The archiving and publishing of the proceedings of the third and fourth plenary sessions of the First National Conference on Primary Health Care Access (April 21, 1990) is made possible, in part, through the generous support of the San Joaquin General Hospital Department of Family Medicine (Stockton and French Camp, California):
John E. Midtling, M.D., M.S. Chair, Department of Family Medicine, Medical College of Wisconsin (Dr Midtling is a Senior Fellow of the Coastal Research Group): I am going to talk about the differentials that David Sundwall referred to, particularly as they relate to physician reimbursement, how that might impact upon our ability in primary care to attract more students into our specialty, and how that might affect our ability to finance our residency programs. Greg Nycz is going to talk about some of the rural differentials.
I think our ability to manage access to primary care depends upon our ability to attract students into the primary care specialties and ultimately to place them in geographic areas that are appropriate. I want to describe the magnitude of the problem that we’re facing in primary care.
Consider the national trends in numbers of office-based family physicians. In 1963, there were about 67,000 family physicians in the U.S. and today there are about 5,000 less. In terms of a percent of the total physicians pool, the decline has been from around 27% down to 10.8%. There has even been an absolute decline, despite a doubling of the physician population.
Even when you factor in the other two primary care specialties – and I use here the federal definition of primary care being family practice, internal medicine, and pediatrics – the decline is from 49% of all total physicians down to 34%. You can see that there has been virtually no growth in total numbers.
Although there has been some expansion in the numbers of internists and pediatricians, this has not offset the decline in the number sof family physicians. And in 1963, there were 53 primary care physicians in the U.S. For every 100,000 people. Today there are just 52 per 100,000. This goes along with what David Werdegar was saying, that we have not done a good job as academic health institutions in producing the numbers of primary care physicians that our society requires.
I think the problem is likely to get worse because it’s been widely recognized that there is a bimodal distribution of family physicians in the U.S. There is one group of residency trained family physicians with a median age of around 35 and a second group of general practice era family physicians with a median age of around 60.
Today, one out of three family physicians is over 55 years of age. There are 24,000 family physicians over 55, and we’re only producing 2400 new graduates each year. So simple arithmetic tells you that as these over-55 cohorts move out into retirement over the next decade, they will not be replaced at the rate of production of our new graduates. I believe that this is going to have a serious impact on rural areas in particular, because family physicians account for 65% of all office-based physicians in rural areas.
In the metropolitan areas of the U.S., we have 56 primary care physicians for every 100,000 people and in the non-metropolitan statistical areas just 42 for every 100,000. But the family physician cohort represents by far the majority of primary care physicians practicing in the non-metropolitan statistical areas of the U.S.
There have been studies which have shown that over the past ten or twelve years there has been about a 45% decline in office-based primary care physicians practicing in our underserved urban areas. So any decline in the numbers of family physicians, I believe, will severely impact the rural areas and the underserved urban areas.
Furthermore, the increases in managed care systems which have found primary care physicians – especially family physicians – to be effective utilizes of resources, has drained family physicians away from the traditional rural ares into more of the urban-suburban locations. This may further exacerbate the impending shortage.
It is clear that we must increase the numbers of family physicians in the U.S., if we are to impact the access problem. However, this has been difficult to do because we suffer from both a shortage of positions and applicants. In fact, we now have fewer residents in family practice than we did in the peak years of 1983 and 1984.
Between 1984 and 1987, the number of first year positions in family practice declined by 11%, the number declined in internal medicine by 6.8% and the number declined in pediatrics by 4.1%. Studies have shown that it is difficult to fund more more than one-third of program costs with patient care revenue. One-third must come from hospital support and the remaining one-third must come from external subsidies.
However, Medicare indirect cost payments have been squeezing the hospitals. There has been a ratcheting down already from 11.59% to 7.7% in the so-called Medicare indirect cost adjustment. This ratcheting down has occurred through a percent adjustment applied for each .1 increase in the resident-to-bed ratio in a hospital. HFCA is now proposing that this be further ratcheted down to 4.4%. If that occurs, that will have a marked impact on hospital graduate medical education payments.
Of the $3 billion paid last year for graduate medical education, roughly $2 billion was in the form of the indirect cost payments. Not only that, the administration is now proposing that the direct cost payment be capped. The direct cost payment is calculated by taking the resident’s salary, fringe benefits, and malpractice costs, teaching and administrative costs, and other costs associated with the residency programs and multiplying that by the percent Medicare says.
The proposal is that the direct payment be capped at the 1987 average resident salary of $25,574, updated to 1991 by the Consumer Price Index, to create a cap at $29,688. Although a 1.8 multiplier would be applied for primary care, this would still provide primary care programs with a direct graduate medical education payment of roughly $10,000 less in funding per resident than was provided as a national average in 1988. Furthermore, Medicare patients participating in managed care systems are excluded from the graduate medical education payment system.
I believe this differentially penalizes primary care programs located in community hospitals that participate in managed care systems. At the same time our federal training grants have been reduced in size.
Since 1978, there has been a decline of about $10 million in available monies for the federal training grants in family practice. Similar trends have been noted in general internal medicine and pediatrics. When you factor in the Consumer Price Index, there has been almost a two-thirds decline in monies available for family practice graduate medical education.
Thus, I believe, the funding of our training programs is becoming overly dependent upon patient care dollars. I feel this is an undesirable situation since it is clearly more cost-effective for hospitals to train residents in the procedurally oriented specialties where hour for hour they may earn up to 15 times or more the fees generated by residents in the primary care specialties. In addition, the cost of maintaining the model clinic in family practice has been an additional cost that the hospitals must bear.
I believe if we are to expand our programs, we need to expand these Title 7 payments for primary care education. We also need to factor into the Medicare graduate medical education cost payments the nation’s need for the specialty being trained. I believe that some sort of multiplier is needed for direct cost payment which would more than compensate for the ratcheting down of the indirect cost payment.
Finally, it’s possible that full implementation of the Resource Based Relative Value Scale [RBRVS] may help increase the percentage of residency training program costs generated through patient care dollars. If the projections are correct, family physicians could undergo a 37% increase in reimbursement and internists a 14%.
Data from Phil Lee’s commission modeled what would happen at full implementation of RBRVS (not what would happen to total family physician income, but what would happen to Medicare-derived income which accounts for about 29% of all physician income in the U.S). They have projected that internal medicine would get about a 14% increase; family practice a 37% increase; opthalmology and thoracic surgery – a 16% increase in reimbursement and in the largely rural areas about a 31% increase.
Since many of our programs are located in the small metropolitan or even rural areas, it’s possible that this could help increase practice plan revenue, increasing the financial stability of our residency programs.
Primary care training programs have also suffered from an increasing shortage of applicants, causing many programs to reconsider plans to expand or actually downsize. I believe any attempt to increase the numbers of primary care physicians in the U.S. must also address predoctoral education programs so as to influence specialty choice among U.S. medical students.
I am sure you’ve all seen the charts as to what has happened to the National Intern and Residency Match in primary care over the past decade. I’ll talk about what happened with the 1990 Match as well. Between the years 1978 and 1988 there was a decline in the ability of family practice programs to match with the U.S. seniors from around a fill rate of about 77% to about 61%. Pediatrics had a similar decline; and internal medicine virtually an identical decline to family practice, falling from 77% down to 63%.
However, general surgery and the surgical sub-specialties have increased their match from about 66% of positions to 80%. OB-GYN, despite student concern about the malpractice issue, has undergone a substantial increase, from 76% to 86%. Some of the surgical sub-specialties, like orthopedics, ENT, have been able to match in excess of 90% of their positions with U.S. seniors.
In 1990, family practice matched only 59.3% of its positions with U.S. seniors and internal medicine only 59% However, the figures would have been even worse, because in 1990 family practice offered 100 fewer positions than in 1988, and internal medicine offered 300 fewer positions than they did in 1988. So we’re offering fewer positions and doing less well in the Match. I think this trend is going to continue for several more years and I’ll show you why.
The Association of American Medical Colleges (AAMC) surveys U.S. seniors at the time they graduate from medical school. You can see over a five-year period of time there has been about a 5% reduction in student interest in family practice; 8% reduction in general pediatrics; and a 40% reduction interest in general internal medicine. However, look at this! A 39% increase in the medical sub-specialties. If these students who graduated in 1987 follow through on those career plans, there will be further sub-specialization in internal medicine.
Other data, based on AAMC surveys done at the time students take the medical college admissions test, are even more worrisome. The AAMC asked those students who subsequently matriculate at a U.S. medical school as to what specialty in which they intended to practice. Over a ten year period, there has been a 56% reduction in student interest in family practice as a career; a lesser decline in internal medicine; but a tremendous increase in student interest in surgery and the surgical sub-specialties.
These students that were surveyed and entered in 1987 will graduate in 1991. IF they follow through on these career plans, I would anticipate a further decline in student interest in primary care.
Well, many have tried to analyze these trends. Some have suggested it may be due to a movement way from the traditional values of social concern and service to an increasing emphasis on financial well being. Certainly studies of undergraduate students have shown this to be the case. Others have suggested students’ increasing concern about their ability to services large educational debt burdens with a career in primary care.
According to the AAMC, the average debt of graduating medical students in the U.S. increased from $5,000 in 1971 to $35,000 in 1987. This represents a 700% increase in student debt over a 15-year period, far outstripping the Consumer Price Index. What this doesn’t tell is that in 1987, 1 out of 14 medical students in the U. S. had educational debt burdens in excess of $100,000.
During the same period of time, there was a marked reduction in the federally subsidized loans for the health professions and the near elimination of the National Health Service Corps scholarship program. Most studies which have looked at the relationship between student debt and specialty choice, I believe, have been flawed because they have not differentiated between subsidized and unsubsidized student debt.
The guaranteed student loan (so called GSL), the national direct student loan (NDSL), and the health profession student loan (HPSL) are all federally subsidized loans. These must be differentiated from the health education assistance loan, the so called HEAL loan.
Unlike the other loans, HEAL debt interest begins accruing immediately at prevailing rates. Loan repayment then begins after the training period but interest accruals markedly increase the size of the principal to be repaid. In some cases initial loan amounts may double by the time the student begins repayment.
Subsidized loans became less available during the 1980’s. More students were froced to take out HEAL loans. Bazzoli, in a study done for the AMA in 1984, was able to show that for every $10,000 increase in HEAL debt burden, it decreased the likelihood of a student entering primary care by 7.5 percentage points. In fact, Bazzoli was able to model as early as 1984 that primary student reliance on HEAL debt at then prevailing debt ratios would decrease the number of students selecting primary care as a career by 1,597 students per year.
Unfortunately, Bazzoli’s mathematical models were never considered and policy extensions were made which implemented the policies of which Bazzoli’s mathematical experiment forewarned. The decline in student interest in primary care as a career could have been predicted and anticipated based upon the policies to fund medical education implemented in the 1980’s.
We now have a very unfortunate situation where HEAL debt burdens are so large in some cases that even the National Health Service Corps loan repayment program cannot service the student debt in many cases, to say nothing of reducing the principal.
While accruing ever larger unsubsidized debt burdens, students face the prospect of declining earning power for the primary care specialties. The normal laws of supply and demand have not influenced primary care physician earnings, since reimbursement rates are fixed by third party payers who reimburse for procedural services at a rate many times that paid for cognitive services.
Like other payers, Medicare has actually exacerbated the primary care access problem by paying sub-specialists more than primary care physicians for the same work, reimbursing urban physicians at a higher rate than rural, providing complete coverage for inpatient care, but requiring patient contributions for outpatient care, reimbursing physician time spent in undertaking a procedure at a rate several times that paid for time spent providing ambulatory care or cognitive services and in many cases not reimbursing for preventive care at all.
Not surprisingly, these differentials have encouraged physicians to choose specialties and practice locations already well supplied. In fact, the gap in income between the primary care and non-primary care specialties has been growing throughout the past decade. If you look at what a family physician earned in 1977, he earned 82% of average physician income. That same individual in 1986 earned just 68% of average physician income. Internal medicine declined from 98% to 91%; pediatrics a decline from 76% to 68%.
If you look at a rural physician in 1977, that individual earned 95% of average physician income. However, in 1986 that same individual earned only 86% of average physician income. If you break it down according to specialty, the gap is even more dramatic. Over a less than 10 year period of time, there has been a 15%-20% decline in family physician earning power as a percent of average physician income; similar declines in pediatrics and internal medicine; but in many of the procedurally oriented specialties, there has been almost a 20% increase in earning power as a percent of average physician income. So, over the decade of the 80’s, the gap in income between primary care and non-primary care has been growing dramatically.
Will RBRVS correct these differentials or will it be too little too late? I believe RBRVS holds great promise to attract and retain more students in the primary care specialties. However, will it augment earnings enough to make a difference? In some cases, as David indicated, we’re looking at two or threefold or more differentials in earning power. Will the preoccupation with the federal budget deficit prelude implementation as planned? HCFA now has a single spigot that can be cranked down.
Enactment of the administration’s current Medicare budget proposal would speed up payment cuts expected under the resource based payment system and use them for 1991 budget savings. Such a use of expected savings for budget reduction would preclude the redistribution of savings to primary care in rural areas as intended.
Phil Lee, Chair of the Physician Payment Review Commission, has recently gone on record as stating that such reductions could severely limit the funds available for crucial payment increases for primary care specialties and for rural areas. That may hasten the retreat from the primary care specialties and the rural areas, further exacerbating the access problem. It seems to me it is critical that we strongly advocate full implementation of RBRVS with savings directed to correct the primary care and rural differentials as originally intended.
I believe that medical education reforms are also necessary to increase the numbers of students selecting primary care as a career. Attention should be given to selecting applicants for medical school with a primary care preference. New efforts must be made to recruit more under-represented minorities into medicine.
There are fewer underrepresented minorities in medicine today that there were ten years ago. Another focus must be the medical school curriculum. Major new primary care curricular changes need to be made, especially at the third year level, during with most students make career decisions. Education in the ambulatory setting is costly and labor intensive.
Such a redirection of educational experiences would require substantial expansion of primary care faculty who are based in the ambulatory setting. Major new predoctoral training incentives, I believe, need to be established through Title 7 programs to provide the resources necessary for these curricular changes.
Perhaps we need to implement a primary care deficit reduction act modeled after Gramm-Rudman-Hollings where state and federal support for medical education would be base upon the production of at targeted number of primary care physicians. I believe that for family practice a reasonable target would be 25% of all U.S. graduates and for the three primary care specialties combined a reasonable target would be 50% of all medical graduates.
I believe that as the major payer of medical education, government can exert considerable influences on the specialty distribution of available training positions.
Finally, I believe that we as primary care educators have a unique role ot play in health services research. I believe that there are many places such as the State of Oregon which is trying to establish a Medicaid rationing program. This program is crying out for research data, too look at what works and what doens’t, what are cost-effective solutions, and what will benefit the greatest number of people. I believe that we in primary care need to provide a leadership role in developing this research agenda, which has great potential for developing strategies to control health care costs and improve access.
Ultimately, if we are to improve our American system of health care, we must be able to allocate resources in a way that is consistent with a well-thought out, national policy applied in a consistent and longitudinal fashion. We must advocate for the application of such a policy and we must learn from the mistakes of the past.
This presentation was preceded by: First National Conference on Primary Health Care Access (3rd Plenary Panel, Part 1, Sundwall)
This presentation was followed by: First National Conference on Primary Health Care Access (3rd Plenary Panel, Part 3, Nycz)