Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Mar 1.
Published in final edited form as: Acad Med. 2016 Mar;91(3):388–394. doi: 10.1097/ACM.0000000000000922

The Association Between Premedical Curricular and Admission Requirements and Medical School Performance and Residency Placement: A Study of Two Admission Routes at The Warren Alpert Medical School of Brown University

Paul George 1, Yoon Soo Park 2, Julianne Ip 3, Philip A Gruppuso 4, Eli Y Adashi 5
PMCID: PMC4767648  NIHMSID: NIHMS713503  PMID: 26422591

Abstract

Purpose

The curricular elements of undergraduate premedical education are the subject of an ongoing debate. The Warren Alpert Medical School of Brown University (AMS) matriculates students via the traditional premedical route (TPM) and an eight-year baccalaureate/MD program—the Program in Liberal Medical Education (PLME)—which provides students with a broad and liberal education. Using the juxtaposition of these two admission routes, the authors aimed to determine whether there is an association between highly distinct premedical curricular and admission requirements and medical school performance and residency placement.

Method

The cohorts studied included all of the PLME (n = 295) and TPM (n = 215) students who graduated from the AMS between 2010 and 2015. Outcome variables consisted of multiple measures of medical school performance, including standardized multiple-choice examination scores and honors grades, and residency placement. The authors employed unadjusted tests of averages and proportions (independent t tests and chi-squared tests) to compare variables.

Results

The TPM students attained marginally, but statistically significantly, higher average scores on standardized multiple-choice examinations than their PLME counterparts. The number of undergraduate premedical science courses completed by PLME students accounted for less than 4% of the variance in key metrics of medical school performance. The residency placement record of the PLME and TPM cohorts proved comparable.

Conclusions

These findings suggest that the association between medical school performance and residency placement and undergraduate premedical curricular and admission requirements is weak. Further study is needed to determine the optimal premedical preparation of students.

The curricular elements of undergraduate premedical education are the subject of an ongoing debate.1–3 On the one hand, a case is being made for greater investment in scientific competencies with an eye toward standardizing the undergraduate premedical curriculum. This outlook is most eloquently articulated in a recent report.1 On the other hand, compelling arguments are being put forth in favor of more liberal education with an eye toward diversifying the undergraduate premedical curriculum.4,5 This view is most apparent in the recent revision of the Medical College Admission Test (MCAT), which seeks to increase the emphasis on the “psychological, social, and biological foundations of behavior” in recognition of the importance of “socio-cultural and behavioral determinants of health and health outcomes.”2,6

In 1985, The Warren Alpert Medical School of Brown University (AMS) launched the Program in Liberal Medical Education (PLME). From the outset, this eight-year baccalaureate/MD program (one of approximately 40 in the United States7) sought to provide students with a broad and liberal education, integrating a social science and behavioral perspective into the educational program, while at the same time increasing racial and ethnic diversity at the AMS. In keeping with this mission, the PLME has encouraged entrants to freely pursue their varied interests in the humanities, social sciences, or natural sciences.8 PLME students are admitted to Brown University based on their performance in high school and are granted provisional admission to the AMS subject to the successful completion of the undergraduate degree requirements. When admitting PLME students to the AMS, traditional undergraduate grade point average (GPA) is not considered. Moreover, neither an MCAT score nor an admissions interview is required. While the undergraduate course requirements for PLME students partially overlap with those of students in the traditional premedical route (TPM), PLME students may receive competency credits equivalent to one biology course, one mathematics course, and one physics course by attaining advanced placement in these subjects. PLME students must attain a grade of A or B for each of three core biology courses; other undergraduate science courses (such as physics and chemistry) require a passing grade only and may be taken on a pass/fail basis. The PLME accounts for approximately 50% of all AMS matriculants.

The AMS also matriculates a similar number of students via the TPM wherein careful consideration is given to students’ GPA, MCAT score, and admissions interview. However, once they are at the AMS, all students are subject to the same medical education program. It is this unique juxtaposition of the two AMS admission routes that made it possible for us to aim to determine whether there is an association between highly distinct premedical curricular and admission requirements and medical school performance and residency placement.

Method

The cohorts studied included all of the PLME (n = 295) and TPM (n = 215) students who graduated from the AMS between 2010 and 2015. We obtained demographic (race/ethnicity, SAT scores, GPA, MCAT score data, etc.) and outcome data (see below) for all students through the AMS Oasis system, which serves as the official medical school registration system. Underrepresented in medicine (URM) status refers to students who identified themselves as black or Hispanic. As of the time of publication, residency placement records were not yet available for students in the class of 2015. The institutional review board of Brown University determined the study to be exempt. Because we analyzed data retrospectively and anonymously by assigning each student a distinct number, the institutional review board did not require informed consent from participants.

Outcome variables

Our outcome variables included: the average of the year 1 and year 2 examination scores, United States Medical Licensing Examination (USMLE) Step 1 score, USMLE Step 2 Clinical Knowledge (CK) score, average of the six core clerkship examination (shelf examination) scores, core clerkship honors grades, doctoring honors grades, induction into the Alpha Omega Alpha Honor Medical Society or the Gold Humanism Honor Society, professionalism ratings, participation in a scholarly concentration (SC) program, publication metrics, and residency placement.

We did not include undergraduate GPA and MCAT scores in our analysis as these data were not available for the PLME cohort. (Brown University does not calculate a GPA for any of its students and PLME students do not take the MCAT.) The USMLE Step 2 CK data does not include the class of 2015, for which data were unavailable. The six core clerkships at the AMS are medicine, surgery, pediatrics, obstetrics/gynecology, psychiatry-clinical neuroscience, and family medicine. Students in the doctoring course, which introduces students to clinical medicine during the first four semesters of the medical school curriculum, are eligible for a grade of honors on the strength of demonstrated professional behavior. Further assessment of student professionalism in the clinical setting relied on evaluation by faculty during the internal medicine clerkship. An SC program constitutes an elective opportunity during which students may gain experience in areas such as medical education, global health, or ethics.9 The PubMed search engine was used to identify peer-reviewed publications by PLME and TPM students.

Our analysis of residency placement made use of data provided by the National Resident Matching Program.10 A discontinuous measure of the competitiveness of residency training programs in individual disciplines was defined by a USMLE Step 1 score averaging ≥ 240 (for competitive disciplines) or < 240 (for noncompetitive disciplines). By applying an average USMLE Step 1 score to each individual medical discipline, we created a continuous variable that reflects the competitiveness of the corresponding residency training programs.11 We assessed the competitiveness of the internal medicine residency training programs that AMS graduates matched to by examining the proportion of those programs affiliated with medical schools in the top 10% and the top quartile of the U.S. News & World Report research ranking of U.S. medical schools.12

Statistical analysis

SPSS version 20 (IBM, Armonk, New York) was used for data analysis. We employed unadjusted tests of averages and proportions (independent t tests and chi-squared tests) to compare variables associated with students admitted to the AMS via the PLME or TPM routes. We then performed multiple logistic regression analysis for dichotomous outcome measures and multiple linear regression analysis for continuous outcome measures. Both the logistic and linear regressions controlled for age, URM status, and the number of undergraduate premedical science courses completed.

Results

PLME students comprised 57.8% (n = 295) and TPM students comprised 42.2% (n = 215) of the AMS classes of 2010–2015 (n = 510).

The average SAT scores attained by the PLME cohort were 716/800 (for the verbal section) and 728/800 (for the mathematical section). The average rate at which PLME students matriculated to the AMS was 88% (range: 83%–98% for individual classes); the majority of PLME students who did not matriculate to the AMS matriculated to other medical schools.

While undergraduate GPA and MCAT score data were not available for the PLME cohort, both metrics were considered in the admission of TPM students to the AMS. For the class of 2015, the average GPA of the TPM student cohort was 3.67 and the average MCAT score was 33. GPA and MCAT score data for previous TPM class years were comparable.

The PLME student cohort was more racially and ethnically diverse, as measured by URM status (n = 72 [24.4%] for PLME students vs n = 31 [14.4%] for TPM students; P < .01), and less likely to major in a science discipline (n = 138 [46.8%] for PLME students vs n = 140 [65.1%] for TPM students; P < .001) than the TPM student cohort. Moreover, there was a statistically significant difference in the number of undergraduate premedical science courses completed by the two student cohorts (Figure 1). PLME students completed fewer premedical biology courses than TPM students, though the intercohort difference was modest. However, the intercohort differences in the number of premedical chemistry, physics, and mathematics courses completed were marked, with PLME students completing half as many of these undergraduate premedical science courses as their TPM counterparts.

Figure 1.

Figure 1

Open in a new tab

Number of undergraduate premedical science courses completed by the Program in Liberal Medical Education (PLME) student cohort (black bars) and the cohort of students who matriculated via the traditional premedical route (gray bars), The Warren Alpert Medical School of Brown University, classes of 2010–2015. Data are shown as the average + 1 standard deviation. Asterisks indicate P < .001 versus the PLME cohort.

Relationship between admission route and medical school performance

The average scores attained in standardized multiple-choice examinations revealed marginal, but statistically significant, differences between the PLME and TPM student cohorts (Figure 2). TPM students attained higher scores across a spectrum of tests that included the year 1 and year 2 examinations, USMLE Step 1 and USMLE Step 2 CK examinations, and six core clerkship shelf examinations. These results remain unchanged in linear regression models even after controlling for age, URM status, and the number of undergraduate premedical science courses completed.

Figure 2.

Figure 2

Open in a new tab

Comparison of standardized multiple-choice examinations scores attained by the Program in Liberal Medical Education (PLME) cohort and the cohort of students who matriculated via the traditional premedical route (premedical), The Warren Alpert Medical School of Brown University, classes of 2010–2015. Data are shown as median and 25th and 75th percentiles (boxes) and over a full range of values for each group (whiskers). Asterisks indicate P < .001 versus the PLME cohort.

Students in the TPM cohort received more honors grades in the six core clerkships than their PLME counterparts (average = 2.8 [standard deviation (SD) = 1.87] vs average = 1.9 [SD = 1.75]; P < .001). These results did not differ in linear regression models after controlling for age, URM status, and the number of undergraduate premedical science courses completed. No statistically significant difference was noted for the two student cohorts in the professionalism rating assigned in the internal medicine clerkship. The average rate at which students contributed to the peer-reviewed literature proved similar for the two student cohorts (0.93 [SD = 1.51] papers for PLME students vs 1.04 [SD = 1.35] papers for TPM students; P = .87). These results did not differ in linear regression models after controlling for age, URM status, and the number of undergraduate premedical science courses completed. No statistically significant difference was noted between the two groups in the average impact factor of the journal in which their papers were published (3.22 [SD = 4.01] for PLME students vs 3.59 [SD = 4.63] for TPM students; P = .65).

The Alpha Omega Alpha Honor Medical Society inducted a higher proportion of TPM students than PLME students (Table 1). Similarly, a higher proportion of TPM students achieved an honors grade in the doctoring course. No statistically significant difference was detected between the two groups in the proportion of students inducted into the Gold Humanism Honor Society or the proportion of students enrolled in an SC program. These results did not differ in logistic regression models after controlling for age, URM status, and the number of undergraduate premedical science courses completed.

Table 1.

Dichotomous Variable Outcomes for Students in the Program in Liberal Medical Education (PLME) Admission Route and the Traditional Premedical Route (TPM), The Warren Alpert Medical School of Brown University, Classes of 2010–2015

Outcome variables PLME, no. (%)a TPM, no. (%)a P value
Inducted into the Alpha Omega Alpha Honor Medical Society 21/202 (10.4) 31/121 (25.6) < .001
Inducted into the Gold Humanism Honor Society 32/202 (15.8) 28/121 (23.1) .10
Received honors grade in doctoring 78/295 (26.4) 79/215 (36.7) < .05
Enrolled in a scholarly concentration program 80/295 (27.1) 74/215 (34.4) .08
Matched in a competitive disciplineb 31/155 (20.0) 26/93 (28.0) .26
Open in a new tab
a

Data represent the number of students achieving the outcome variable and the number of students eligible for each category and, in parentheses, the percentage of students the achieving outcome variable for each category.

b

The authors defined competitive disciplines as disciplines for which the average United States Medical Licensing Examination Step 1 score is ≥ 240; the authors calculated results using chi-squared analysis.

We hypothesized that the number of undergraduate premedical science courses completed by PLME students would contribute significantly to their medical school performance. In particular, we anticipated a strong relationship between the number of undergraduate premedical science courses completed and the average year 1 examination score. In fact, the association though statistically significant was weak (r2 = 0.023; P = .02) (Figure 3). We observed a similarly statistically significant but weak association between the number of undergraduate premedical science courses completed and the USMLE Step 1 score (r2 = 0.037; P = .001). No correlation was noted between the number of undergraduate premedical science courses completed and the average shelf examinations (P = .75) or USMLE Step 2 CK (P = .78) scores. We interpreted these data as indicating that the number of undergraduate premedical science courses completed by students in the PLME cohort, while highly variable, accounted for less than 4% of the variance in key metrics of medical school performance. These same relationships were examined for the TPM cohort; in no case was a significant correlation noted.

Figure 3.

Figure 3

Open in a new tab

The relationship between the number of undergraduate premedical science courses completed and medical school performance for Program in Liberal Medical Education students, The Warren Alpert Medical School of Brown University, classes of 2010–2015. Graphs show the relationship between the number of undergraduate premedical science courses completed and the average year 1 examination score (Panel A), USMLE Step 1 score (Panel B), average of the six core clerkship shelf examination scores (Panel C), and USMLE Step 2 CK score (Panel D). Also examined was the relationship between the number of undergraduate premedical science courses completed and the competitiveness of the residency training program attained, as indicated by the discipline-specific average USMLE Step 1 score for the national cohort (Panel E); the authors defined competitive disciplines as disciplines for which the average USMLE Step 1 score is ≥ 240.

Relationship between admission route and residency placement

We used several metrics to assess residency placement for the classes of 2010–2014 (PLME [n = 190] and TPM [n = 132] students). A somewhat greater proportion of PLME students than TPM students entered primary care disciplines (family medicine, internal medicine, pediatrics, or medicine/pediatrics) (109/190 [57.4%] vs 62/132 [47.0%]; P = .07), although this difference was not statistically significant. As described above, we used the average USMLE Step 1 score required for residency placement in a specific medical discipline (a discontinuous variable) as a surrogate for the competitiveness of the corresponding residency training programs. In this regard, no statistically significant difference was noted between the two student cohorts; the continuous variable of the average USMLE Step 1 score for individual medical disciplines yielded virtually identical averages for the two student cohorts (224.3 [SD = 8.11] for PLME students vs 225.2 [SD = 20.14] for TPM students; P = .60).

We carried out an additional analysis to assess the two cohorts’ placement rates with competitive residency training programs affiliated with leading medical schools. Given the absence of a ranking system for residency training programs, we chose to selectively examine placement with internal medicine residency training programs by aligning the residency programs with the annual U.S. News & World Report ranking of the affiliated medical schools. No statistically significant difference in placement was noted between the two cohorts. Among the total of 64 students who applied to internal medicine residency training programs, 61 (95.3%) students matched with residency programs at hospitals that are the primary affiliate of their medical school. Of these 61 students, 33 (54.1%) were in the PLME cohort, while 28 (45.9%) were in the TPM cohort. Students who went on to match with internal medicine residency training programs at hospitals affiliated with medical schools in the top 10% of the U.S. News & World Report ranking represented 16/33 (48.5%) and 12/28 (42.9%) (P = .80) of the PLME and TPM cohorts, respectively. Similar results were obtained for internal medicine residency training programs affiliated with medical schools in the top quartile of the U.S. News & World Report ranking (26/33 [78.8%] for the PLME cohort and 19/28 [67.9%] for the TPM cohort; P = .39).

Finally, we carried out an analysis to determine the correlation between the number of undergraduate premedical science courses completed and residency placement. We found a weak but statistically significant association was discerned between the number of undergraduate premedical science courses completed and residency placement as indicated by the average USMLE Step 1 score required by individual medical disciplines (r2 = 0.023; P = .05).

Discussion

From its inception, the PLME has aimed to bridge undergraduate and medical education and at the same time increase racial and ethnic diversity at the AMS. While the PLME maintains a number of premedical science requirements, an MCAT score has not been required for admission to the AMS. Additionally, though PLME students complete multiple traditional premedical science courses, they often do so on a pass-fail basis, which precludes the calculation of a GPA. The absence of the GPA and MCAT score requirements for matriculation to the AMS encourage PLME students to explore academic areas beyond those traditionally pursued in the course of undergraduate premedical education.

Few studies have examined the performance of students enrolled in combined baccalaureate/MD programs. One recent study of seven combined baccalaureate/MD programs observed no apparent difference in the medical school grades or licensing examination scores the baccalaureate/MD students obtained relative to the scores obtained by students who had taken the TPM.13 Another study observed no difference in academic performance between students in the Early Assurance Program at the Icahn School of Medicine at Mount Sinai and those admitted to the same institution through the TPM.14 Drees et al15 concluded that graduates of the combined baccalaureate/MD program at the University of Missouri-Kansas City outperformed students who had taken the TPM in areas such as problem solving and professionalism. The above notwithstanding, the PLME differs from other combined baccalaureate/MD programs in several important aspects, most notably in waiving the GPA, MCAT score, and admissions interview requirements and in the large percentage of students admitted to the AMS through this route.

Prominent among our findings was that TPM students achieved marginally, but statistically significantly, higher average scores in standardized multiple-choice examinations relative to their PLME counterparts. It is unlikely that this difference can be attributed to an inherent difference in test-taking ability given the excellent SAT scores achieved by PLME students. It is plausible that the rigors of the TPM promote the acquisition of a broad knowledge base, as well as the requisite test-taking skills that are needed to excel in the standardized multiple-choice examination paradigm. It is also plausible that in the TPM students who perform poorly on standardized multiple-choice examinations are being deselected based on their MCAT score. The aforementioned considerations may have also contributed to the intercohort differences in the attainment of honors grades in the doctoring course and in the Alpha Omega Alpha Honor Medical Society induction rates. However, the association between these differences in medical school performance and residency placement, as measured by the average USMLE Step 1 score required by individual medical disciplines, is a weak one. Furthermore, the association between in-training examination scores and future performance as a physician is uncertain at best.16,17

Professionalism ratings proved comparable for the two student cohorts. Similarly, no statistically significant difference was noted in the rate at which the two student cohorts were inducted into the Gold Humanism Honor Society or in the proportion of students electing to participate in an SC program. The scholarly accomplishments of students, as measure by the peer-reviewed publication rate and impact factor of the journal they were published in, proved similar for the two cohorts. In the aggregate, results unrelated to medical school standardized multiple-choice examination performance reveal substantial similarity in medical school performance outcomes for the two student cohorts.

We hypothesized that differences in medical school performance would be related to the number of undergraduate premedical science courses completed by PLME students. Our data did not fully support this hypothesis. Note was made of a weak (although statistically significant) association with year 1 examination and USMLE Step 1 scores but not with shelf examinations and USMLE Step 2 CK scores.

In the absence of data on the performance of medical school graduates during and beyond their residency training, we elected to rely on residency placement as an indirect indicator of the readiness of students for this transition, using the average USMLE Step 1 score required for residency placement in a specific medical discipline. In this regard, PLME students and their TPM counterparts proved comparable.

Given the absence of a system that evaluates and compares residency programs for their quality, we chose to undertake a limited assessment of the competitiveness of the residency placement attained by AMS graduates in internal medicine. In using the U.S. News & World Report ranking of the medical schools affiliated with the residency programs, we assumed that the medical schools’ ranking could serve as a surrogate for the competitiveness of the residency programs in question. This approach also disclosed no statistically significant difference between the PLME and TPM cohorts. We consider the examination of internal medicine residency training programs to have greater validity as a residency placement measure given the relatively large number of subjects involved and the broad range of competitiveness as compared with other medical disciplines (e.g., surgical specialties) for which all resident positions are highly competitive. Our broad interpretation of these results is that PLME students are at no distinct disadvantage with regard to residency placement relative to the TPM student cohort. We also conclude that the number of undergraduate premedical science courses AMS graduates complete is not a strong predictor of success in medical school or of achieving competitive residency placement.

A limitation of our study is the absence of longer-term outcome data that might have allowed for an assessment of the performance of graduates as practicing physicians. Future studies may compare outcome variables such as entering an academic career or being funded by the National Institutes of Health. In addition, the available data did not allow for a comparison of the scholastic performance of the two student cohorts prior to entering the AMS in that we do not have SAT scores for the TPM cohort or MCAT scores for the PLME cohort. Obtaining our TPM students’ SAT scores upon matriculation in the future will allow us to better control for confounders. Finally, given that a single educational site was the source of our data, the generalizability of our findings to other settings may be limited. In particular, the PLME has long provided the AMS with an impetus to develop a curriculum that is effective for students with highly variable premedical science backgrounds.

Conclusions

Our study may serve to inform future decisions as to the relative utility of premedical curricular and admission requirements. Our findings suggest that the association between medical school performance and residency placement and undergraduate premedical curricular and admission requirements is weak. This supports the view that highly distinct curricular and admission requirements may be compatible with comparable medical school and residency placement outcomes. Emphasizing a broader range of premedical competencies in the humanities may constitute an alternative approach in preparing students for the lifelong practice of medicine. We encourage further study to determine the optimal premedical preparation of students.

Acknowledgments

The authors appreciate the assistance of Eileen T. Palenchar, Hilary F. Sweigart, and Kathleen Chien in gathering the data included in this paper and the input provided by Ilene B. Harris, PhD, and David S. Guzick, MD, PhD.

Funding/Support: This study was supported in part by NIH grant R25HD068835 (to P. George and P.A. Gruppuso).

Footnotes

Other disclosures: None reported.

Ethical approval: The study was determined to be exempt by the institutional review board of Brown University.

Contributor Information

Paul George, Associate professor of family medicine and associate professor of medical science, The Warren Alpert Medical School of Brown University, Providence, Rhode Island.

Yoon Soo Park, Assistant professor of medical education, University of Illinois College of Medicine, Chicago, Illinois.

Julianne Ip, Clinical associate professor of family medicine and associate dean of medicine, The Warren Alpert Medical School of Brown University, Providence, Rhode Island.

Philip A. Gruppuso, Professor of pediatrics, The Warren Alpert Medical School of Brown University, Providence, Rhode Island.

Eli Y. Adashi, Professor of medical science and former dean of medicine, The Warren Alpert Medical School of Brown University, Providence, Rhode Island.

References

RESOURCES