Personal view or opinion piece
Open Access

Medical School Admissions: Defining the Future of Healthcare Delivery in the United States

John Burkhardt[1][a], Mahshid Abir[1], Steven Durning[2]

Institution: 1. University of Michigan, 2. Uniformed Services University of the Health Sciences
Corresponding Author: Dr John Burkhardt ([email protected])
Categories: Educational Strategies, Selection, Research in Health Professions Education
Published Date: 30/09/2019


Issue:  Systemic health disparities in the U.S. are unlikely to be successfully addressed without evidence-based, multilevel policy interventions.  To date, most research investigations and policy interventions have been focused on incentivizing career choice aimed at ameliorating health disparities and correcting medical bias in practicing doctors.  Much less work has been undertaken on utilizing educational policy, practices and data to affect positive change “upstream” of the current practicing healthcare workforce.  An underutilized area in educational policy and predictive analytic use is in the medical school admissions process.  Admissions policies for selective educational institutions have been tailored in the past to provide for the public good and therefore represent a natural first intervention area.  

Evidence:  In the past, numerous policies have been enacted to incentivize physician behavior in ways to produce more primary care physicians, increase the physician work force in underserved areas, and in general improve health care equity.  Despite these interventions, health disparities because of physician workforce shortcomings persist.   With regard to medical school admissions approaches, several current and past standards have demonstrated that the selection process can be modified to alter the incoming cohort of student-doctors.

Implications:  We propose that medical education policy has not been sufficiently leveraged to act in concert with existing health service-based interventions.  Specifically, we argue the medical school admissions process represents an underutilized opportunity to fundamentally change the physician workforce and address downstream health care system issues.  Expanding what may constitute desirable for selection based on societal needs versus academic credentials is one possible remedy to the current approach.  An opportunity also lies in leveraging the power of large data methods to provide predictive analytical input into the admissions process.  Reconceptualizing the admissions process in these ways could allow for a complementary strengthening of the current programs designed to ease health disparities.  In addition to the continuation of current diversity-enrichment programs in health professional education, expansion to more novel and cross-disciplinary solutions, such as those discussed in this article, could be actively encouraged.

Keywords: Admissions; Diversity; Health Disparities; Policy Reform; Predictive Analytics


Admissions policies are once again at the forefront of public discourse as a result of “Operation: Varsity Blues (Pascus, 2019).”  In that sting operation by the FBI, numerous affluent parents allegedly bribed and manipulated admission criteria in order to gain admission for their children to selective undergraduate programs.  While the alleged bribery and money laundering were far from simple in nature, two factors that are generalizable to a discussion on admissions policy are the relative value of standardized testing and the adoption of more lenient admission standards in special cases (athletes, legacies, etc.). 

Standardized testing, and its relative importance in educational section processes, have been discussed in detail in other works (Phelps, 2005; Donnon, Paolucci and Violato, 2007; Davis et al., 2013).  However, “Operation: Varsity Blues” demonstrates how those in power can manipulate admissions policies in order to provide unfair advantages for some (Alon and Tienda, 2007; Alon, 2009; Alon, 2014).  It has been alleged that not only did the students involved in the case utilize high-priced tutoring available to wealthy families they also manipulated accommodations for special needs students and in some cases bribed proctors to take the tests on the behalf of their children (Pascus, 2019).  Whatever one’s opinions on the use of standardized testing in admissions, this level of manipulation calls into question how objective these results might be in certain circumstances.

Similarly, athletics can be seen as a measure of applicant achievement,(Fried, 2007) a potential means for increased campus diversity (Fried, 2007), and both a financial (Brown, 1996) and reputation boon (Clark et al., 2009; Jones, 2009) for the higher educational institution.  As such, differential criteria for admission for some applicants (Fried, 2007) can be a rational approach.  However, once again, the recent scandal demonstrates how manipulation of admissions criteria can result in “crew recruits” from schools without crew teams (Pascus, 2019). While the evidence presented from “Operation: Varsity Blues” might cause someone to grow cynical about admissions policies it also provides an opportunity to consider how the admissions process could be reformed to act in the service of the public good instead for those most willing to cheat. 

To date, similar scandal as seen in “Operation: Varsity Blues” has not touched medical school admissions.    Medical school admission is the decisive moment of entrance to the profession in the United States (U.S.) and therefore is arguably the most important selection process in medical education for determining the physician workforce.  According to the Association of American Medical Colleges (AAMC) in 2014, 28,125 (57% of applicants) individuals who completed a pre-medical curriculum, took the Medical College Admission Test (MCAT), and applied to at least one medical school had their applications declined (Association of American Colleges, 2015).  Given the high financial cost of the application process (Patterson, Baldwin and Olsen, 2009) and other opportunity costs for excluded applicants, as well as the increasing need for physicians (Scheffler et al., 2008; Association of American Medical Colleges, 2010), using a selection model that is efficient and optimized is critical.  Ongoing large-scale discussions of the fairness of admissions may allow for a reconsideration of which criteria are used in the basis of medical school admissions. We propose that medical school admissions process should be reframed as a major opportunity to define healthcare delivery in the next generation. 

Medical students are typically selected after consideration of a number of factors including grade point average (GPA), admissions tests (MCAT in the U.S.), letters of recommendation, volunteer and research experiences, and interviews.  Admission decisions are based on the limited information available to the admissions committee, customarily using an applicant’s past performance to predict future success (Eva et al., 2004; Donnon, Paolucci and Violato, 2007).  We believe admissions decisions are based on professional norms and established selection factors.  However, how these constructs reflect today’s physician workforce needs deserves close evaluation.  Previously, studies on the medical school admissions process have primarily focused on examining the ability of typical selection factors to predict short-term educational outcomes such as medical clerkship performance or national licensing examinations (Eva and Reiter, 2004; Eva et al., 2004; Donnon, Paolucci and Violato, 2007).  The resultant literature largely informs the predictive validity of commonly collected admissions committee data on some educational outcomes, while the overall importance of admissions decisions in determining health work force capacity has not been explicitly explored (Eva and Reiter, 2004; Eva et al., 2004; Kreiter et al., 2004). 

Given the 90-95% likelihood of medical school completion once admitted, the physician population of the health care system becomes largely set at the time of admission.  We propose programmatic evaluation of admission decisions beyond completion of medical school, to include consideration of the number of admitted students that fill key health care system needs 5-10 years after their admission.  Some of these outcomes we would propose have been considered in the past, such as the production of primary care physicians at medical schools such as Michigan State University College of Human Medicine (Switzer et al., 2018).   However we would argue for a much more data-driven approach based on predictive analytics in medical school admissions.  Such an approach could move beyond a focus on culture creation and general consideration of applicant career interest toward the use of predicted probability consideration of students most likely to practice primary care in the future.  Future practice with underserved populations and those most likely to address persistent health and health care disparities could similarly be selected.

Applicant selection based on socioeconomic class has been explored in the business literature and can be considered in the medical education context.  Persistent social stratification along socioeconomic status has been explained in elite firm professional hiring practices through bias in the interview process.  Here, non-academic factors associated with higher social status become strong predictors of interview evaluation scores (Rivera, 2011).  For example, there can be an increased likelihood of hiring applicants with common experiences in expensive camps, participation in selective social and eating clubs, or sports that are more common amongst the elite (e.g. crew, lacrosse).  Bonds created through these common life experiences can result in unconscious bias towards certain applicants and are often expressed unintentionally versus as explicit discrimination.  Reassigning admissions interviews so that they are more often completed by those in areas of health care system need, may help to co-opt this form of bias.  If we recruit those that are similar to us, utilizing interviewers practicing in areas of need may increase the likelihood of meeting system needs.  Whether as a result of one of the above mechanisms, or others, there are persistent systemic issues of under-representation in medicine for some minority groups (The Rationale for Diversity in the Health Professions: A Review of the Evidence, 2006) and that is  impacting the ability of the health care system to prepare for the future.   

The medical school selection process does not occur in a vacuum and admissions policies have impacts outside equity and single institutional concerns.  We believe disproportionately low minority representation disadvantages medically underserved groups.   African Americans and other minority groups can have issues of trust with the medical community based on a history of mistreatment and experimentation.  Pervasive distrust of the medical profession is an ongoing issue in many minority communities (Corbie-Smith, Thomas and George, 2002).  African-Americans have especially been the victims of inhuman experimentation by the medical community including during their period of enslavement in the United States (Gamble, 1993) as well as more recently in the infamous Tuskegee Syphilis trial (Gamble, 1997).  Continued mistrust of physicians is multifactorial (Kennedy, Mathis and Woods, 2007) and results in enduring concerns about participation in medical research (Shavers, Lynch and Burmeister, 2002) as well as fears of “being experimented on” or financially “exploited” during routine medical care (Armstrong et al., 2008).  This persistent distrust by patients of color is maintained, even when controlling for socioeconomic factors and geographical differences (Corbie-Smith, Thomas and George, 2002), though these effects may also have complex intersectional effects (Armstrong et al., 2007).  Distrust of medical providers also extends beyond patient-physician interactions and influences the trust felt between pediatric patents’ parents and their child’s physician (Rajakumar et al., 2009).  While historical mistreatment is likely an ongoing contributor to feelings of mistrust (Gamble, 1997), a lack of cultural competence (Kennedy, Mathis and Woods, 2007) and implicit bias on the part of medical professionals may perpetuate these issues (Dovidio et al., 2008).  Mistrust of this type continues to this day (Saha et al., 1999) and is an important consideration as the population moves toward a majority-minority state (Maxwell, 2014; Colby and Ortman, 2017).

Under-represented Minority (URiM) physicians treating medically underserved populations has been shown to correlate with increased trust in the provider and improved clinical outcomes (The Rationale for Diversity in the Health Professions: A Review of the Evidence, 2006)  Additionally, URiM physicians choose primary care practice in underserved populations at a higher rate than other students (The Rationale for Diversity in the Health Professions: A Review of the Evidence, 2006).  “Traditional majority” students who study with URiM students have been shown to also practice in medically underserved populations at higher rates, report increased capability to care for minority patients, and a heightened interest in equitable health access (The Rationale for Diversity in the Health Professions: A Review of the Evidence, 2006).  A continued lack of equal representation in the medical profession from URiM groups can thus be seen as sub-optimally serving members of those communities.  Similarly, efforts have been made to predict students most likely to address persistent physician shortages for underserved rural populations (Puddey et al., 2015).  Once again, students who have similar experiences as the underserved population, in this case growing up in small rural communities, are more likely to choose to serve in those same communities (Puddey et al., 2015).

Several proposed mechanisms have been explored in the medical literature to explain biases and disparities in health care delivery (Caterinicchio, 1979; Thom and Campbell, 1997; Thom et al., 1999; Thom, 2001; Nelson, 2002; Thom et al., 2002; Smedley, Stith and Nelson, 2003b; Smedley, Stith and Nelson, 2003a; Staff, 2004; Egede, 2006; Traylor et al., 2010).  One major reported mechanism resulting in care disparities are underlying biases of caregivers and their application of stereotypes in treating patients of color (Smedley, Stith and Nelson, 2003b; Staff, 2004).  A related mechanism is greater levels of provider uncertainty in the delivering of care to underserved minority patients (Smedley, Stith and Nelson, 2003b; Staff, 2004).  Of equal importance is the role of systemic inequality that results from the administration of health systems as a whole, including health care plan managers and other economic considerations in health care delivery (Smedley, Stith and Nelson, 2003a; Staff, 2004).  Of specific interest with regard to how admissions decisions may impact these issues is through consideration of patient-physician concordance (Saha et al., 1999; Cooper and Powe, 2004; Street et al., 2008; Traylor et al., 2010; Greenwood, Carnahan and Huang, 2018; Shen et al., 2018) and issues of patient-physician trust (Caterinicchio, 1979; Anderson and Dedrick, 1990; Gamble, 1993; Kao et al., 1998; Doescher et al., 2000; Corbie-Smith, Thomas and George, 2002; DeVille and Kopelman, 2003; Piette et al., 2005; Armstrong et al., 2007; Kennedy, Mathis and Woods, 2007; Armstrong et al., 2008; Dovidio et al., 2008; Rajakumar et al., 2009).  As such, we focus on this aspect of the multifactorial origin of health disparities for patients from historically underserved populations.   

Physician-Patient trust can be affected by physician behavior and the level of this trust impacts medical outcomes:  Physician-patient trust has been studied for decade (Caterinicchio, 1979) and with increasing focus with the expansion of managed care in the United States.(Thom and Campbell, 1997)  Researchers have attempted to quantify (Anderson and Dedrick, 1990; Thom et al., 1999; Epstein et al., 2005) physician-patient trust in order to study its effects.  Within these studies, physician attributes (Weng, 2008), behaviors (Thom, 2001; Thom et al., 2002; Kim, Kaplowitz and Johnston, 2004), and even the insurance mechanisms of payment (Kao et al., 1998) impact patient-physician trust.  Of specific importance to the consideration of admission decisions is the demonstrated benefit of shared personal and ethnic backgrounds between physician and patient (Saha et al., 1999; Doescher et al., 2000; Street et al., 2008).  Patient behavior, especially compliance with physician recommendations, can be influenced by their level of trust in their health care provider (Piette et al., 2005).  Physician-patient trust and rapport has even been correlated with decreased hospitalizations and emergency room utilization in the elderly population (Parchman and Burge, 2004).  As such patients, who have less access to care and those who receive worse care even when access is equal might have both forms of disparity ameliorated through a data-driven, societal need-based admissions process.

While the intervening time in training between admissions and practice may weaken the associations between applicant characteristics and eventual career behavior, medical school admission is an essential and decisive moment for determining the physician pipeline and thus one of the final points where meaningful intercession is most likely to be possible.  What we propose is a frameshift from research focused on explaining behavior to include one that incorporates the ideas of machine learning and predictive analytics focused on precise prediction without specific or clear causation (Yarkoni and Westfall, 2017).  In this way we would add to the current research approach, one that is decisively outcome focused over one that is explanatory focused.  Machine learning can come in many forms (Mair et al., 2000), however for the purposes of our discussion we believe it most important to utilize this new technology to make connections between existing data not immediately apparent to human researchers (Kourou et al., 2015). Utilization of predictive analytics in medical school admissions policies are not unknown (J. C. Burkhardt et al., 2016; J. Burkhardt et al., 2018), but are generally limited in scope and application.  We believe that research focused on data mining (Witten et al., 2016) of associations on a national scale may yield the types of information that could be implemented to address current disparities.  This is exploratory work by nature.  However not being able to fully predict these results must not preclude us from expanding what might be of possible in reforming our health education and health care systems.

The problems of health care inequality that are related to a lack of physician background diversity are unlikely to change without intervention.  Current medical school selection policies may therefore result in intensification of health disparities, having a major impact on the future of healthcare delivery in the United States.  Further, a substantial physician shortage is projected in the U.S. by the Department of Health and Human Services (U.S. Department of Health and Human Services, 2008).  This shortage is due to both supply (e.g. freeze on governmental Graduate Medical Education funding, aging physician population) and demand (e.g. increased per-patient health care utilization) shortfalls (U.S. Department of Health and Human Services, 2008).  If the medical profession wants to meet the needs of the health care system with physicians, and not mid-level providers, educators need to rapidly expand the number of medical students while maintaining or improving the level of competence in the profession.  While the physician shortage is likely to affect the larger U.S. population, minority and low SES patients are likely to be most impacted with increased wait times to get care and poorer clinical outcomes (U.S. Department of Health and Human Services, 2008).  Therefore, we believe that  medical school admissions policy in the future must be cognizant of their potential role in mitigating future health disparities through thoughtful integration of workforce development needs into their selection goals.   

Given the potential connections between student selection and projected health care workforce needs, new research to guide reevaluation of medical school admission policies and practices is critical.  We propose focusing selection criteria research on additional outcomes.  Selection criteria that can demonstrate a positive predictive value for placing graduates into areas of physician shortage (specialty and location-based), better addressing health disparities, and/or other explicit policy needs should be considered in addition to commonly collected data. Medical school admissions rely on a belief in a meritocratic process; whereas fairness, equity, and societal benefit also deserve close consideration.  Medical school admissions reform should guide selection of candidates with the highest likelihood of future success and responsiveness to the needs of society to include various previously underserved communities.  This is a matter of social justice and equity and can likely be accomplished without a decrease in practicing physician ability. 

How we select applicants for entry to the medical field is a critical societal issue that extends beyond professional boundaries and must be considered in the larger context of the societal contract between physicians and the general population.  As a profession, we need to commit to studying the medical school selection process using robust research practices. Establishing the benefits of the current selection factors and exploring novel considerations should be a policy and arguably is a research priority.  By creation of an expanded medical education literature on admissions, which has a foundation in theory and uses large-scale data to draw its conclusions, medical educators can redefine the evidentiary basis for our admissions policies and shape the future of healthcare delivery.  As medical educators we must embrace our potentially hidden additional responsibility for determining the quality of care for our fellow citizens and be more active in advocating for change through our selection policy choices. 

Take Home Messages

  • Systemic health disparities in the U.S. are unlikely to be successfully addressed without evidence-based, multilevel policy interventions
  • An underutilized area in educational policy and predictive analytic use is in the medical school admissions process
  • The medical school admissions process represents an underutilized opportunity to fundamentally change the physician workforce and address downstream health care system issues
  • An opportunity also lies in leveraging the power of large data methods to provide predictive analytical input into the admissions process
  • Admissions policy changes could complement the current health services programs designed to ease health disparities

Notes On Contributors

John Burkhardt, M.D., Ph.D. is an Assistant Professor in the Departments of Emergency Medicine and Learning Health Sciences at the University of Michigan Medical School.  His research focuses on critically reevaluating medical education policy to increase fairness, equity, and ultimately address health disparities. ORCID iD

Mahshid Abir, M.D., M.Sc. is an Assistant Professor in the Department of Emergency Medicine at the University of Michigan Medical School.  Dr. Abir’s research focus is in evaluating the continuum of acute care delivery in the United States, including pre-hospital, emergency, inpatient, and ambulatory care with a focus on addressing policy-related questions pertaining to utilization, quality, efficiency, outcomes, and costs of acute care delivery in these settings.

Steven Durning, M.D., Ph.D. is a Professor in the Department of Medicine at the Uniformed Services University of the Health Sciences.  His current research interests include enhancing our understanding of clinical reasoning and its assessment to include unraveling the phenomenon of context specificity and incorporating biologic measures into this empirical work.


Disclaimer: The views expressed herein are those of the authors and not necessarily those of the Department of Defense or other federal agencies.


Alon, S. (2009) 'The Evolution of Class Inequality in Higher Education: Competition, Exclusion, and Adaptation', American Sociological Review, 74(5), pp. 731-755,

Alon, S. (2014) 'Continuing to Build a Theory of Inequality in Higher Education: Claims, Evidence, and Further Directions', American Sociological Review, 79(4), pp. 817-824,

Alon, S. and Tienda, M. (2007) 'Diversity, Opportunity, and the Shifting Meritocracy in Higher Education', American Sociological Review, 72(4), pp. 487-511, 

Anderson, L. A. and Dedrick, R. F. (1990) 'Development of the Trust in Physician scale: a measure to assess interpersonal trust in patient-physician relationships', Psychological reports, 67(3_suppl), pp. 1091-1100,

Armstrong, K., McMurphy, S., Dean, L. T., Micco, E., et al. (2008) 'Differences in the patterns of health care system distrust between blacks and whites', Journal of general internal medicine, 23(6), pp. 827-833,

Armstrong, K., Ravenell, K. L., McMurphy, S. and Putt, M. (2007) 'Racial/ethnic differences in physician distrust in the United States', American journal of public health, 97(7), pp. 1283-1289,

Association of American Colleges (2015) The Official Guide to Medical School Admissions. [Online]. Available at:

Association of American Medical Colleges (2010) Physician Shortages to Worsen Without Increases in Residency Training. [Online]. Available at:

Brown, R. W. (1996) 'The revenues associated with relaxing admission standards at division IA colleges', Applied Economics, 28(7), pp. 807-814,

Burkhardt, J., DesJardins, S. L., Teener, C. A., Gay, S. E., et al. (201) 'Predicting Medical School Enrollment Behavior: Comparing an Enrollment Management Model to Expert Human Judgement', Academic Medicine, 93 (11S) pp. S68-73. 

Burkhardt, J. C., DesJardins, S. L., Teener, C. A., Gay, S. E., et al. (2016) 'Enrollment Management in Medical School Admissions: A Novel Evidence-Based Approach at One Institution', Acad Med, 91(11), pp. 1561-1567.

Caterinicchio, R. P. (1979) 'Testing plausible path models of interpersonal trust in patient-physician treatment relationships', Social Science & Medicine. Part A: Medical Psychology & Medical Sociology, 13, pp. 81-99,

Clark, J. S., Apostolopoulou, A., Branvold, S. and Synowka, D. (2009) 'Who knows Bobby Mo? Using intercollegiate athletics to build a university brand', Sport Marketing Quarterly, 18(1), p. 57,

Colby, S. L. and Ortman, J. M. (2017) 'Projections of the size and composition of the US population: 2014 to 2060: Population estimates and projections',

Cooper, L. A. and Powe, N. R. (2004) Disparities in patient experiences, health care processes, and outcomes: the role of patient-provider racial, ethnic, and language concordance. Commonwealth Fund New York, NY., last accessed: September 26, 2019

Corbie-Smith, G., Thomas, S. B. and George, D. M. M. S. (2002) 'Distrust, race, and research', Archives of internal medicine, 162(21), pp. 2458-2463,

Davis, D., Dorsey, K., Franks, R. D., Sckett, P. R., et al. (2013) 'Do Racial and Ethnic Group Differences in Performance on the MCAT Exam Reflect Test Bias?', Academic Medicine, 88(5), pp. 593-602,

DeVille, K. and Kopelman, L. (2003) 'Diversity, Trust, and Patient Care: Affirmative Action in Medical Education 25 years After Bakke', Journal of Medicine and Philosophy, 28(4), pp. 489-516,

Doescher, M. P., Saver, B. G., Franks, P. and Fiscella, K. (2000) 'Racial and ethnic disparities in perceptions of physician style and trust',

Donnon, T., Paolucci, E. and Violato, C. (2007) 'The Predictive Validity of the MCAT for Medical School Performance and Medical Board Licensing Examinations: A Meta-Analysis of the Published Research', Academic Medicine, 82(1), pp. 100-106,

Dovidio, J. F., Penner, L. A., Albrecht, T. L., Norton, W. E., et al. (2008) 'Disparities and distrust: the implications of psychological processes for understanding racial disparities in health and health care', Social science & medicine, 67(3), pp. 478-486,

Egede, L. E. (2006) 'Race, ethnicity, culture, and disparities in health care', Journal of general internal medicine, 21(6), pp. 667-669,

Epstein, R. M., Franks, P., Fiscella, K., Shields, C. G., et al. (2005) 'Measuring patient-centered communication in patient–physician consultations: theoretical and practical issues', Social science & medicine, 61(7), pp. 1516-1528,

Eva, K. W. and Reiter, H. I. (2004) 'Where judgement fails: pitfalls in the selection process for medical personnel', Adv Health Sci Educ Theory Pract, 9(2), pp. 161-74.

Eva, K. W., Rosenfeld, J., Reiter, H. I. and Norman, G. R. (2004) 'An admissions OSCE: the multiple mini-interview', Med Educ, 38(3), pp. 314-26,

Fried, B. H. (2007) 'Punting our future: College athletics and admissions', Change: The Magazine of Higher Learning, 39(3), pp. 8-15,

Gamble, V. N. (1993) 'A legacy of distrust: African Americans and medical research', Am J Prev Med, 9(6 Suppl), pp. 35-38,

Gamble, V. N. (1997) 'Under the shadow of Tuskegee: African Americans and health care', American journal of public health, 87(11), pp. 1773-1778,

Greenwood, B. N., Carnahan, S. and Huang, L. (2018) 'Patient–physician gender concordance and increased mortality among female heart attack patients', Proceedings of the National Academy of Sciences, p. 201800097,

Jones, A. (2009) 'Athletics, applications, & yields', College and University, 85(2), p. 10,

Kao, A. C., Green, D. C., Zaslavsky, A. M., Koplan, J. P., et al. (1998) 'The relationship between method of physician payment and patient trust', Jama, 280(19), pp. 1708-1714,

Kennedy, B. R., Mathis, C. C. and Woods, A. K. (2007) 'African Americans and their distrust of the health care system: healthcare for diverse populations', Journal of cultural diversity, 14(2), p. 56,

Kim, S. S., Kaplowitz, S. and Johnston, M. V. (2004) 'The effects of physician empathy on patient satisfaction and compliance', Evaluation & the health professions, 27(3), pp. 237-251,

Kourou, K., Exarchos, T. P., Exarchos, K. P., Karamouzis, M. V., et al. (2015) 'Machine learning applications in cancer prognosis and prediction', Computational and structural biotechnology journal, 13, pp. 8-17,

Kreiter, C., Yin, P., Solow, C. and Brennan, R. (2004) 'Investigating the Reliability of the Medical School Admissions Interview', Advances in Health Sciences Education, 9, pp. 147–159.,

Mair, C., Kadoda, G., Lefley, M., Phalp, K., et al. (2000) 'An investigation of machine learning based prediction systems', Journal of Systems and Software, 53(1), pp. 23-29,

Maxwell, L. A. (2014) 'US school enrollment hits majority-minority milestone', Education Week, 34(1), p. 1,

Nelson, A. (2002) 'Unequal treatment: confronting racial and ethnic disparities in health care', Journal of the National Medical Association, 94(8), p. 666,

Parchman, M. L. and Burge, S. K. (2004) 'The patient-physician relationship, primary care attributes, and preventive services', FAMILY MEDICINE-KANSAS CITY-, 36(1), pp. 22-27,

Pascus, B. (2019) 'Every charge and accusation facing the 33 parents in the college admissions scandal', CBS News. Available at:  (Accessed: May 31, 2019).

Patterson, D. G., Baldwin, L.-M. and Olsen, P. M. (2009) 'Supports and obstacles in the medical school application process for American Indians and Alaska Natives', Journal of health care for the poor and underserved, 20(2), pp. 308-329,

Phelps, R. (2005) Defending Standardized Testing. Psychology Press.

Piette, J. D., Heisler, M., Krein, S. and Kerr, E. A. (2005) 'The role of patient-physician trust in moderating medication nonadherence due to cost pressures', Archives of internal medicine, 165(15), pp. 1749-1755,

Puddey, I. B., Mercer, A., Playford, D. E. and Riley, G. J. (2015) 'Medical student selection criteria and socio-demographic factors as predictors of ultimately working rurally after graduation', BMC Med Educ, 15, p. 74.

Rajakumar, K., Thomas, S. B., Musa, D., Almario, D., et al. (2009) 'Racial differences in parents' distrust of medicine and research', Archives of pediatrics & adolescent medicine, 163(2), pp. 108-114,

 The Rationale for Diversity in the Health Professions: A Review of the Evidence (2006).

Rivera, L. (2011) 'Ivies, extracurriculars, and exclusion: Elite employers' use of educational credentials', Research in Social Stratification and Mobility, 29, pp. 71-90,

Saha, S., Komaromy, M., Koepsell, T. D. and Bindman, A. B. (1999) 'Patient-physician racial concordance and the perceived quality and use of health care', Archives of internal medicine, 159(9), pp. 997-1004, https://doi.rg/10.1001/archinte.159.9.997

Scheffler, R. M., Liu, J. X., Kinfu, Y. and Dal Poz, M. R. (2008) 'Forecasting the global shortage of physicians: an economic- and needs-based approach', Bulletin of the World Health Organization, 86(7), pp. 497-576,

Shavers, V. L., Lynch, C. F. and Burmeister, L. F. (2002) 'Racial differences in factors that influence the willingness to participate in medical research studies', Annals of epidemiology, 12(4), pp. 248-256,

Shen, M. J., Peterson, E. B., Costas-Muñiz, R., Hernandez, M. H., et al. (2018) 'The effects of race and racial concordance on patient-physician communication: A systematic review of the literature', Journal of racial and ethnic health disparities, 5(1), pp. 117-140,

Smedley, B. D., Stith, A. Y. and Nelson, A. R. (2003a) 'Assessing potential sources of racial and ethnic disparities in care: patient-and system-level factors', in  Unequal treatment: confronting racial and ethnic disparities in health care.   National Academies Press (US).

Smedley, B. D., Stith, A. Y. and Nelson, A. R. (2003b) 'Assessing potential sources of racial and ethnic disparities in care: the clinical encounter', in  Unequal treatment: confronting racial and ethnic disparities in health care.   National Academies Press (US).

Staff, I. o. M. (2004) Unequal treatment: Confronting racial and ethnic disparities in healthcare. National Academies Press.

Street, R. L., O’Malley, K. J., Cooper, L. A. and Haidet, P. (2008) 'Understanding concordance in patient-physician relationships: personal and ethnic dimensions of shared identity', The Annals of Family Medicine, 6(3), pp. 198-205, 

Switzer, R., VandeZande, L., Davis, A. T. and Koehler, T. J. (2018) 'Are Michigan State University medical school (MSU-CHM) alumni more likely to practice in the region of their graduate medical education primary care program compared to non-MSU-CHM alumni?', BMC Medical Education, 18(1), p. 113.

Thom, D. H. (2001) 'Physician behaviors that predict patient trust', The Journal of family practice, 50(4), pp. 323-328,

Thom, D. H. and Campbell, B. (1997) 'Patient-physician trust: an exploratory study', The Journal of family practice, 44(2), pp. 169-176,

Thom, D. H., Kravitz, R. L., Bell, R. A., Krupat, E., et al. (2002) 'Patient trust in the physician: relationship to patient requests', Family practice, 19(5), pp. 476-483,

Thom, D. H., Ribisl, K. M., Stewart, A. L., Luke, D. A., et al. (1999) 'Further validation and reliability testing of the Trust in Physician Scale', Medical care, pp. 510-517,

Traylor, A. H., Schmittdiel, J. A., Uratsu, C. S., Mangione, C. M., et al. (2010) 'The Predictors of Patient–Physician Race and Ethnic Concordance: A Medical Facility Fixed‐Effects Approach', Health services research, 45(3), pp. 792-805,

U.S. Department of Health and Human Services (2008) The Physician Workforce: Projections and Research into Current Issues Affecting Supply and Demand.

Weng, H.-C. (2008) 'Does the physician's emotional intelligence matter?: Impacts of the physician's emotional intelligence on the trust, patient-physician relationship, and satisfaction', Health care management review, 33(4), pp. 280-288,

Witten, I. H., Frank, E., Hall, M. A. and Pal, C. J. (2016) Data Mining: Practical machine learning tools and techniques. Morgan Kaufmann.

Yarkoni, T. and Westfall, J. (2017) 'Choosing prediction over explanation in psychology: Lessons from machine learning', Perspectives on Psychological Science, 12(6), pp. 1100-1122,




There are no conflicts of interest.
This has been published under Creative Commons "CC BY-SA 4.0" (

Ethics Statement

Formal ethical approval was not sought as this work is an opinion paper that expresses the authors' viewpoint. While it is based of the current literature, no additional human subject research was performed in the creating of this manuscript. As such there were no research subjects that could be put at risk and was outside of the purview of IRB review.

External Funding

This article has not had any External Funding


Please Login or Register an Account before submitting a Review

Ken Masters - (14/02/2020) Panel Member Icon
An interesting piece describing problems of admission policies to medical schools as they impact on healthcare delivery in the US.

Some issues:

• Sub-headings: Even though the normal sub-headings do not apply in an opinion piece, some sub-headings in this piece would have been useful as signposts. The argument does develop very well, but it develops in stages, and identifying these stages would allow readers to follow the path more easily. This is especially important in order to distinguish between the current situation, policies and practices, and those proposed by the authors.

• I do like several of the proposal given, although I’m concerned with the lack of detail. While the paper is strong on describing the problems, it is somewhat weak on the details of the solutions. While I realise that an opinion piece cannot give a great deal of detail, it does require some, especially as several of these proposals have been suggested before, and then collapsed (or abandoned) when the implementation detail becomes apparent. A particular example is “Applicant selection based on socioeconomic class”. This is not a new idea – the problem is how? Are the authors suggesting a quota system? How would it be managed? How would the socioeconomic class be measured, especially given that it is not fluid? How would it be compared to the population? Would regional differences be taken into account, and how? What about ethnicity and gender, and what if a person claims to be from one group, when visually they appear not to be so? Would DNA and other intrusive tests be required? Using the examples that the authors use: what if a person came from a really poor background, but had been able to finish school at an expensive school (e.g. through a scholarship), or had received a sponsorship to attend an academic training camp – would that count against them? What if a person has recently moved to the US, and they were not in a minority group in their own country, but they are in the US? Are medical schools prepared to go to court if sued by a student with superior grades who does not gain admission simply because they are a member of a majority group? These are issues that cannot be left for later, as it is exactly these kinds of issues that cause problems and lead to program failure. It would have useful if the authors could point to literature (even non-US) where these questions have been appropriately answered.

• Although the proposal calls for machine-learning as a solution, the authors really do need to tread carefully here. The questions outlined above are problematic because, no matter the decision, somebody will be aggrieved. Transferring the decision-making to an Artificial Intelligence system does not solve the real problems – perhaps the only problem it solves is culpability: the medical school can blame the computer, rather than the administrator or a committee. There are several examples of how biases (e.g. racial and gender) have been (usually inadvertently) built into AI systems, and any AI admissions system would need to be carefully designed, otherwise the school will simply hide behind the “black box” and shrug, but transferring decision-making does not transfer ethics.

So, overall, the paper is strong on identifying the problems, and proposing that we need methods of admission that deal with these problems, and that they should be data-driven. I don’t think there is much argument there, and these problems have been recognised for many years (as is evidenced by the literature cited here and elsewhere). The question is how do we deal with the details? It is in dealing with these that the authors’ expertise and insights are really required.

Possible Conflict of Interest:

For transparency, I am an Associate Editor of MedEdPublish.

Nandalal Gunaratne - (01/10/2019)
This is based on admission policies in the US medical schools. However the concerns raised about underdeveloped areas and health inequity is common to all countries.
The opinions and take home messages are apt. My concern is that picking those from underprivileged areas may not necessarily work for the benefit of those in such areas.
Medical school is 5 to 7 years. And those who come in can radically change in those years and may try their best to avoid their places of origin to work in better areas, larger hospitals and in specialised areas due to social and economic benefits.
The way a "worth" of a clinician is measured by us and society must also change.
Possible Conflict of Interest:


Felix Silwimba - (30/09/2019)
informative and trans-formative article.