Altering specialty balances.

A policy issue that has reemerged in current reform discussions is the appropriate balance of generalist and specialist physicians to meet national needs. This is not a new issue. Much health policy of the late 1960s and early 1970s was concerned with whether there would be enough physicians of the right specialties in the right places to meet the new access demands created in part by passage of Medicare and Medicaid. A large number of state, federal, and private sector programs were created, initially to increase the aggregate output of our schools of medicine and osteopathy and later to alter specialty balance. Notable initiatives were development of family medicine as a formal specialty, the National Health Service Corps, expansion of community and migrant health centers in underserved areas, educational initiatives in primary care at the undergraduate and the graduate levels, growth of training programs for nurse practitioners and physician's assistants, and state and private initiatives in loan repayment and other forms of underserved, primary care practice support.

The GMENAC report of 1980, which projected an overall physician surplus for the next two decades, led many to conclude that the overall increase in physicians would take care of problems of geographic and specialty maldistribution. This perception, combined with a general social malaise, resulted in limited activity and research on this topic in the '80s.

Toward the end of the decade, the issue began to draw increased attention again because of the closure of rural hospitals and the resulting attention on physician supply; the AIDS epidemic and its impact on innercity public hospitals; the decline of medical student interest in generalist careers; comparisons of the U.S. system with those of other developed countries, such as Canada; escalation in health care costs, with some evidence that generalists are more efficient; and the demand by managed care organizations for more generalist providers. Considerable debate and discussion have ensued in many professional and policy organizations. AMA and AAMC have endorsed the addition of generalist physicians, and both the Commission on Graduate Medical Education (COGME) and the Physician Payment Review Commission (PPRC) have proposed moving toward a target of total residency positions' being 110 percent of U.S. medical school graduates, with 50 percent being in generalist positions. Similar considerations have been under discussion in the Clinton Task Force Workgroup on Workforce Policy. The 50 percent goal has generated considerable attention and enthusiasm because of its simplicity and general acceptance of this direction for change.

What are the policy issues and considerations that such proposals raise?

* Rationale for more generalists. It is generally believed that more generalists are needed for access reasons in inner-city and rural areas as well as for cost effectiveness. HMO demand primarily falls into the latter category. Careful estimates need to be made of the access need both nationally and at the state level. Whereas such need unquestionably exists, it is likely to involve a relatively small percentage of all physicians. Judgments will need to be made as to whether general or targeted strategies are best applied to the access rationale.

Some studies support the cost effectiveness argument. One recent analysis indicated a $5 billion savings with increased generalists. The Medical Outcomes Study has found that specialists use more resources than generalists do and general internists use more than family physicians, controlling for case mix. More research is needed in this critical area. Demand by managed care organizations depends on the extent of managed care growth and on the type of managed care. Most estimates of generalist utilization come from group and staff models, while penetration and recent growth is in the IPA type of organization, which utilizes generalists more in line with the fee-for-service experience.

* Definitional issues. Care needs to be taken to be sure we are using the same definitions. Which physicians are generalists? A common definition is family practice, general internal medicine, and general pediatrics. Including OB/GYN or internal medicine subspecialties makes a considerable difference in supply or need estimates. Osteopathy needs to be included, as do substitution contributions of nurse practitioners and physician's assistants. Almost all datasets label physicians as either generalists or specialists, ignoring physicians with "mixed" practices. An additional definitional issue is when the specialty determination is made--medical student specialty intent, PGY-1 residency selection, postresidency placement, or ultimate practice specialty. The geographical area or type of nstitution also needs to be specified, e.g., smaller counties have and likely need a lower ratio of specialists than does an entire state, and specialty hospitals will have fewer generalist residents than will community hospitals concentrating on family practice.

* Percentage or per capita. There is a tendency to express generalist targets in terms of a percentage of all physicians in a given area or organization. While this may have the value of simplicity, it has the flaw that the total numbers will continue to increase over the next 20 years, so that 50 percent in 1995 will be considerably less per capita than 50 percent in 2015. Canada has 46 percent generalists, or 103 per 100,000 population, while Great Britain has

42 percent generalists, but only 59/100,000 population. Targets for generalist or specialist supply should be expressed in per capita terms wherever possible.

* Approaches to change. Some argue that this is not an issue that needs public sector intervention. They maintain that we really don't know if the current balance is inappropriate, that market forces such as managed care demand will correct the supply, and that the problems in overcorrection in a technology-driven profession are serious. On the other side, advocates for change argue that access and cost imperatives are absolutely clear and that academic systems are too insulated from market forces to rely on them alone. They also point out that production of specialists is primarily driven by the service needs of teaching hospitals and not in consideration of regional or national needs. Most discussion of reform focuses on some system of allocation of residency positions by either national or state public commissions or by private organizations such as ACGME. Allocation has been linked in some proposals to a national pool of GME funding skimmed from all payers for distribution to programs approved for residency slots. Whereas there are substantial analytic, political, and even legal issues with such plans, it is hard to believe that substantial reallocation will occur through market forces. The bulk of analysis to date indicates a moderate to serious oversupply of specialists over the next two decades. When combined with overall reduction in training slots such as envisioned in the COGME 110 percent proposal (reduction from 24,000 PGY-1 positions to 19,000), transition issues abound. Support for staff physician or midlevel provider resident substitutes is a minimum requirement. Additional resources will be needed to increase the capacity for training generalists and midlevel substitutes. The phasing in of the changes will have to be made so as not to disrupt patient care activities in teaching hospitals. Whereas residency position control may be the most direct way to rapidly alter generalist and specialist production, other changes in undergraduate admission and curricula, as well as in practice environment and compensation, need to be addressed. There is also considerable current discussion about specialist retraining to generalism. A series of professional, educational, and cost issues need to be resolved before this can be a major solution.

While I have stated above my reservations about the impact of market forces on educational institutions, significant changes in compensation could be important and could lessen the need for regulatory educational solutions. The editor of JAMA has called for any health reform proposal to increase generalist incomes to 90 percent of the average for all physicians by 1996. If that were a realistic possibility, some of the efforts described above might be reconsidered. The jury is out on which is more likely to happen and will most improve patient care and community health status.

David A. Kindig, MD, PhD, is Director and Professor, Programs in Health Management, Department of Preventive Medicine, University of Wisconsin, Madison.

David A. Kindig, MD, PhD, is Director and Professor, Programs in Health Management, Department of Preventive Medicine, School of Medicine, University of Wisconsin, Madison.