American Pediatric Society Presidential Address 1998: What is the Future for Academic Pediatrics?

Ladies and gentlemen, members of the Academic Pediatric Societies, and guests. It is a great honor to be President of the American Pediatric Society. I cannot adequately convey to the Society my gratitude for allowing me to serve in this capacity.

Cost containment, in the form of managed health care in the United States over the past decade, has had a profound impact on clinical departments in the medical schools of this nation, which in turn has affected the support of biomedical research. As the clinical services of academic health centers focus increasingly on the development of comprehensive and integrated delivery systems, surpluses from clinical practice revenues have become squeezed progressively by cost competition and the ability to use surplus clinical income to support research has diminished.

A recent survey which I conducted of U.S. and Canadian medical schools revealed that a mean of 56.9% of total departmental support for all U.S. and Canadian schools was derived from clinical income (Table 1). Schools reported that a mean of only 10.7% of departmental support came from federal research grants and a mean of an additional 11.0% came from nonfederal research funds. Several departments reported that 100% of departmental support was derived from clinical income.

Table 1 Percentage of U.S. and Canadian pediatric department support derived from clinical practice income or research

In this regard, it is important to note that federal, and in particular National Institutes of Health (NIH), support for pediatric research is clustered in a relatively small number of departments (Table 2). Only 32 (or 25%) of U.S. pediatric departments have more than 10 NIH grant awards; 35 departments have none. As a result, a relatively small number of medical students, pediatric residents, or fellows are exposed to NIH-supported pediatric investigators. A somewhat larger number have been exposed to research in some form and to research faculty who may have been supported by nonfederal research funding or by funds derived previously from surplus clinical income.

Table 2 NIH awards to U.S. pediatric departments

Moy and associates (1) reported that there was an inverse relationship between growth in NIH awards in the past decade and managed care penetration among U.S. medical schools (Fig. 1). Medical schools in all markets had comparable rates of growth of NIH awards from 1986 to 1990. Thereafter, medical schools in markets with high managed care penetration had slower rates of growth in the dollar amounts and in the number of NIH grant awards compared with schools in markets with low or medium managed care penetration.

Figure 1

RO1 awards to clinical departments in different managed care markets: 1986-1995 (modified with permission from JAMA 278:217-221, 1997; copyright 1997, American Medical Association).

Recognition of these economic realities has some disquieting consequences. First, it means that clinical departments will be less able to support a faculty member who spends a significant portion of his or her time in research but who does not recover commensurate support for salary or research-related expenses from extradepartmental research sponsors. Individuals who claim research time without research support are becoming less affordable. In most cases, these individuals should turn their energies to teaching effectively and enhancing the efficiency and productivity of their department's clinical practice. Acceptance of this notion suggests that whereas previously the successful clinical academician was expected to demonstrate qualities as a "triple threat" with excellence in teaching, research, and clinical care, this is no longer a reasonable expectation on which to build a department. Although such "triple threats" still exist, they are unusual; they are, however, likely to become the leaders of our academic enterprise in the future. Nevertheless, for most faculty members productivity is not equal in the three missions and the need to maximize the productivity of every faculty member may result in a reduction of the breadth of his or her individual mission. Similarly, departments can ill afford clinicians who work infrequently and inefficiently; the productive, grant-supported researcher who is less effective as a teacher or clinician should devote the largest portion of his or her time to the laboratory.

Research in a clinical department has become a highly competitive business, but it is a business in which pediatric departments must be prepared to compete if we are to continue to advance the knowledge base so essential to continuing to improve child health and to provide an appropriate environment for the training of both pediatric practitioners and future pediatric academicians.

What must we do not only to survive but to enhance the pediatric research enterprise in the future? I believe we must address the development of appropriate departmental and school research infrastructure issues and also address the manner in which subspecialty training is conducted.

In planning for research in clinical departments, it is useful to distinguish between two types of research programs: contract research and investigator-initiated research. Contract research implies a deliverable product, most often based on the execution of a research protocol designed by the sponsoring agent. Investigator-initiated research, on the other hand, more commonly is supported by governmental or philanthropic sponsors and is based on a search for new knowledge with less promise of a deliverable product of near-term economic value, but rather a principal intent of contribution to the body of science and the scientific literature. Both types of research and researchers must be valued and rewarded if they are to be successful, but their demands on a department and the institution will be substantially different.

Academic departments and medical schools often have not appreciated or responded to the needs of the pharmaceutical research sponsors. These corporations seek efficiency in patient recruitment and sponsor interaction, rapid responsiveness, and sometimes a willingness to surrender what the academician deems to be his or her own imprimatur on the research protocol. Protocols must be executed as written. To be successful, the faculty also should be trained in developing and carrying out clinical trials research.

I believe there are at least three elements essential for the development of effective investigator-initiated research. First, faculty recruitment must be focused on selecting the best candidates from a national pool, even at the assistant professor level, and on the development of resources necessary to attract the candidate who may be chosen.

Second, a strong infrastructure must be present to support such research. The issue of critical mass is also important. It is rarely possible for investigators to remain competitive for NIH grants in isolation: witness the clustering of NIH support for pediatrics among a relatively small number of institutions (Table 2). They must have adequately equipped laboratories and access to necessary core institutional resources and especially have the opportunity to interact with talented and vigorous scientific colleagues, usually from both basic and clinical science departments.

Third, the department and institution must be committed to the continuing investment required for the support of competitive and adequately funded faculty engaged in investigator-initiated research. Time allocated to carry out research must be commensurate with the commitments made to research sponsors in the grant application process. Faculty with salaries in excess of the NIH salary cap must not be pressured to make up this difference with clinical activities to generate clinical income. Those salary increments that may be required must be assumed as a departmental or school rather than an investigator responsibility.

What approaches must we pursue as a pediatric research community to ensure the continuity and growth of research (both investigator-initiated and mission-oriented contract research) as we move forward in the era of clinical care cost constraints? Of paramount importance is the training of future faculty and, in the case of pediatricians, of both M.D.'s and M.D./Ph.D.'s largely derived as products of our subspecialty fellowship training programs. We must accept the reality that training today for a successful career in biomedical research is as demanding in time and commitment as training for clinical practice.

Let us turn now to the issue of training and look at our collective subspecialty training programs, both in terms of the number of trainees referable to the perceived demand for these subspecialists and also at the current design of these training programs.

In an attempt to obtain accurate information about the near-term future of academic pediatric departments in the United States and Canada and to obtain a cross-sectional glimpse of current subspecialty programs and the trainees currently enrolled, I surveyed all pediatric departments in the United States and Canada during this past year. Departments were asked to list the types of subspecialty fellowships offered, the number of fellowship positions offered at the first-year fellowship level (usually PL-IV) and the number of positions filled at the first-year level, the number of fellows at each level, and the number and percentage of current fellows who planned a career in academic medicine (Table 3). Additional information requested included the total current faculty by specialty for each department, the projected new faculty in each specialty that the department planned to hire during the next 3 years, and/or the number of subspecialty faculty positions that might be eliminated during the next 3-year time frame. These data permitted a reasonably accurate assessment of the likelihood of finding an appropriate academic position for any current fellow who wished to pursue a career in academic medicine after completion of his or her training program.

Table 3 Information requested on a survey of U.S. and Canadian pediatric departments

The number and percentage of the 144 U.S. and Canadian pediatric departments who offer subspecialty training in any given area is shown for subspecialty in Figure 2. Fewer than half of pediatric departments offer subspecialty training in any area other than in neonatology. Only a very small number of departments offer training in rheumatology (8%), toxicology or experimental therapeutics (3%), whereas a relatively large number of departments offer training in neonatology (66%), infectious diseases (41%), and hematology/oncology (41%).

Figure 2

Number of U.S. and Canadian pediatric departments offering fellowship positions.

The number of positions offered in each subspecialty compared with the number and percentage of filled positions at the first-year level is shown in (Table 4). None of the subspecialty disciplines filled all available first-year positions in U.S. and Canadian medical schools. The fill rate ranges from highs of 96% in allergy and immunology, to lows of 60% in academic ambulatory pediatrics and 53% in toxicology and experimental therapeutics.

Table 4 U.S. and Canadian fellowship positions: the number of positions offered and filled in each specialty at the first-year fellowship level

Sixty-two percent of fellows in training indicated that they wished to pursue a career in academic pediatrics, whereas 38% indicated a desire to pursue a career in the private sector. The range varied markedly from department to department, with some departments reporting that 100% of fellows were interested in an academic career, whereas in others 100% of fellows planned careers in community practice. Departments in which 100% of fellows indicated either an academic or community practice career choice most often were departments that offered few fellowship positions.

Of the 2,375 fellows in training, 1,473 indicated that they had academic career aspirations. The fellows with academic aspirations by specialty are compared with the total number of new faculty positions that the pediatric chairs reported might be available during the next 3-year time frame (Fig. 3). There are many more fellows with academic aspirations currently in training than there are reported available positions for allergy/immunology, cardiology, critical care, emergency medicine, endocrinology and metabolism, hematology/oncology, gastroenterology and nutrition, infectious diseases, neonatology, renal, and toxicology and experimental therapeutics (Fig. 3). In contrast, more faculty positions are projected compared with trainees in academic ambulatory pediatrics, adolescent medicine, genetics, neurology, pulmonology and rheumatology.

Figure 3

U.S. and Canadian fellows with academic career aspirations compared with new faculty positions in academic years 1998-2000. The comparison excludes international medical graduates with academic aspirations.

These data clearly suggest that we are training, in many areas, more subspecialists than can be absorbed by the academic community. What will the job market look like in the private sector?

According to data derived from information provided by The Future of Pediatric Education II Task Force, on managed care staffing patterns, we already have too many pediatric cardiologists, pediatric gastroenterologists, pediatric hematologist-oncologists, and pediatric pulmonologists than required by health maintenance organizations (HMOs) per 100,000 enrolled population (2) (Table 5).

Table 5 Pediatric subspecialist physician full-time equivalents per 100 000 population

Whether or not we are training too many subspecialists to be absorbed in the future in the aggregate, that is, for academic endeavors, private practice, public sector service, or for industry, seems likely based on current data, but the ultimate answer to this question is specialty-specific and depends on the preparation of the trainees to compete effectively in the appropriate setting and the degree to which HMO staffing patterns rather than fee-for-service staffing patterns become the U.S. norm.

The clinical care committee of the North American Society of Pediatric Gastroenterology and Nutrition has recommended already that future training in the field should not exceed the number of American Board of Pediatrics-certified subspecialists at present (3). To accomplish that goal as projected into the future, it would require that the departments train 25 to 50% fewer individuals in this field during the next decade (Fig. 4). Even with this reduction, there would be sufficient trainee graduates to fill not only private practice positions but academic positions as well.

Figure 4

The projected supply and demand for each of the next 10 years for each of the supply and three demand scenarios (modified with permission from J Pediatr Gastroenterol Nutr 26:106-115, 1998). FTE, full-time equivalent.

What are the implications of these data for the future design of subspecialty training programs? The majority of clinical subspecialists in pediatrics currently practice in academic medical centers. Those who work outside of academic medical centers, frequently in a managed care environment, must see general pediatric patients as well, except in fields such as critical care, neonatology, and emergency medicine.

The evolution of patterns of subspecialty care and practice patterns within academic centers within recent years suggests that subspecialty training programs may require redesign. Many senior pediatric academicians are opposed to the evolution of different tracts within subspecialty training programs, and most defend passionately the need for a program of 3 years' duration. It should be noted, in this regard, that for an individual who is board-certified in internal medicine, certification to care for adolescent patients can be achieved after 2 years. This puts the pediatric community who may wish to train adolescent medicine specialists or to attract such individuals into pediatrics at a competitive disadvantage. The same situation pertains to emergency medicine physicians. Many academicians recognize that few fellowship training program graduates are prepared to compete as independent investigators and that additional training beyond 3 years is required to better prepare the fellowship graduate for academic independence. This additional training frequently is difficult to finance.

To focus what I hope will be an increasing dialogue on this subject, I submit the following. Pediatrician-subspecialists who practice in the general community, most commonly in neonatology and allergy and immunology, but in smaller numbers in all other subspecialties as well, generally are engaged in full-time clinical practice. Subspecialists in an academic setting tend either to be clinician-teachers or basic or clinical research physician-scientists. Each of these different types of subspecialists may require training more appropriately tailored to meet their needs. Such training could differ in scope and duration (not necessarily shorter than 3 years) and probably should be carried out by fewer programs and possibly by programs uniquely directed toward various paradigms.

The community-based subspecialists require training in their clinical subspecialty discipline, with more training time spent in an outpatient setting than currently is the norm.

The clinical subspecialists who work in an academic center will require the same subspecialty clinical training but perhaps also an additional year in a particular super-subspecialty, for example, if a cardiologist, in interventional cardiac catheterization or echocardiography and the like, or if in gastroenterology and nutrition, in hepatology for transplantation or in invasive endoscopy.

Relatively few centers have the necessary critical mass or the patient base to train these super-subspecialists, and I believe that we should look critically at the number and location of training facilities and at the number of special-interest subspecialists required to handle patient care needs, whether such needs are handled in an academic setting or elsewhere in the community.

We must continue to train physician-scientists; in fact, the future of pediatric care depends on the creation of new knowledge. These individuals must have the same basic clinical training as the other two groups, but they should anticipate additional training in research. The duration of additional research training will depend, in part, on prior research training and experience of the fellow. This type of training may occur in only a handful of centers (perhaps 20 or 30) which may differ in location depending on the individual subspecialty under consideration. In this regard, it is of interest that relatively few pediatric departments have NIH training grants (Table 6). Forty-eight U.S. pediatric departments or 37.5% report at least one NIH training grant. The total number of NIH training grants held by all pediatric departments is 157. Only 27 departments have an NIH training grant in more than one discipline, 13 in more than three, and only seven departments have five or more NIH training grants.

Table 6 U.S. pediatric departments with at least one NIH training grant

Today, in many subspecialty programs, those fellows who do not wish to participate in meaningful basic research may conduct clinical trials, but they frequently are asked to write case reports or establish databases to fulfill the requirements for research in fellowship training. In my own experience, as well as from reports of other training program directors in many disciplines, these trainees invariably end up in virtually full-time clinical practice within or outside the academic center.

To facilitate the training and to ensure the success of the physician-scientist group requires identification of appropriate mentors and the identification of committed support by the institution or from NIH for the extended period required to achieve independence, namely 4 to 7 years, with 75 to 80% protected time. Physician-scientist awards, clinical investigator awards, career development awards sponsored by NIH, or various other not-for-profit foundations are important sources of support for such trainees. We need to lobby our representatives in Congress to push for legislation to provide expanded sources of ongoing support for this type of postgraduate training for this group of trainees.

Is there a future for pediatric academic medicine and pediatric research in this reconfigured integrated health care delivery system? My answer is emphatically yes! But it requires a clear understanding of fundamental departmental, institutional, local, and even national goals and needs. At the departmental and individual institutional level, it requires an accurate assessment of research capabilities and potential and a commitment to accountability and assessment of faculty and training productivity. At the national level, it requires a valid assessment of near- and long-term subspecialty manpower needs. A true assessment of specialty manpower needs is highly complex, for the needs and opportunities are markedly different from one specialty to another. I believe a reassessment and reconfiguration of the fundamental structure and scope of our subspecialty fellowship training programs also is necessary, and I call upon the Academic Research Societies and AMSPDC, working with the subspecialty societies and other groups, such as the American Board of Pediatrics, to collectively address these manpower and training issues by the creation of a special task force dedicated to this endeavor.

In addition, I believe we must collectively work with our academic colleagues in other disciplines to lobby Congress for support for sustenance of the research infrastructure of academic health science centers. For example, the arbitary cap on administrative costs, particularly when coupled with increased regulatory mandates, has increasingly shifted the cost of research projects from NIH to institutional grantees. If implemented, the contemplated use of artificial regional benchmarks for reimbursement of facilities costs are sure to add substantially to the problem by preventing research-intensive institutions from renewing outdated facilities.

The Biomedical Research Support Grant Program of the National Center for Research Resources is still authorized by Congress but has not received an appropriation since 1992. This is a terrible loss of support for new research scientists. Another research resources program that has been seriously eroded is the Shared Instrumentation Grant Program, which gives researchers the opportunity to use critical new technologies and equipment that would be too expensive for individual research investigators to obtain via a single proposal. Funding for this program has declined from $32 million in FY 1991 to the current level of $21 million per year. If we truly want to stretch our research dollars, we need to expand these programs, not reduce them.

Perhaps the most critical threat to the research enterprise of academic pediatrics and academic research in general is in the area of clinical research. If we are to develop the next generation of clinical investigators, we must maintain the vitality of clinical research by getting research discoveries from the laboratory to the patient. The General Clinical Research Center Program, supported by the National Center for Research Resources, is at the heart of the infrastructure of the highest-quality clinical investigations within academic centers, yet funding for approved centers always seems to be below study section recommendations and new centers, despite favorable reviews, have difficulty achieving funding at all.

The future of pediatric research in our academic centers is now at stake to a degree that few of us have witnessed previously. Most of us teach medical students, house staff, faculty, and other health care professionals, and we interact with the lay public as well. We must seize these opportunities to convince all of these various groups to become lifelong advocates in the service of children and supporters of pediatric research. I would hope particularly that as champions of academic pediatrics we would continue vigorously to pursue our historic role of nurturing research and that we would review and reconfigure the size and scope of our training programs and departments proactively so that we may create the climate in which pediatric research can continue. I call upon the membership of the American Pediatric Society, as did my immediate predecessors, Dr. Fred Battaglia and Richard Johnston, to engage thoroughly in all of these initiatives so that future generations of children can be assured of a healthier life than those who have gone before.

Thank you for honoring me by permitting me to serve as your President.