bounce back, especially in the U.S. Although there are cycles related to health policy, there are long term deteriorations. My hat is off to those who have fought so hard for improved health policy (support, reimbursement, inclusion, etc.), especially those who have succeeded such as in Toronto. The following is a piece I wrote for the state FP legislative list serve. One of the rural docs in the Midwest was wondering when the situation would return (like a pendulum)....
Do not believe in pendulums or any sort of return to a more advantaged position without significant effort. Consider a particle moving in a vortex, viewed from the side it looks like a pendular motion, but it is being sucked down with increasing frequency. Viewed from above it moves in a circle, never getting anywhere in a relative sense, but moving faster and faster, until lost from sight. Viewed from below it is not seen, until it suddenly appears and disappears. We have to plug the drain and let the level of primary care rise. It will take admissions changes and health policy to plug the drain, and education and admissions to increase the flow of family medicine likely students.
1. Admissions of the lowest income students (most likely to choose FP, PC poverty) to allopathic medical schools declines by 400 each year (out of 16000 matriculants), they are replaced by 400 of the highest income students who are the least likely to choose family medicine, rural practice, primary care, and poverty locations. Only 40 - 50% of the high income students recognize significant health access problems in the nation compared to 80 - 90% of the lowest income students (AAMC Minorities in Medicine XI). We are slowly headed toward physicians who have less and less of the abilities and attitudes most needed. In Canada there have been similar notations of loss of the lowest income students. In the US the lowest income quartile students choose FP at 22% compared to 13% for the highest quartile Cooter R Economic Diversity in Medical Education The range of choice is similar for rural vs urban born (20+% vs 10%) and older vs younger (13% for those less than 30 and 23% for those over 29 at graduation)
2. Admissions of Instate Born Medical Students (those with the most connections to a state) have declined at slightly less than 1 percentage point a year for the last 40 years. Instate born students have a 30 - 50% greater choice of family medicine, are the most likely to be retained in their state of birth or state of medical school for practice, and are more likely to choose poverty primary care locations. Admissions of rural born students has also been declining at slightly less than 1 percentage point a year. Older students decline with increasing national applicant pool numbers Older Medical Student Career Choices Pools are up lately and are related to increasing college degree numbers. Rural and older student applications (as measured by MCAT takers 1991 - 1999) have been a constant 13% and 24% for these applicants, they are just not being admitted. Rural and older students have 50 - 100% greater choice of family medicine. Birth Origins and FP Choice States also tend to do the same with college funding, scholarships and admissions, favoring those with the highest scores that are the most likely to leave the state with their human capital investments. Understanding Higher Education and Income
3. Health policy favoring primary care was able to cure maldistribution from the supply side for the 1995, 1996, and 1997 graduating classes with 750 rural fp a year and 750 additional office-based PC in poverty locations (900 each entering for all sources from US and international). Health policy also built primary care choice from 28% to 41% for the allopathic classes graduating 1965 to 1978 (creation of Medicare and Medicaid and superior funding then). This health policy impact also increased office based primary care in poverty locations for allopathic docs from 2.4% to 4.4%. In the 1990s reforms, the office based pc in poverty hit a record 5.6% of all allopathic physicians (data as of Masterfile 2004 locations). We are now down below 3% office-based PC in poverty in the US and the office based primary care totals are nearing pre-Medicare levels. Medicaid cuts and state level changes (co-pays, restrictions on prescriptions, decreases in Medicaid eligibility) appear to have been particularly destructive, especially for rural and underserved practices.
4. In the current US health policy, there will be increasing jobs for specialists and subspecialists until health care costs topple businesses and governments and continue (with rising prison costs) to squeeze education out of state budgets. More black males of college age are in jails and prisons in the US than are in college. Rural males are increasing in the prison category and decreasing in the college category at rapid rates. Black males, rural males, Hispanics, and Mexican Americans are admitted at rates approaching 1 out of 1000 compared to 1 out of 200 for all whites, urbans, and US average. Asian college age students range from 1 in 60 for Vietnamese to 1 in 20 Indian-Pakistani citizens or residents gaining medical school admission. US foreign born admissions to allopathic medical schools are 16% of the entering class of 16000 with rates up to 30% in the eastern most urban schools. Indian-Pakistani students choose family medicine at 2.2%, the lowest level of any group determined so far. Rural interested senior medical students are interested in service to the underserved at levels of 60% compared to 40% for black and Hispanic and 11.5% for 1995 US allopathic medical school seniors.
5. The Side Effects of Selecting for Family Medicine (admissions of rural born, lower income, older students and those born in counties of less than 1 million) result in more rural docs, more office based primary care docs, more office-based pc docs in poverty, and more psychiatrists. The side effects of selecting for research or subspecialty (early admissions, younger, higher MCAT, urban, foreign born) are more subspecialists, more doctors for counties of over 1 million, increased economic impact of health on the most urban areas, increased health care costs, decreasing quality, decreasing access, increasing health disparities, and likely worsening state budgets in education and other areas. The correlation between allopathic medical school admissions of rural born students and graduation of rural physicians of all types is +0.92. (look closely, you will likely never see such a correlation ever again). It is no surprise that with the continued declines of rural born students, that allopathic schools have had declining graduation of rural physicians for decades and this will continue.
6. I have not seen any work where any movement occurred from specialty care to primary care, only the reverse. See Primary Care Retention to see how this even impacts NPs and PAs. The retraining specialist efforts were a failure in the 1990s with perhaps the single exception of those who first trained in primary care (family medicine) before specializing. Family medicine has had a net gain of residents from other specialties in the first two years for decades. Even the forced choice of family medicine in the managed care classes (1994 - 1998) has not had excessive attrition rates. About 1 percent of 1997 - 2003 FP Residency Graduates have been lost to fellowships related to FP and another 1 % to other specialties. Retention Within the Specialty of Family Medicine
7. Family medicine also multiplies the distributional tendencies of students, in other words, rural born students choosing fp multiply % choice of rural locations (double or more), older students (50% increase), black and Hispanic students, especially female, 60% rates of poverty location. Accelerated graduates had the most dramatic increases with over double the choice of rural and underserved urban compared to those from the same program, the same state, or national FP figures. Accelerated FP graduates did 3 years med school and 3 years FP residency at the medical school location and had 50% total choice of rural (50%) plus 30% choosing urban poverty locations (zip codes with underserved center or at least 20% of the pop in poverty). There was no better distributional primary care program however this model was terminated at the hands of national and FP accrediting bodies. For internal medicine, pediatrics, the distributional (rural, older) students have a 0.7 - 0.9 multiplier (divider??) for distributional choices. Choice of IM or Peds means losing ground in rural vs urban and no enhancement of poverty location. I do not have data on ethnicity except for FP so cannot comment in this area and non-FP choices. This impact on distribution again may have much to do with lower income student preferences for FP and also more likely to choose poverty locations, at least when graduating from most US medical schools (not the most exclusive ones, however).
Also just did some work comparing the 22 elite mostly eastern, highest MCAT, highest research graduation allopathic schools compared to the 104 other allopathic schools. In the elite 22 schools, the distributional students did not distribute, in the 104 the rural and older students and instate born did choose FP, PC, poverty locations, psych, at higher percentages. Those observing from the perspective of the 22 schools or those most familiar with these schools and less so with schools further west may truly believe that the distributional students (rural, older, lower income) do not distribute! Those most familiar with the schools in the Midwest and South are painfully aware of these differences.
The AMA Masterfile also has data on research choices as well. In the elite 22 schools, the younger, foreign born, and urban chose research in greater numbers (but older graduates still chose research in higher percentages, nearly double that of younger), in the 104 schools the younger, foreign born, out of state, and urban did not choose research in higher numbers. Admissions of students for the 104 schools in the hopes of more researchers did not work and likely distorted workforce toward subspecialty care and urban location. Even in the 22 elite schools that graduated more researchers, the graduates were the most likely to end up over 1000 miles away (correlation between higher % researcher and % over 1000 miles away is +0.44). Researcher retention (keeping a research graduate within your walls to do research or at least within 60 miles) is less than 25% for all schools and researchers (1987 - 1994 graduates of allopathic schools compared to 2004 locations of graduate). Research By the Ages
A 1 point increase in the MCAT score for a medical school class is associated with a gain in physician researchers by 1 additional researcher every other year. A 1 point increase will decrease family physician production by 6 - 9 family physicians each year. The bioscience MCAT score is up 1 point in the past decade. It is a bit of a stretch but this may well predict 6 - 9 fewer FP physicians graduating from a school with such an increase compared to 10 years ago. The MCAT scores held up when using any one of 5 sets of MCAT data from 2000 - 2003, and when using MCAT rank instead of score. MCAT rank order has likely not changed in decades. MCAT has a +0.72 correlation with graduation of researchers and +0.79 with NIH dollars, +0.52 for GME funding, and +0.5 - 0.6 for % graduation of orthopedists, ophthalmologists, ENT, dermatologists, and careers related to general surgery and subspecialties. Those not attending an elite medical school have 30 - 50% less chance of becoming one of the MCAT related subspecialists. Correlation of MCAT with foreign born %, Asian %, out of state % is over +0.4 - 0.5.
MCAT is correlated with choice of FP at a school -0.64 (higher MCAT means fewer FP graduates) and rural FP at -0.62, -0.54 for rural Peds and rural IM, -0.42 for ob-gyn, -0.38 for anesthesia. MCAT CorrelationsThe correlation between choice of FP at the state level is +0.62 compared to a state ratio of HS Graduation in 1986 divided by college graduation rates. A state with breadth of education and greater access and emphasis at the lower levels is likely to graduate family physicians and primary care physicians. A state with depth of education (gifted, merit scholarships based on scores not income levels, higher college graduation rates) is likely to graduate researchers and subspecialists. Bright Future Rankings
This, my friends, is not a pendulum. If medical schools do not understand admissions for graduation of more researchers, it is clear that they do not understand admissions at all and are not likely to invest the necessary resources without major changes. The graduation of family physicians, rural physicians, psychiatrists, and researchers involves admissions of dedicated, mature, versatile, people that absolutely depend upon life experience and education (FP and high school graduation rates, researchers and college graduation rates). In each of the 4 groups it is the older graduates who are more successful. There is little doubt that the complexity for each career required is similar and that such careers are the most demanding in medicine. For each of these complex careers each new day is learning and application, for other physicians, most (not all) are increasingly separated from the primary sources of learning, their patients. The separations from patient care, learning, and true societal needs are most dramatic for those involved in teaching and administration.
From the recent AAMC Physician Workforce Conference: The US continues to recruit over half of family physicians from other nations to fill residency positions. The lower the socioeconomic status of the nation, the more the tendency tend to come to the US. The US also recruits from UK, Canada, Australia, and other developed nations, forcing them to also recruit more from nations with less health and education resources. Medical educators in Ghana are actually proud of the fact that half of their physicians are in the US when they should be doing their best to retain health and education resources in their nation. Canada should expect massive increases in recruitment efforts from US rural and primary care and academic sources. If there is a boom stock market any time in the next 2 years, experts predict massive retirements of US academic physicians. US medical schools are increasingly turning to faculty in other countries to teach in US medical schools.
Robert C. Bowman, M.D.
rbowman@unmc.edu