Career Resources
Strategies to attract medical students to the specialty of child neurology
Rachel M. Werner MD,and Daniel Polsky PhD
Carl E. Stafstrom MD, PhD
Pediatric Neurology
Volume 30, Issue 1 , January 2004, Pages 35-38
Thu, 28 Jul 2005, 00:12

Introduction
A shortage of child neurologists in the United States was identified by the American Academy of Neurology (AAN) Workforce Task Force [1]. In 1998, there were approximately 1080 active child neurologists. Staffing ratio data indicated that this was 20% below the demand for child neurology services. This shortage is projected to remain unchanged through 2020.

One critical aspect of this shortage is the ability to supply new child neurologists through recruitment of students from U.S. medical schools. With the recent concern about the decrease in the number [2, 3, 4 and 5] and quality of U.S. medical students choosing neurology and child neurology as a career [1], some have suggested improving the neurology education of medical students [6]. Others have suggested specific changes to the undergraduate curriculum in neurology to ensure a minimum level of competency in clinical neurology for graduating medical students [7]. There have also been recommendations to encourage MD, PhD students to enter the field of neurology through financial supports and other programs [1]. Evidence is required to better understand which proposals would successfully attract medical students into child neurology.

We undertook this study to learn more about the qualities of highly regarded medical schools that may play a role in attracting students to the field of child neurology. We hypothesized that the extent of the neuroscience curriculum, the reputation of the academic medical center, and the size of the fellowship training program in child neurology may play a role in the number of students entering the field of child neurology from a given medical school.

Methods
Survey sample
The two target groups of medical schools for our survey were highly regarded medical schools that were either "successful" or "unsuccessful" at graduating medical students who chose child neurology as their primary specialty. We defined highly regarded medical schools as the top 50 medical schools from the U.S. News and World Report's "Top Graduate Schools" ranking in 2001 [8].

We determined the successful and unsuccessful medical schools by using the 2001 American Medical Association Physician Masterfile. The Masterfile is an administrative database of all physicians who began a residency program in the United States. For each of the top 50 medical schools, we counted the number of active physicians practicing in the United States in 2001 whose primary specialty on the Masterfile was listed as "Child Neurology" and who graduated from medical school between 1980 and 1998. We then ranked the top 50 medical schools by the number of child neurologists they graduated. The medical schools in the first quintile (schools that produced the most child neurologists) were defined as the "successful" group and those in the last quintile (schools that produced the least child neurologists) were defined as the "unsuccessful" group. We contacted all medical schools in these two groups.

Data collection
Members of the Division of Child Neurology involved in teaching residents or medical students were identified at each of the schools through departmental information published on their website. Those identified were contacted by phone to confirm their involvement in the medical school curriculum and residency training programs. Child neurologists who could not answer specific questions about the neuroscience curriculum or the Division of Child Neurology were asked to provide referrals of faculty that could.

A telephone survey was administered to participants during the summer of 2002. Surveys were developed based on a focus group with child neurology residents in 2001 who were asked to identify factors that were important to them in choosing to enter the field of child neurology. The topics the survey covered were the components of the preclinical and clinical neuroscience curriculum at each institution, and the structure of the Division of Child Neurology, including faculty, fellows, residents, students, research, and clinical components.

Secondary sources were used for additional characteristics of the medical schools. The top hospitals in the country for neurology and pediatrics were identified by U.S. News and World Report's "Best Hospitals" rankings [9 and 10], and national research rank of the medical school overall and the ranking for top neuroscience graduate schools in the U.S. were identified by the U.S. News and World Report's "Top Graduate Schools" rankings [8].

Analysis
We calculated the percentage of "yes" responses for the dichotomous variables and the mean values for the continuous variables for the high-suppliers and low-suppliers of child neurologists. P values were calculated for comparisons between the high and low-suppliers with ÷2 test for dichotomous variables and t test for continuous variables.

Results
Total number of child neurologists by U.S. medical school
The number of child neurologists produced at the top 50 medical schools in the United States varied widely (Table 1). The school that was the highest supplier of child neurologists, Columbia University College of Physicians and Surgeons, produced twice as many as any other school. Four schools produced no child neurologists over the 18 years we studied.

Table 1. Ranking of the top 50 medical schools by the number of child neurologists they supplied between 1980 and 1998

We contacted the 9 schools that graduated 6 or more child neurologists and the 13 schools that graduated zero or one child neurologist. Among the schools supplying a high number of child neurologists, we completed interviews with 8 of the 9 (89%). Of the low-supplying schools, we completed interviews with 8 of the 13 schools (62%).

Characteristics of medical schools that produce the most and fewest child neurologists
As shown in Table 2, it appears that it is more likely for the neuroscience curriculum to span both the first and second years of medical school and to expose students to child neurology earlier in medical school among the schools that were more successful in producing child neurologists. The small number of schools sampled was not statistically powered to detect these differences. The schools that produced the most child neurologists had higher academic reputations in pediatrics, neurology, and the medical school overall. The rank of the graduate programs in the neurosciences trended toward a higher ranking among those schools producing more child neurologists. However, this was not statistically significant. The size of the Divisions of Child Neurology and the child neurology training program were larger at the schools that produced the most child neurologists as well. Overall, our statistically significant findings suggest the medical schools that produced the most child neurologists had stronger neuroscience curricula, stronger academic reputations, and larger Divisions of Child Neurology.

Table 2. Characteristics of medical schools that have supplied the most and the fewest child neurologists

Discussion
In our brief survey of medical schools, we determined that medical schools that have produced more child neurologists have a stronger neuroscience curriculum and a stronger academic reputation than the schools that do not produce as many child neurologists. These schools also have larger Divisions of Child Neurology and larger training programs in child neurology. By looking specifically at the influence of medical school on the choice of a single subspecialty, child neurology, this conclusion was similar to a more general body of research examining the role of student attitudes toward specialties [11], characteristics of the student body [12], and differences in curriculum [13] on specialty choice [14 and 15]. Other research which focused on whether economic factors such as expected income might affect choice of specialty [16, 17 and 18] were largely inconclusive.

An academic reputation or neuroscience curriculum alone clearly is not the only difference between those medical schools that are most successful at producing child neurologists and those that are least successful. In a ranking of medical schools based on overall reputation [8], the schools that produce the most child neurologists are spread throughout the list, with at least one school falling in each quintile. Similarly, the majority of schools with overall rankings in the top ten have not produced more than a few child neurologists over the 18 years of our survey.

One explanation is that the schools that have been most successful in producing child neurologists have both strong academic reputations and have strong curricula in neuroscience. Although a strong curriculum in neuroscience is certainly necessary to spark medical students' interest in child neurology, alone it may not be enough. A strong academic atmosphere, with the benefits that accompany it such as high-quality residents and faculty, may also be necessary to foster medical students' interest in child neurology into a career choice. Either condition alone may not be sufficient to create the environment required to support students interested in child neurology.

It is also possible that students are being turned away from the field of child neurology at some schools. For example, third-year medical students who have had a positive experience in their preclinical neuroscience course and are considering a career in child neurology might by discouraged from entering the field if they had a poor experience in their neurology rotation. This could happen at schools where the Department of Neurology is not strong or the residency training program does not attract high-quality candidates. These qualities could be perceived as unattractive by students with a potential interest in the field.

There are limitations to our study. The small number of schools included in the study limits the statistical significance or our results, and hence our ability to draw more definitive conclusions. However, we have been able to generate hypotheses about what medical school qualities might contribute to encouraging medical students to enter the field of child neurology. More research is required to further explore our hypotheses and other factors that might impact a medical student's decision to enter a chosen field of medicine. It is also possible that a number of variables are not independent of one another. For example, the number of fellows and faculty in child neurology may be a function of the size of a medical school. Similarly, the national rank of individual departments is likely to be related to the rank of the medical school overall. Unfortunately, the size of this study does not allow us to conduct multivariate analysis. Finally, during the time period that our study spans, there has been a trend in medical school education toward an integrated approach in learning and teaching neuroscience in parallel with other portions of the medical school curriculum, such as anatomy, pharmacology, and pathology. There has also been a shift toward an increasing number of schools requiring a neurology clerkship for their students. Because our ranking of schools by the number of child neurologists they have produced is a function of 18 years and the specific characteristics of those schools reflect one snapshot in time, any relation between the two is subject to error.

With rare exception, everyone we spoke with agreed that there was a shortage of child neurologists in the United States and something needed to be done to alleviate the shortage. The ability to attract more high-quality students to the field of child neurology is critical. Our findings suggest that our attention should be focused on prestigious academic centers that have more resources to create an atmosphere that is appealing to prospective applicants. These schools must implement a curriculum in neuroscience and child neurology that specifically exposes them early and maintains their interest in the field of child neurology.

References
1. W.G. Bradley, Neurology in the next two decades. Neurology 54 (2000), pp. 787–789.     

2. Anonymous. Graduate medical education. JAMA 2000;284:1159-72

3. Anonymous. Graduate medical education. JAMA 1998;280:836-45

4. Anonymous. Graduate medical education. JAMA 2001;286:1095-107

5. Anonymous. Graduate medical education. JAMA 2002;288:1151-64

6. P.D. Charles, How much neurology should a medical student learn? A position statement of the AAN Undergraduate Education Subcommittee. Acad Med 74 (1999), pp. 23–26.    

7. D.J. Gelb, C.H. Gunderson, K.A. Henry, H.S. Kirshner and R.F. Jozefowicz, The neurology clerkship core curriculum. Neurology 58 (2002), pp. 849–852.     

8. America's best graduate schools 2002. U.S. News and World Report 4-9-2001;58

9. America's Best Hospitals 2002: Pediatrics [serial online] 2002 [cited 7-12-2002]. Available from: URL: www.usnews.com/usnews/nycu/hosptl/rankings/specreppedi.htm

10. America's Best Hospitals 2002. Neurology and Neurosurgery [serial online] 2002 [cited 7-12-2002]. Available from: URL: www.usnews.com/usnews/nycu/hosptl/rankings/specihqneur.htm

11. W.M. Zinn, A.M. Sullivan, N. Zotov et al., The effect of medical education on primary care orientation: Results of two national surveys of students' and residents' perspectives. Acad Med 76 (2001), pp. 355–365.    

12. J.H. Senf, D. Campos-Outcalt, A.J. Watkins, S. Bastacky and C. Killian, A systematic analysis of how medical school characteristics relate to graduates' choices of primary care specialties. Acad Med 72 (1997), pp. 524–533.    

13. D. Campos-Outcalt and J. Senf, A longitudinal, national study of the effect of implementing a required third-year family practice clerkship or a department of family medicine on the selection of family medicine by medical students. Acad Med 74 (1999), pp. 1016–1020.    

14. M.E. Whitcomb, T.J. Cullen, L.G. Hart, D.M. Lishner and R.A. Rosenblatt, Comparing the characteristics of schools that produce high percentages and low percentages of primary care physicians. Acad Med 67 (1992), pp. 587–591.    

15. C.J.M. Martini, J.J. Veloski, B. Barzanksy, G. Xu and S.K. Fields, Medical-school and student characteristics that influence choosing a generalist career. JAMA 272 (1994), pp. 661–668.    

16. J. Thorton and F. Esposto, How important are economic factors in choice of medical specialty?. Health Econ 12 (2003), pp. 67–73.

17. J. Thorton, Physician choice of medical specialty: Do economic incentives matter?. Appl Econ 32 (2000), pp. 1419–1428.

18. N. McKay, The economic determinants of specialty choice by medical residents. J Health Econ 9 (1990), pp. 335–357.

This topic is obviously of great interest and importance to everyone in our specialty. The authors document only 162 U.S. medical graduates entering the field of child neurology in the 19-year period from 1980 through 1998, which would give an average of fewer than 9 students per year. The authors concluded that medical schools with extensive neuroscience curricula, strong academic reputations, and large child neurology training programs were most successful in attracting students to child neurology. While those factors are certainly important, I feel that the authors have overlooked the most important reason why students choose this or any other specialty—the presence of an inspirational role model. Personal experience with a mentor who models intellectual excitement, compassion, and sheer joy of interacting with children who have neurologic problems and their families must surely have influenced many, if not most of us to choose child neurology as our specialty. Although this factor cannot be easily quantified by empiric research, its influence is undeniable.

At the University of Wisconsin Medical School (Madison), we do not currently have a child neurology training program, yet we have been extraordinarily successful in recruiting medical students to enter pediatric neurology in recent years. Since 2001, eight graduating medical students have chosen to pursue a career in child neurology, including five students from the class of 2004 alone. While these data are too recent to have been included in Werner and Polsky's survey, the trend counters their conclusions. First, our school's “academic reputation” (an inaccurate concept at best), at least according to the authors, would be in the lower quintiles of the top 50 medical schools. Second, our students are exposed to a formal neuroscience curriculum in only the first of the two preclinical years. Third, we are not a large division (4 faculty). Our success, therefore, must have another explanation. I propose that it is the enthusiasm of child neurology faculty and the interest they take in actively mentoring students that form the basis for students' decision to choose child neurology as a specialty.

Most first- and second-year medical students know little about child neurology, or perhaps even that the specialty exists! Therefore early exposure is critical; the sooner students have contact with child neurologists, the greater the chance that their interest will deepen. This goal can be achieved in several ways, in addition to those suggested by Werner and Polsky: (1) Offer students the opportunity to “shadow” a faculty member for a few clinic sessions or ward rounds during their first or second year of medical school. (2) Offer summer stints in the laboratories of faculty members pursuing research in topics related to developmental neuroscience. (3) Volunteer to teach in Introduction to Clinical Medicine courses in the first two years, in which students first learn history taking and clinical examination skills. (4) Volunteer to serve as an advisor to first- and second-year medical students. (5) Become active in (or form) a SIGN club (Student Interest Group in Neurology). Many U.S. medical schools have begun SIGN groups, sponsored by the American Academy of Neurology. We meet periodically at the homes of neurology faculty (and other neuroscience-related fields such as neurosurgery, rehabilitation medicine, etc.) to discuss career options, course offerings, and clerkship scheduling. SIGN meetings allow the opportunity for “peer mentoring,” with students invited from all four medical school classes. These gatherings allow faculty and older students to share information about residency training and application details (this is becoming increasingly critical for child neurology as the possibility of a match program looms). All of these opportunities are invaluable in making the personal connection with aspiring students.

We agree with Dr. Stafstrom that mentors are another important factor in attracting medical students to the specialty of child neurology. Our article on strategies to attract medical students to the field of child neurology did not report on the importance of mentorship [1], however, it was a factor we thought would be important and attempted to quantify in our survey of medical schools. In fact, among the medical schools producing the most child neurologists, 75% of the divisions of child neurology had faculty who had won teaching awards, compared with 25% of the divisions of child neurology in medical schools producing the fewest child neurologists. Schools that produced the most child neurologists were also more likely to see mentorship as an important role of their division (50% vs 13%). However, when asked about opportunities for mentorship in their division, the likelihood of being involved in a neurology interest group or offering students the opportunity to do summer research with faculty in the division were similar across the two groups of medical schools (33% vs 38% and 100% vs 88%, respectively). As Dr. Stafstrom correctly points out, empirically quantifying mentorship is a difficult task, and thus we question the validity of our findings on the presence of mentorship. However, the importance of mentorship was recently underscored in a survey we did of child neurology residents. Residents reported that mentors were the second most influential exposure to child neurology, after their clinical experiences in their third and fourth years of medical school [2].

While mentors are important to attract new candidates to the field of child neurology, in most cases mentors alone are insufficient to correct the shortage of medical students choosing to enter the specialty of child neurology. Medical schools with the resources to create an atmosphere that is appealing to prospective applicants will be most successful at cultivating the enthusiasm that mentors initiate.

References
[1] R.M. Werner and D. Polsky, Strategies to attract medical students to the specialty of child neurology, Pediatr Neurol 30 (2004), pp. 35–38.

[2] D. Polsky and R.M. Werner, The future of child neurology: A profile of child neurology residents, J Child Neurol 19 (2004), pp. 6–13.