waiting for the doctor glut, or is the cavalry really coming?

6
PERSPECTIVES Waiting for the Doctor Glut, or Is the Cavalry Really Coming? GENE A. H. KALLENBERG, MD, RICHARD K. RIEGELMAN, MD, PhD, LISA I. K. HOCKEY, MD SINCE 1980, a great deal has been written about the "doctor glut" predicted by the final report of the Graduate Medical Education National Advisory Committee (GMENAC). 1 As one medical journalist recently put it, "... the words 'doctor' and 'glut' have become as wedded in medical vocabulary as 'coronary' and 'bypass.'"2 Unfortunately, one of the report's essential conclusions, that the predicted glut is for specialty physicians and not for primary care physicians, has received far less attention. 3-5 In re- sponse to the predicted glut many medical school administrators have already reduced or are pres- ently considering reducing the sizes of future classes. 6 Many of us who toil away in the field of primary care continue to feel that our country is still seriously deficient in its supply of properly trained primary care physicians. We are especially concerned by the rush to reduce the number of medical students before real solutions to the problem of the dispropor- tionate mix of specialists and generalists are in place. We feel that this approach, combined with the present administration's attempts to cut federal funding for all postgraduate medical training, could result in a serious decline in the future annual pro- duction of primary care physicians both in absolute numbers and in relation to the numbers of specialists. Reconsideration of some widely accepted facts about primary care physician supply and demand may help to ensure that our policies are appropri- ately focused. SUPPLY Primary Care Supply Forecasts-- Overly Optimistic? In 1980 the GMENAC report predicted that by 1990 there would be an overall excess number of physicians and that the excess would be concen- trated predominantly in a variety of (sub)special- ties.~ The committee based its conclusions on extrap- olations from then-current data on medical student, resident and physician populations (supply) plus es- timates of existing and future levels of patient utiliza- Received from the George Washington University Health Plan, 1229 25th Street, N.W., Washington, DC 20037. Address correspondence and reprint requeststo Dr. Kallenberg. TABLE 1 Percentages of Active PhysiciansWho Are Primary Care Providers: Comparisonof Achievementswith Projections 1970 1975 1980 1983 1985 1990 Achieved lo 37.9 38.4 38.5 38.4* 38.8* Projected (GMENAC) 40.6 42.3 *Approximates 35 per cent if recalculated using new data collected since 1981 that differentiate and separate additional subspecialists from the ranks of general internists. tion (demand). Two notable exceptions to the antici- pated glut were in the primary care fields of family practice and general internal medicine, which were expected to be in approximate balance with de- mand by 1990. ~ The forecasts for an adequate supply of primary care physicians were supported by the substantial growth in the 1970s of family practice training pro- grams and, to a lesser extent, internal medicine pro- grams.7, 8Also important was the slightly later devel- opment of increased interest in primary care expressed by those selecting internal medicine resi- dencies in the late 1970s and early 1980s, as demon- strated by the decrease in those choosing subspe- cialty training from 75 per cent in 1977 to 59 per cent in 1981 - 82. 8These two groups (the other being pedi- atrics) were to have accounted for most of the pro- jected increase in the percentage of primary care physicians, which was to have risen from 38.2 per cent of the total number of physicians in 1974 to 40.6 per cent in 1985 to 42.3 per cent in 1990. 9 However, the increases forecasted for primary care slowed appreciably by 1980 and have remained relatively flat since that time (making achievement of even GMENAC's modest goals unlikely (Table 1). Stein- wachs notes that ff a more strict definition of primary care were used the numbers would be closer to 21 per cent in 1970 and 32 per cent in 1980. n For about the last six years, 12- 13 per cent of medical school graduates have entered family practice, and 35- 37 per cent have entered internal medicine (including those headed for subspecialty training).12 For family practice the current number of first-year residents (approximately 1,900) falls well short of the GMENAC prediction for the years 1980- 1985 of 2,347. 7 Moreover, the new family physicians have largely just replaced older GPs who are retir- 251

Upload: gene-a-h-kallenberg

Post on 21-Aug-2016

213 views

Category:

Documents


1 download

TRANSCRIPT

PERSPECTIVES Waiting for the Doctor Glut, or Is the Cavalry Really Coming?

GENE A. H. KALLENBERG, MD, RICHARD K. RIEGELMAN, MD, PhD, LISA I. K. HOCKEY, MD

SINCE 1980, a great deal has been written about the "doctor glut" predicted by the final report of the Gradua te Medical Education National Advisory Committee (GMENAC). 1 As one medical journalist recently put it, " . . . the words 'doctor' and 'glut' have become as wedded in medical vocabulary as 'coronary' and 'bypass.'"2 Unfortunately, one of the report's essential conclusions, that the predicted glut is for specialty physicians and n o t for primary care physicians, has received far less attention. 3-5 In re- sponse to the predicted glut m a n y medical school administrators have a l ready reduced or are pres- ently considering reducing the sizes of future classes. 6

Many of us who toil a w a y in the field of primary care continue to feel that our country is still seriously deficient in its supply of properly trained primary care physicians. We are especially concerned by the rush to reduce the number of medical students before real solutions to the problem of the dispropor- tionate mix of specialists a n d generalists are in place. We feel that this approach, combined with the present administration's attempts to cut federal funding for all pos tgraduate medical training, could result in a serious decline in the future annua l pro- duction of primary care physicians both in absolute numbers and in relation to the numbers of specialists.

Reconsideration of some widely accepted facts about primary care physician supply and demand m a y help to ensure that our policies are appropri- ately focused.

SUPPLY

Primary Care Supply Forecasts-- Overly Optimistic?

In 1980 the GMENAC report predicted that by 1990 there would be an overall excess number of physicians and that the excess would be concen- trated predominantly in a variety of (sub)special- ties.~ The committee based its conclusions on extrap- olations from then-current da t a on medical student, resident and physician populations (supply) plus es- timates of existing and future levels of patient utiliza-

Received from the George Washington University Health Plan, 1229 25th Street, N.W., Washington, DC 20037.

Address correspondence and reprint requests to Dr. Kallenberg.

TABLE 1

Percentages of Active Physicians Who Are Primary Care Providers: Comparison of Achievements with Projections

1970 1975 1980 1983 1985 1990

Achieved lo 37.9 38.4 38.5 38.4* 38.8* Projected (GMENAC) 40.6 42.3

*Approximates 35 per cent if recalculated using new data collected since 1981 that differentiate and separate additional subspecialists from the ranks of general internists.

tion (demand). Two notable exceptions to the antici- pa ted glut were in the primary care fields of family practice and general internal medicine, which were expected to be in approximate ba lance with de- m a n d by 1990. ~

The forecasts for an adequa te supply of primary care physicians were supported by the substantial growth in the 1970s of family practice training pro- grams and, to a lesser extent, internal medicine pro- grams.7, 8 Also important was the slightly later devel- opment of increased interest in primary care expressed by those selecting internal medicine resi- dencies in the late 1970s and early 1980s, as demon- strated by the decrease in those choosing subspe- cialty training from 75 per cent in 1977 to 59 per cent in 1981 - 82. 8 These two groups (the other being pedi- atrics) were to have accounted for most of the pro- jected increase in the percentage of primary care physicians, which was to have risen from 38.2 per cent of the total number of physicians in 1974 to 40.6 per cent in 1985 to 42.3 per cent in 1990. 9 However, the increases forecasted for primary care slowed appreciably by 1980 and have remained relatively flat since that time (making achievement of even GMENAC's modest goals unlikely (Table 1). Stein- wachs notes that ff a more strict definition of primary care were used the numbers would be closer to 21 per cent in 1970 and 32 per cent in 1980. n

For about the last six years, 12- 13 per cent of medical school graduates have entered family practice, and 35- 37 per cent have entered internal medicine (including those h e a d e d for subspecialty training).12 For family practice the current number of first-year residents (approximately 1,900) falls well short of the GMENAC prediction for the years 1980- 1985 of 2,347. 7 Moreover, the new family physicians have largely just replaced older GPs who are retir-

251

~.S~. Kallenberg etal., WAITING FOR THE DOCTOR GLUT, OR IS THE CAVALRY REALLY COMING?

TABLE Z

Subspecialization Rates (Percentages) for Internal Medicine Residency Graduates, 1 9 7 7 - 7 8 Through 1 9 8 4 - 8 5

1 9 7 7 - 7 8 75 1 9 7 8 - 7 9 69 1 9 7 9 - 8 0 63 1980-81 58 1 9 8 1 - 8 2 59* 1 9 8 2 - 8 3 62* 1 9 8 3 - 8 4 60* 1 9 8 4 - 8 5 61"

*Recalculated based on 15 subspecialties (including clinical pharma- cology, critical care, geriatrics, and nutrition) rather than the traditional 11 used in prior calculations. General internal medicine fellows excluded, s

ing.9 Within internal medicine the most recently pub- lished da ta about residency graduates demon- strates a possible trend back toward subspecialization despite various redefinitions of what kinds of fellows should be included in the cal- culations (Table 2). These da ta also indicate an ab- sence of any self-regulation by teaching centers or fellowship directors in curtailing the number of sub- specialty fellowship training positions. Such trends will result in the majority of internists' having sub- specialized by 1990, with the ratio of subspecialists to generalists continuing to increase thereafter, s

One other relatively recent change in the de- mography of the current medical student population might significantly affect the future ranks of primary care physicians. Women now comprise 30- 40 per cent of m a n y medical student classes. ~s Further- more, several investigators have est imated that fe- male physicians work 20- 40 per cent fewer hours during their practice lives than do male physicians, primarily because of family responsibilities. |4 Addi- tionally, some have estimated that 50 - 70 per c e n t of these women will marry fellow physicians. ~s The in- c reased family pressures within a two-professional family and the options created by two professional incomes will often result in less than the expected two for two FTE practitioners e n g a g e d in delivering patient care. As one recent study suggests, these factors could reduce GMENAC's surplus manpower projections by as much as 28 per cent. 14 While this applies to all physicians, primary care would be af- fected substantially, because approximately 25 per cent of all women physicians enter family practice or general internal medicine residencies, s. 16

Current Realities and Future Policies

Given this leveling off of the number of primary care graduates , the supply situation could become even worse with a policy of limiting medical school class size at a time when graduates ' specialty pref- erences m a y be turning back in the wrong direction. As Steinwachs has pointed out, it is at the entrance to

g radua te medical education that one can have the biggest impact on the generalist-to-specialist ratio between now and 1990. Whereas a 10 per cent pref- erence shift from specialist to primary care would yield a 5.6 per cent "positive" change in the ratio of generalists to specialists by 1990, decreas ing medi- cal school classes by 10 per cent would cause a greater percentage reduction for primary care phy- sicians than for specialists over the s a m e period. This is in part due to the relatively shorter training time for a generalist, n

The reasons for the stabilization of both the in- terest in such positions evinced by medical school graduates and the number of residency positions avai lable also lie in current policies. First, while the reasons for specialty choice are very complex, today's graduates with their ever-increasing tuition payback burdens cannot help but be influenced by the still-significant sa lary differentials between pri- mary care and specialist careers. ~7" ~s

Second, having been stable in numbers over the past several years, ~9 it is highly unlikely that the number of primary care residencies will increase in the nea r future. In fact, m a n y ambulatory care train- ing programs are facing reductions in their subsidies by direct federal and state grants because of both federal and state budgetary crises. Because of the expectation of a doctor glut, the present administra- tion has repeatedly proposed eliminating funding for all postgraduate training programs/`° Under the existing reimbursement systems few primary care programs could achieve complete financial self-suf- ficiency from patient care incomes alone. ~. ~~' 22 Even the new proposals for separat ing resident salary support from Medicare payments to teaching hospi- tals m a y genera te a divisive financial competition for resident positions be tween specialists and gen- eralist training programs. 23" 24

Perhaps the reimbursement gap between inpa- tient and outpatient services will be closed by DRG- like changes in the physician's fee component of health care financing. Perhaps new mechanisms will evolve for both financing and changing the pro- portions of primary care vs. specialty residency po- sitions.23, 24 And perhaps new medical school admis- sion policies will be adopted favoring those oriented toward careers in primary care. But for now these potential solutions remain on the drawing board and will not be adopted quickly.

Thus, in the immediate future the number of physicians actually trained in primary care pro- grams is unlikely to increase substantially; the GMENAC forecast will not be achieved, and gener- alists will continue to be substantially outnumbered by their specialist colleagues. Therefore, a signifi- cant proportion of primary care will continue to be delivered by subspecialty-trained physicians forced

JOURNAL OF GENERAL INTERNAL MEDICINE, Volume 2 (3ul/Aug), 1987 253

by financial pressure to make their livings outside their fields of training a n d interest. Despite argu- ments supporting their pr imary ca re contributions, a glut of physicians inappropr ia te ly t rained and forced to pract ice against their original professional desires m a y well prove to have substantial undesir- able effects on patients, the doctors themselves, and the heal th ca re system. 2s'~9

DEMAND

Forecasting D e m a n d - Epidemiology and Patient Behavior

The need for heal th ca re services, and particu- larly pr imary ca re services, can be est imated very differently depend ing on the methods used. One major a p p r o a c h is cal led "demand-based" in that estimations a re ext rapola ted from current heal th ca re service utilization rates (which can va ry great ly depend ing upon which patient population, geogra- phy, or heal th ca re system is used). The other major a p p r o a c h is called "needs -based" in that it follows the assumption of providing full services to the entire populat ion for all conditions present ba sed on epi- demiologic measurements of d isease incidences. The GMENAC recommendat ions acknowledged that reality was best r ep resen ted by a b lend of these approaches , which they called an "adjusted needs- b a s e d model. ''9 It is still clear, however , that any such estimates will face constant revision as changes occur in the basic heal th of the population, technology, the access to and configuration of the heal th ca re system, and the extent to which the American people a re willing to f inance their heal th c a r e needs , s°' 3~

As a single example of an unexpec ted and sud- den c h a n g e in as basic a factor as the heal th (or incidence of disease) of the American people, one can point to the rise of AIDS and re la ted illnesses from virtual nonexis tence to near-epidemic propor- tions. The Public Health Service projects that by 1990 there will be 270,000 cases of AIDS, requiring $8 to $16 billion worth of care . s2 With the inc reased em- phasis on outpatient t reatment of less acutely ill pa- tients and the needs of those with AIDS-related com- plex or just HIV positivity, a growing burden is being p laced on pr imary ca re physicians. These recent events could conceivably chal lenge GMENAC m a n p o w e r predictions if the AIDS problem con- tinues to grow at the current pace .

Another c h a n g e that is a l r e ady occurring is the aging of the Amer ican population, which, by the y e a r 2000, will make "geriatricians" of us all. How- ever, the burden of delivering and coordinat ing the vast majority of ca re for the a g e d will fall to the fam- ily physicians and the genera l internists. The elderly h a v e a wide var ie ty of complex problems, and car-

ing for them requires a large investment of time even in the ambula tory ca re setting. Some h a v e further pointed out that the demographics of the "graying of the baby-boom could turn the surplus of physicians predicted by GMENAC in the y e a r 2000 into a short- a g e ten yea r s later. ''5

How people perce ive their needs for heal th ca re also affects the d e m a n d for pr imary ca re services. As those who pract ice pr imary ca re know full well, there is enormous variat ion in patients' thresholds for seeking care. Studies of symptom preva lences suggest that only a small minority of patients with symptoms seek care. 3s Systems of ca re that allow unlimited access, such as m a n y HMOs, recognize the enormous potential d e m a n d for services. 34 Even when ca re is provided without cha rge patients often come in only after it is t oo late to maximize their outcomes. A recent Robert Wood Johnson study sug- gests that some 28 million Americans still h a v e a difficult time getting access to our heal th ca re sys- tem. ss Because of this, the Institute of Medicine re- cently a d v o c a t e d a system of communi ty-based pri- mary ca re that great ly expands the role of pr imary ca re physicians, getting them out of their offices and into the community. This app roach b roadens the definition of heal th ca re needs to include those that will benefit most from medical ca re ra ther than just those that r each the doctor's office, se

Along the same lines, one must consider two addit ional sources of need for pr imary ca re services. While the new family pract ice g radua te s a n d the National Health Service Corps m a d e significant progress in the 70s at redress ing the geographica l imbalances in physician distribution, there remain about 1,950 pr imary ca re heal th m a n p o w e r short- a g e a r e a s that need over 4,300 pr imary ca re physi- cians to serve them. s7 Furthermore, 35 million Ameri- cans a re currently without heal th insurance, an increase of 25 per cent since 1977. s8 Inasmuch as insurance cove rage is a key factor that determines use of heal th ca re services, s9 should our society rec- tify these inequalities, predictions about a ba l anced supply of pr imary ca re physicians could be very inaccurate .

New Health Care Delivery Systems Produce New Demands

Beyond accura te ly predicting the future patient ca re needs for pr imary ca re physicians, the impact of newly evolving heal th ca re del ivery systems must be cons idered in calculating the actual amount of inc reased patient ca re time that will be provided by new additions to the ranks of such physicians. We choose to discuss this impact on the "demand" side of the ledger, because it does, in fact, represent de- mands of work and time spent by pr imary ca re phy-

Z54 Kallenberg eta/., WAITING FOR THE DOCTOR GLUT. OR IS THE CAVALRY REALLY COMING?

sicians. Some of these will opera te on our specialist bre thren as well, but others m a y be particularly fo- cused on those who del ivery pr imary care.

The movement toward prepa id and corpora te medical ca re del ivery systems in which physicians a re sa lar ied employees m a y further acce le ra te the trend toward a reduction in the a v e r a g e physician's number of pat ient-contact hours. In private pract ice there a re strong financial incentives to work extra hours. Often the physician's income must first cover the largely fixed costs of overhead . Any extra hours a r e assoc ia ted with minimal addit ional costs and often result in exponent ia l increases in income. As physicians take sa lar ied positions in institutional de- livery systems, the financial incentives for such extra hours d e c r e a s e in e x c h a n g e for the expectat ion of "regular hours" and the a c c e p t a n c e of limits on earn ing potential. 4

As the del ivery of medical ca re evolves into larger, more complex and often corporate ly orga- nized systems, other new and e x p a n d e d d e m a n d s on physicians ' time follow. Cost control, quality as- surance, and pee r review activities become more important as large groups of patients and physi- cians a re affected by a particular heal th organiza- tion's policy decisions. Such activities require physi- c ian participation as well as more complex, more careful, a n d more t /me-consuming communicat ion and documenta t ion of care. Tarlov has pointed out the increasing role for "manager-physic ians , ''4 while others h a v e discussed the "ga tekeeper" func- tion that pr imary ca re physicians will increasingly p lay in p repa id heal th ca re systems. To perform this function well, such physicians will require new skills, which they will have to leam, as well as the time n e e d e d to use them. 4°

While the overall system m a y become more effi- cient as def ined in terms of dollars spent, one might specula te that the number of pr imary ca r e physi- cians required to ach ieve the same amount of pa- tient-contact hours might actually increase. Because of the complexity of what would h a v e to be mea- sured to demonst ra te this, da t a on this issue a re diffi- cult to come by. Consequently, conclusions about the overall effects of such new systems of heal th ca re del ivery on physician productivity and projected physician requirements a re still far from clear. Esti- mates va ry from a 40 per cent increase to an 80 per cent d e c r e a s e in forecas ted surpluses. 4~-4s

Managing Technology and Maintaining Quality

Other forces a re at work to reduce the number of pat ient-contact hours. The explosion in medical technology and medical information has had a dra- matic impact on the importance of continuing medi- cal educat ion for the pr imary ca re physician, s~ Such activities legitimately take time a w a y from direct pa-

tient care. While all physicians h a v e CME needs, the b read th of pr imary ca re a n d the "first-contact" aspec t of its clinical setting make the acquisition and ma in tenance of current medical knowledge very difficult and its impact often and immediately felt.

Also, as ambula tory sites assume a g rea te r role in the educat ion of medical students, pr imary ca re physicians will be increasingly called upon to per- form teaching and receptor functions. All these ac- tivities take time a w a y from direct pat ient care.

OVERSUPPLY- -COMPETIT ION AND UNNECESSARY SERVICES? . . . OR THE

CHANCE TO SLOW DOWN?

Is Oversupply In Primary Care Different?

Rather than settle for a projected "near bal- ance" in pr imary ca re physician supply (which, as we h a v e demonstra ted, there a re m a n y reasons to doubt), we bel ieve that an excess of such physicians could be a ve ry desi rable event.

Some have con tended that physicians will re- spond to an oversupply by creat ing their own de- m a n d in an effort to mainta in or increase their in- comes. 44 They would accomplish this by ordering unneces sa ry services for which they a re re imbursed and "churning" or creat ing unneces sa ry follow-up visits.

This a rgument has much more validity when appl ied to of ten-procedure-or iented specialists. Be- cause of present re imbursement inequalities, when a specialist performs an unneces sa ry procedure the re imbursement is still substantial. This is much less the case for the pr imary ca re physician, whose equivalent "cognitive" time with a patient brings far less proportional income. Indeed, m a n y of the pri- ma ry ca re physician's more cost-effective and pre- vention-oriented activities a re not yet even reim- bursed. 45 The financial sett/ng also makes a difference. In a fee-for-service system the physician is paid for wha teve r he does; if he does more he is paid more. With the rapidly expand ing trend toward p repaymen t in HMOs, PPOs, and IPAs, there is a strong disincentive on the physician's propensi ty to c rea te his own inc reased demand . 46 Because such systems will involve m a n y more pr imary ca re physi- cians than specialists, their income-limiting effects will be disproport ionately felt.

Finally, our patients m a y have more to s ay about the future structure of heal th ca re than the financial forces that now a p p e a r to be shaping heal th ca re delivery. Patients would like their pri- mary ca re continuity physicians to be available, good listeners a n d effective communicators, and in- terested in preserving their wellness as well as ex- plaining and treating their illnesses. Patients also don't want physicians to keep them waiting! For the

JOURNAL OF GENERAL INTERNAL MEDICINE, Volume 2 (Jul/Aug), 1987 2BS

2,500- 3,500 patients "required" to support a pr imary ca re physician, ~' 47 meet ing all these expectat ions m a y be difficult. Meeting them in the 13 - 15 minutes currently a l located to pr imary ca re physicians dur- ing the a v e r a g e patient visit m a y be impossible. 48 If we a re serious about meet ing these expectat ions we will inevitably be forced to see fewer patients per hour. 3~

Unlike the specialists, the pr imary ca re physi- c ian is usually the first cal led and the last called. The pr imary ca re requirement of ongoing continuity of ca re requires emotional time off to replenish one's stores of tolerance, compassion, and caring. Each d a y we work in our offices trying to t each patients to allow some time for themselves, to take things a bit slower, to set more realistic limits on their work re- sponsibilities, and to spend more time with their spouses, children, and families. All this for a more healthful life. Yet we ourselves often don't follow this advice. 49

The a v e r a g e pr imary ca re physician works 60- 70 hours per week, including all professional and professionally re la ted activities, and the a v e r a g e income of internists in pr imary ca re pract ice is in the r ange of $80,000-90,000. 50 Thus, the a v e r a g e in- come per hour is in the r ange of $30, hardly an ex- cessive ra te whe n compared with the incomes of other professionals and businesspersons in our soci- ety. So we suggest that, ra ther than getting paid at exorbitant rates for their time, pr imary ca re physi- cians m a y be working excess ive hours in order to ach ieve their incomes.

A Better Solution--Fewer Hours

Perhaps a better solution would be for a sub- stantially increased n u m b e r of pr imary ca re physi- cians to each work f ewer hours. This app roach might d e c r e a s e salar ies until the inequities in reim- bursement be tween technical and cognitive skills and services a re corrected, s~ However, fewer work hours might h a v e positive benefits, both in reducing the number of impaired or at least frustrated and overworked physicians a n d in improving the overall quality of pr imary care.

Only time will tell whether we will develop the commitment nece s sa ry to increase the number of practicing pr imary ca re physicians and redress our country's lopsided imbalance be tween generalists a n d specialists. In our opinion, this will require new criteria for medical school admission, reapport ion- ment of training resources such as res idency posi- tions, and a combination of limits on further tuition increases and a redress ing of the imbalance in sala- ries be tween generalists and specialists. Until these changes a re accomplished, any reduction in total physician output, while certainly reducing the num- ber of specialists, m a y also result in the unwan ted

consequence of an inadequa te supply of properly t ra ined pr imary ca re physicians. This would surely make future pr imary ca re pract ice even more har- ried and d e m a n d i n g for both patient a n d physician.

Though an inc reased supply of pr imary ca re physicians might result in an artificial creat ion of demand, an equal ly likely possibility is that pr imary ca re physicians would utilize the increased time to fulfill the m a n y unmet needs and expectat ions of their patients, their own families, and themselves. In the meantime, we eage r ly await the pr imary care doctor glut. In our opinion, it can ' t come too soon!

REFERENCES

1. Summary Report of the Graduate Medical Education National Advi- sory Committee (Vol. 1 ). DHHS Publication No. HRA 81-651. Wash- ington, DC" U.S. Department of Health and Human Services, 1980

2. Stelmach W J, Sammons JH. Is there really a doctor glut? New Physi- cian 1984; September: 8-11

3. Louria DB. Sounding Board, Copingwiththeapproachingdoctor glut. N Engl J Med 1979;300:1047-9

4. Tarlov AR. Special report, Shattuck Lecture--the increasing supply of physicians, the changing structure of the health-services system, and the future practice of medicine. N Engl J Med 1983;308:1235-44

5. Haas WH, Crandall LA. A word of caution on the implications of the GMENAC report (letter). N Engl J Med 1983;308:1543

6. Schroeder SA, Gerbert B, Showstack JA, Chapman S. Reduction in medical-school class size (letter). N" Engl J Med 1984;311:1583

7. Ricketts TC, DeFriese GH, Wilson G. Trends in the growth of family practice residency training programs. Health Affairs 1986;5(4):89-96

8. Cox MW, Aday LA, Levey GS, Andersen RM. National Study of Inter- nal Medicine Manpower: X. Internal medicine residency and fellowship training: 1985 update. Ann Intern Med 1986; 104:241-5

9. Interim Report of the Graduate Medical Education National Advisory Committee (DHEW Publication No. HRA 79-633). Washington, DC: U.S. Department of Health, Education and Welfare, 1979

10. American Medical Association. Physician characteristics and distribu- tion in the United States, 1985. Chicago: AMA, 1986

11. Steinwachs DM, Levine DM, Elzinga D J, Salkever DS, Parker RD, Weisman CS. Changing patterns of graduate medical education: ana- lyzing recent trends and projecting their impact. N Engl J Med 1982;306:10-4

12. Graettinger JS. Datagram, results of the NRMP for 1986. J Med Educ 1986;61:617-9

13. Crowley AE, Etzel SI, Petersen ES. Undergraduate medical education. JAMA 1984;252:1525-32

14. Lanska M J, Lanska DJ. Effect of rising percentage of female physi- cians on projections of physician supply. J Med Educ 1984;59:849-55

15. Shapiro R. Women in medicine. New Physician 1984;March: 10-4 16. Crowley AE. Summary statistics on graduate medical education in the

United States. JAMA 1984;252:1545-53 17. Steinberg EP, Lawrence RS. Where have all the doctors gone: physi-

cian choices between specialty and primary care practice. Ann Intern Med 1980;93:619-23

18. Geertsma RH, Romano J. Relationship between expected indebted- ness and career choice of medical students. J Med Educ 1986;61:555-9

19. NRMP directories of hospitals/programs 1979-1985, National Resident Matching Program, Evanston, IL

20. U.S. Department of Health and Human Services. The fiscal year 1986 budget: health professions education

21. Lee PR. LeRoy LB. Paying for primary care--t ime for a change? Am J Med 1980;68:319-21

22. Ciriacy EW, Liang FZ, Godes JR, Dunn LD. The cost and funding of family practice graduate education in the United States. J Faro Pract 1985;20:285-95

23. Prout DM. Challenges facing general internal medicine in the 99th Congress. J Gen Intern Med 1986; 1:44-8

~.$6 Kallenberg et aL, WAITING FOR THE DOCTOR GLUT, OR IS THE CAVALRY REALLY COMING?

24. Whitcomb ME. Sounding Board, The federal government and gradu- ate medical education. N Engl J Med 1984;31 O: 1322-4

25. Schroeder SA. Western European responses to physician supply. JAMA 1984;252:373-84

26. McCue JD. Training internists: insights from private practice. Am J Med 1981 ;95:772-4

27. Kroenke K. Ambulatory care: practice imperfect. Am J Med 1986;80:339-42

28. Spiegel JS, Rubenstein LV, Scott B, Brook RH. Who is the primary physician? N Engl J Med 1983;308:1208-12

29. Peterson ML. The place of the general internist in primary care (edito- rial). Ann Intern Med 1979;91:305-6

30. Blendon R J, Altman DE. Special Report, Public attitudes about health-care costs. N Engl J Med 1984;311:613-6

31. Harris JE. How many doctors are enough? Health Affairs 1986;5(4):73-83

32. Public Health Service. PHS plan for prevention and control of aids. Public Health Rep 1986; 101:341-8

33. Banks MH, Beresford SA, Morrell DC, Waller J J, Watkins CJ. Factors influencing demand for primary medical care in women aged 18- 44 years: a preliminary report. Int J Epidemio11975;4:18g-g5

34. Garfield SR. The delivery of medical care. Sci American 1970;222:15-23

35. Updated report on access to health care for the American people: The Robert Wood Johnson Foundation Special Report, Number One. Princeton, N J: The Robert Wood Johnson Foundation, 1983

36. Community oriented primary care; a practical assessment. Vol. 1, committee report (National Academy Press, April 1984)vii:106

37. HMSA Statistics as of Dec. 31. 1986. Office of Data Analysis and Management, Bureau of Health Professions, Department of Health and Human Services

38. Mundinger MO. Sounding Board: Health service funding cuts and the

declining health of the poor. N Engl J Med 1985;313:44-7 39. Wilensky GR, Berk ML. Poor, sick and uninsured. Health Affairs

1983;2(2):91-5 40. Eisenberg JM. The internist as gatekeeper: preparing the general

internist for a new role. Ann Intern Med 1985; 102:537-43 41. Steinwachs DM, Weiner JP, Shapiro S, Batalden P, Colt-in K, Wasser-

man F. A comparison of the requirement for primary care physicians in HMOs with projections made by the GMENAC. N Engl J Med 1986;314:217-22

42. Jacobsen S J, Rimm AA. Primary care physicians: HMOs versus the GMENAC (letter). N Engl J Med 1986;315:324

43. Rudzinski K, Katzoff K. Impact of productivity changes on 1990 physician requirements and anticipated surplus. Presented at 1985 Public Health Conference on Records and Statistics, August 15, 1985

44. Rice TH. The impact of changing Medicare reimbursement rates on physician-induced demand. Med Care 1983;21:803-15

45. Schemer RM, Weifeld N, Ruby G, Estes EH. A manpower policy for primary health care. N Engl J Med 1978;298:1058-62

46. McPhee S J, Myers LP, Lo B, Charles G. Cost containment confronts physicians. Ann Intern Med 1984; 100:604-6

47. Kress JR, Singer J. HMO Handbook: a guide for development of prepaid group practice health maintenance organizations. Maryland: Aspen Systems Corporation, 1975:115

48. Medical Practice in the United States: A special report of the Robert Wood Johnson Foundation. New Jersey, 1981

49. McCue JD. The effect of stress on physicians and their medical prac- tice. N Engl J Med 1982;306:458-63

50. Owens A. Are you still losing out to inflation? Med Economics 1984; 17(September): 181-90

51. Schwartz H. Doctors debate how to split the fees. Wall St J January 15, 1985