For several decades, a substantial number of U.S. medical schools have offered combined baccalaureate/MD (BA/MD) degree programs, granting conditional acceptance to medical school at the time of admission to the affiliated undergraduate program. As of 2011, 57 of the 141 U.S. medical schools that were accredited by the Liaison Committee on Medical Education offered one or more combined degree programs.1 Programs vary in size. The typical program is small, enrolling fewer than 10 students per medical school class, but several enroll more than 50 students per year. The stated missions of these programs also vary. Some target the recruitment of honors students or those from disadvantaged backgrounds, others focus on integrating the premedical and medical learning environment, and some others emphasize the preparation of graduates for primary care or for scientifically oriented careers.1 Data regarding comparative educational outcomes of these programs are limited. Two program-specific analyses were published over two decades ago,2,3 and the latest review of national data was published in 1997.4 The latter was limited to an analysis of United States Medical Licensing Examination (USMLE) scores in combined degree versus traditional MD programs.
Proponents of combined programs argue that the early assurance of access to a medical education lessens the extreme academic pressures often associated with being a “premed” during college. The need for conformity to the presumed expectations of medical school admission committees is reduced, allowing exploration of other areas of intellectual interest outside of the student’s “comfort zone.” Critics of combined degree programs most often question students’ ability to make a mature career decision at the point of college application. A truncated college experience (at present, about 20% of combined degree programs involve less than the eight years traditionally needed to complete college and medical school)1 also generates concerns among some educators about the maturity of these students as they enter medical school.5
An additional, perhaps predominant, theme throughout the history of combined degree programs concerns the practical, societal advantages of accelerating the path to the MD. This formed a part of the initial rationale for the programs in the 1960s, as the need for physicians to serve those newly insured by Medicare stimulated a rethinking of the structure of medical education.6 A similar reassessment is under way in the wake of the 2010 Patient Protection and Affordable Care Act, and also because of increasing concern over the individual and societal costs of a medical education.7,8
The role that accelerated BA/MD programs may play in the national debate regarding the length and costs of medical training should be informed by the educational outcomes of such programs. Accordingly, the aim of our study was to compare the educational outcomes of students in a large accelerated medical program with those of students in the traditional pathway over a 15-year period.
Setting and participants
Northwestern University is a private, not-for-profit institution based in Evanston, Illinois, and classified by the Carnegie Foundation for the Advancement of Teaching as a “research university/very high research activity.” The Northwestern University Feinberg School of Medicine (hereafter “Feinberg”) is located in downtown Chicago, admits approximately 165 students each year, and, as of this writing (2014), has a total student body (first year through fourth year) of 686. We included all 2,583 students who matriculated to Feinberg between 1999 and 2013 in our study sample.
The Honors Program in Medical Education (HPME) at Northwestern University was created in 1961 under the leadership of John A.D. Cooper, MD, PhD, the dean of sciences at Northwestern University Medical School and later president of the Association of American Medical Colleges. The purpose of the program was to provide an integrated and accelerated curriculum for gifted students interested in the study of medicine.9 This and a similar program at Boston University were the first of their kind and undoubtedly served as models for other programs. The HPME program initially required only two years of undergraduate study before matriculation to medical school. Since the late 1980s, three years of undergraduate study have been required, to allow greater flexibility and individuality in the undergraduate experience. Students may enroll in the Weinberg College of Arts and Sciences, the School of Communications, or the McCormick School of Engineering and Applied Science. The cohort size peaked at 60 students (approximately one-third of the medical school class) during the 1970s and 1980s but is currently approximately 20 students per matriculating class, a size designed to maximize the medical school’s ability to mentor and track progress of these students during the premedical phase.
The current mission and focus is to provide a broad and challenging undergraduate education free from the pressures imposed by the intense competition for acceptance to medical school. Students are expected to take advantage of the reduced pressure by exploring areas of interest in medical research, public health, global health, and health policy, building a foundation on which they can expand when they enter medical school. Students are permitted to major in any subject in the School of Arts and Sciences. In the School of Communications and in the McCormick School of Engineering they are restricted to a human communication science major or a biomedical engineering major, respectively. The HPME program does not require any specific course work beyond the student’s major and the usual premedical science courses required by Feinberg. HPME students are not required to sit for the Medical College Admission Test (MCAT). The GPA required for continuation in the program has been maintained at a level meant to ensure appropriate competence for medical training but to temper pressures for academic perfection and allow some exploration and risk taking in college course selection. During the 15-year period represented by the present data, the overall GPA threshold requirement was 3.0, with the additional requirement that the GPA in required science courses be 3.2 or above. The average GPA of matriculating students was consistently well above this threshold, however (the average overall GPA for the last decade was 3.63).
When students matriculate to the medical school phase of the program, they are fully integrated into the class and complete all four years of the required curriculum. Feinberg’s grading system is pass/fail in the two preclinical years of the curriculum and honors/high pass/pass/fail in the clerkships during the third and fourth years of medical school.
Variables and measurement
We collected demographic data on gender, race, and ethnicity from Feinberg’s student records database. We assessed three measures of academic performance: membership in Alpha Omega Alpha (AOA), student quintile rankings from the Medical Student Performance Evaluation (MSPE), and USMLE Step 1 and Step 2 Clinical Knowledge scores. AOA, the national medical honor society, allows the induction of up to one-sixth (16.7%) of each medical school graduating class. To construct the MSPE prepared for each student’s residency application, quintiles of student rankings, based on scores determined each August in the last year of medical school, were used. Each student received a weighted score based on grades in the seven required clerkships. Weights were determined by clerkship length (e.g., a grade from a 12-week clerkship contributed more to the overall score than a grade from a 4-week clerkship). Points were deducted for failures in either preclinical or clinical years. USMLE Step 1 and Step 2 Clinical Knowledge scores were obtained from the student records database. USMLE Step 2 Clinical Skills results were not explored for this analysis, as they date back only to 2005 and are reported only as pass or fail; on this examination, there was an overall failure rate of Feinberg students of 0.9% (13 students out of 1,512 over a nine-year period).
Match success was recorded starting in 2003 in Feinberg’s student records database and is defined as a student matching in a residency position. Students’ choices of specialties were obtained from school records of the National Residency Matching Program, American Urological Association, and San Francisco Match results.
Analyses were conducted using Stata commercial statistical software (StataCorp, 2011, College Station, Texas). We used chi-square tests to compare gender, race/ethnicity, AOA, and specialty choice between HPME and non-HPME students. We computed specific comparisons of proportions using two-sample z tests only when the overall chi-square test was statistically significant (e.g., comparing the proportion specializing in emergency medicine only when the overall test for specialty choice by HPME was significant). We compared age at admission to medical school and USMLE scores using two-sample t tests. Quintiles of academic performance were compared using a Wilcoxon rank–sum test. We used an alpha of 0.05 to determine statistical significance.
This study was reviewed by the Northwestern University IRB (study number STU00097799) and given exempt status.
Table 1 displays demographic characteristics of the 2,583 medical students who matriculated from 1999 to 2013. A total of 560 (21.7%) were HPME students. In both groups, just under half were women.
There were significant racial/ethnic and age differences between HPME and non-HPME students. More than three-quarters of HPME students were Asian/Pacific Islander (409; 76.1%) compared with 432 (23.6%) of non-HPME students. One hundred five (19%) of the HPME students were white versus 1,046 (57.1%) of the non-HPME students. Minorities underrepresented in medicine constituted 15 (2.8%) of the HPME students versus 285 (15.5%) of the non-HPME students. HPME students were on average 2.2 years younger than non-HPME students at admission.
For HPME and non-HPME students, there were no significant differences in completion of the MD degree (455/468 [97.2%] versus 1,280/1,336 [95.8%]) or dismissal (4/468 [0.9%] versus 7/1,336 [0.5%]). Table 2 compares measures of academic performance by HPME versus non-HPME students among students who matriculated from 1999 to 2011 and whose degrees were either completed or pending at the time of data collection. AOA selection and USMLE Step 2 were available for students who matriculated before 2011; quintile performance was available for students matriculating 2000–2010. There were no significant differences between HPME and non-HPME students in the proportions inducted into AOA, distributions in quintiles, or average USMLE Step 1 or Step 2 scores.
Match success and specialty choice
Approximately 98% of both HPME and non-HPME students matched successfully (261/268 [97.4%] and 770/785 [98.1%], respectively). There were significant differences in specialty choice (see Table 3). More than a third (161; 35.8%) of HPME students entered categorical internal medicine residency programs, compared with 261 (20.6%) of non-HPME students. HPME students were significantly less likely than non-HPME students to choose emergency medicine—25 (5.6%) versus 110 (8.7%), respectively—and obstetrics–gynecology—9 (2.0%) versus 67 (5.3%), respectively.
This study demonstrates that students enrolled in Northwestern’s combined BA/MD program over a 15-year period achieved an academic performance level in medical school equivalent to that of the non-BA/MD students. The vast majority of the BA/MD students completed only three years of undergraduate education. Yet, we found no significant differences in the rate of completion of the MD degree, USMLE performance, AOA selection, overall class rank, and Match success between the two groups of students. Anecdotal concerns about inadequate preparation for medical school among students enrolled in accelerated BA/MD programs are not supported by these findings.
These findings can contribute to current discussions about the appropriateness and content of typical medical school selection criteria and premedical education. The HPME students we studied did not sit for the MCAT, and the GPA threshold to remain in the program was considerably lower than that typically needed for admission to competitive programs. These results may also illuminate the role that accelerated BA/MD programs can play in reducing the duration and cost of a medical education.
In their Perspective “‘How to Fix the Premedical Curriculum’ Revisited,” Gunderman and Kanter10 call for “students to identify their own intellectual passions and pursue them in depth. We need to foster a willingness to take risks, try out new things, and be creative.” The need for physicians who are compassionate, team-oriented lifelong learners may not be best supported by a premedical experience that emphasizes competitiveness for high GPA and top MCAT scores.
We recently informally surveyed 57 undergraduate HPME students and asked them to identify course work that they may not have undertaken had they been in a conventional premedical curriculum. Students frequently cited the value of experience in foreign language, music, history, and other liberal arts areas. Several students responded that they would not have pursued majors in engineering because of the traditionally lower GPAs among engineering students. The following two examples of their comments are typical:
I do not think I would have pursued a degree in philosophy, which would be a shame.… I have seen a huge improvement in my writing and critical thinking as … the philosophy classes push those attributes to the limit, just as the science courses require a more formulaic and problem-solving form of thinking.
If not in HPME, I would not have chosen to study biomedical engineering. Having the freedom to not worry about my GPA and what it would look like to a medical school admissions board allowed me to pursue a tougher major. Being an engineer has … taught me how to problem solve in a very practical way. Moreover, it has given me the skills to understand medicine and the human body from a more technical perspective, and I hope it will be good background to have if I want to pursue a more technical field in medicine.
Our data also help inform recent discussions and initiatives aimed at accelerating the path to the MD degree. Potential advantages of a reduced overall time to degree include a reduction of overall debt, an increase in the number of students considering a career in primary care fields,11 and younger physicians entering practice.12 One approach would be to reduce the length of the medical school curriculum.7,12 Supporters argue that a more widespread use of competency-based education and assessment allows for more individualized and therefore, in some cases, more rapid progress through the medical curriculum. Shortening the duration of medical school would cut 25% of the overall tuition and expenses and allow students to begin repaying medical school debt one year sooner.13 In Canada, the medical schools of McMaster University and the University of Calgary have already adopted a three-year medical school curriculum. There are also a few schools in the United States offering three-year curricular options for selected students, including New York University School of Medicine, Texas Tech University Health Sciences Center School of Medicine, and Mercer University School of Medicine.
Although truncating the medical school years would generally yield the greatest reduction in overall educational expenses given that the extraordinary levels of indebtedness accumulated by the majority of U.S. students are largely incurred during medical school,14 the reduction of the premedical curriculum by a year would also reduce students’ expenses substantially. Moreover, the adoption of a three-year medical school curriculum as the norm may be unwise.15 As the competency domains our students must achieve to succeed in the ever more complicated world of health care expand, it is hard to imagine a curriculum that can effectively prepare students with one less year of medical education. Furthermore, educators continue to struggle with effective competency-based assessment. It is currently difficult to envision an assessment system that would provide clear confidence that a student is ready to forgo a year of medical school. Others have raised concerns about the readiness of current students to enter residency even after four years of medical school education.16
Our study demonstrates the viability of an approach that reduces the time spent in premedical education rather than in medical school. BA/MD programs should be more central to this debate, particularly if other accelerated programs can be shown to have similar educational outcomes. The majority of BA/MD programs currently require four years of undergraduate and four years of medical school; many have lengthened the undergraduate portions of their BA/MD programs, citing concerns about inadequate preparation and possible immaturity of students matriculating to medical school. Our data suggest that these concerns may not be warranted, and there may, in fact, be advantages to reducing the undergraduate portion.
The distinctive racial and ethnic composition of our program, with a sizable majority reporting an Asian/Pacific Islander heritage and a paucity of students from groups underrepresented in medicine, also warrants discussion. There are, however, no comprehensive data available on racial and ethnic diversity in other BA/MD programs for comparison. The demographics of our program may simply result from its visibility in particular communities and high schools. Should the racial and ethnic composition of other BA/MD programs be similar to ours, however, the consequences for broader adoption of this pathway to the MD degree would need to be addressed, especially the implication of any apparent barriers to the enrollment of underrepresented minorities.
This and other potentially unique features of the HPME constitute the major limitation of this study; also, it reports the educational environment at only one undergraduate institution and one medical school. In addition, the program we have described here maintains traditional, highly selective admission criteria at the college entry level; some combined programs have other recruitment goals. Therefore, the findings may not be generalizable to other combined BA/MD programs. In addition, admission into this program is predicated on very high academic standards in high school and superlative SAT scores. The group we used for comparison with the BA/MD students (i.e., the non-HPME students) attended a variety of colleges and universities. Although we know their undergraduate GPAs, we do not know whether the two groups were comparable when they were admitted into college. This could be a limitation as well.
We found significant differences in the specialty choices of HPME and non-HPME students, most notably a higher proportion of HPME students choosing internal medicine. Our study did not allow for a more detailed understanding of this finding, nor did it examine long-term outcomes such as career satisfaction, involvement in research, or leadership roles among HPME graduates compared with other Feinberg graduates. These are among the questions warranting further study.
Nevertheless, the data we have presented here represent the first comprehensive assessment of academic performance in an accelerated BA/MD program to be published in over two decades. The equivalence of academic outcomes of the BA/MD and non-BA/MD students should stimulate more discussion about the value of such programs in ameliorating the length and cost of a medical education. The academic success of these students in medical school absent the usual emphasis on undergraduate GPA and MCAT scores is also supportive of ongoing efforts to redefine medical student selection criteria.
Acknowledgments: The authors wish to acknowledge Diane B. Wayne, MD, for her thoughtful and constructive comments during the preparation of this report, and Nicole Fancher and Brian Agne for their assistance with data collection.