Is Medical Student Choice of a Primary Care Residency Influenced by Debt?

Abstract

Context

The average medical student accumulates more than $120,000 in debt upon graduation.

Objective

The purpose of this study was to investigate whether medical student debt affects residency choice.

Design

This was a cross-sectional research study.

Setting

This study was a 5-year analysis of student debt and residency choice for 2001–2005 graduates from 3 US medical schools (n = 2022): Tulane University School of Medicine, New Orleans, Louisiana; University of South Florida College of Medicine, Tampa, Florida; and Louisiana State University School of Medicine in New Orleans.

Main outcome measures

Individual student data were collected from offices of financial aid for debt prior to and during medical school to determine total debt at graduation. Total debt (adjusted for inflation) was compared with residency match results coded according to specialties listed in the Graduate Medical Education Directory 2005–2006. Graduates were coded into either primary care (PC) or nonprimary care (NPC) specialty categories. Logistic regression for the choice of a PC residency was used with 4 predictors: (1) total debt, (2) medical school, (3) year of graduation, and (4) number of years of training required for a residency program.

Results

Mean total debt for the study population was $89,807 (SD = 54,925). Graduates entering PC did not have significantly less total debt than those entering NPC ($87,206 vs $91,430; P = .09). Further, total debt was not a predictor of a PC residency after adjusting for medical school, year of graduation, and years of training in residency (P = .64).

Conclusion

There is no association between PC residency choice and debt. We conclude that medical students make residency decisions on the basis of a complex set of factors.

Introduction

Choosing a specialty is the most important professional decision that medical students make after entering medical school. The factors affecting medical student residency choices are complex and include personal interests as well as practical lifestyle issues. Many studies have examined the relationship between student debt and residency choice, particularly during the 1990s when there was a focus on creating more primary care (PC) physicians.[1–10] Most of these studies relied on surveys, questionnaires, or focus groups and did not relate actual debt to residency choice. One exception is a 1993 study by Spar and colleagues[3] that examined the relationship between actual student debt and specialty choice for medical students graduating from 6 private medical schools from 1988 to 1990. These investigators found no difference in student indebtedness among residency choices. More recent studies with surveys to examine the effect of debt on the specialty choices of graduating medical students have concluded that there is no relationship.[6–10] We undertook this study to investigate the role of student debt on residency choice with data obtained from financial aid offices.

Since 1990, medical school tuition has increased dramatically, and the average debt incurred by US medical students is now averaging over $120,000.[11] We investigated the relationship between indebtedness and residency choice. Unique to this study, actual student debt information was collected directly from the respective financial aid offices of 3 institutions over a 5-year period and compared with student residency matches. The influence of medical school, year of graduation, and number of years of training required for a residency program were part of the study.

Methods

Three institutions participated in this study: one private, Tulane University School of Medicine (New Orleans, Louisiana), and two public, Louisiana State University (LSU) School of Medicine (New Orleans) and the University of South Florida (USF) College of Medicine (Tampa, Florida). The institutional review boards at each of the 3 participating medical schools granted approval to participate in the study. The office of financial aid at each school provided actual student debt, which included both undergraduate and medical school debt for the graduating students from the classes of 2001–2005. Graduation year and residency choice, as determined from the residency match, were also obtained from each medical school. Any identifying information for students was blinded prior to study entry and data analysis.

Residency choices, as listed in the Graduate Medical Education Directory 2005–2006,[12] were coded with the specialty in which the student matched. Plans for fellowship training were not considered in this analysis. The analysis included 2022 graduates over the 5-year study period: 830 from LSU, 715 from Tulane, and 477 from USF. Graduates were not included if they matched into a transitional year (n =12) or into a combined program with PC and 1 nonprimary care (NPC) specialty (eg, internal medicine/neurology).

The residency choices of graduates were classified as either PC or NPC: PC specialties included family medicine, internal medicine, pediatrics, internal medicine/family medicine, and internal medicine/pediatrics. NPC specialties included anesthesiology, child neurology, dermatology, diagnostic radiology, emergency medicine, general surgery, neurology and neurosurgery, obstetrics/gynecology, ophthalmology, orthopaedics, otorhinolaryngology, pathology, physical medicine and rehabilitation, plastic surgery, psychiatry, radiation oncology, and urology.

Statistical Analysis

Debt prior to medical school and debt during medical school were combined to yield total student debt at medical school graduation. Debt was reported in 2001 dollars with a 1.6% adjustment for inflation (yearly Consumer Price Index [CPI]) for the class of 2002, a 3.9% adjustment for the class of 2003, a 6.6% adjustment for the class of 2004, and a 9.6% adjustment for the class of 2005. The CPI reports changes in prices paid by urban consumers for a representative basket of goods and services.

The independent samples t test was used to compare total debt for graduates entering a PC vs an NPC residency program. Logistic regression for the choice of a PC residency was used with 4 predictors: (1) total debt, (2) medical school, (3) year of graduation (2001–2005), and (4) number of years of training required for residency program. Inferences were made at the 0.05 level of significance with no correction for multiple comparisons.

Results

Table 1 describes selected characteristics of the 3 institutions and their students. The private medical school (Tulane) differed notably from the 2 state-funded medical schools in tuition and fees, but the other characteristics in the 3 profiles were only moderately different. The mean total debt at graduation for the 2022 graduates was $89,807 (SD = 54,925). Graduates entering PC specialties did not have less total debt than those entering NPC specialties ($87,206 vs $91,430; P = .09).

Table 1.

Comparison of Medical Schools

Characteristic	Tulane University	Louisiana State University	University of South Florida
Status	Private	State-funded	State-funded
Founded	1834	1931	1971
Male/female (%)	54 of 46	50 of 50	58 of 42
Self-classified as white (%)	74	81	72
Yearly tuition/fees ($)	38,000	8800	10,900
Mean debt ($)	110,000	80,000	70,000
Received financial aid (%)	85	90	80
No debt at graduation (%)	10	18	10
Undergraduate science majors (%)	47	64	56

Table 2 shows the logistic regression for the choice of a PC residency with 4 predictors: total debt, medical school (Tulane, LSU, and USF), year of graduation (2001–2005), and years of training required for residency program. No predictor was significant when entered separately: total debt (P = .09), medical school (P = .10), year of graduation (P = .37), and years of residency training required (P = .20). After adjusting for the other 3 variables, debt was not an independent predictor of choosing a PC residency (P = .64).

Table 2.

Logistic Regression for the Choice of a Primary Care Residency With 4 Predictors: (1) Total Debt, (2) Medical School (Tulane University, Louisiana State University, University of South Florida), (3) Year of Graduation (2001–2005), and (4) Number of Years of Training Required for Residency Program (n = 2022)

Predictor	Significance With 1 Predictor	Significance With 4 Predictors
Total debt	.09	.64
Medical school	.10
Tulane		.53
LSU		.16
USF		.13
Year of graduation	.37
2001		.31
2002		.70
2003		.97
2004		.61
2005		.66
Years of training in residency	.20	.20

LSU = Louisiana State University; USF = University of South Florida

Discussion

This study is unique because it obtained actual debt from the financial aid offices at 3 medical schools over a 5-year period. There was no association between PC residency choice and debt.

In the 1980s, several studies that examined level of debt and residency choice in house officers conflicted as to whether a relationship existed.[1, 13–15] These studies, which were based on survey data, including information derived from the graduation questionnaire (GQ) administered by the Association of American Medical Colleges, suggested that increasing debt levels led to a diminished interest in PC residency choices. After reviewing data from the 1992 GQ, Kassebaum and Szenas[16] concluded that residency choices were not significantly affected by educational debt, and in their 1993 update, they found less association between surgery and surgical subspecialty career choice and debt when compared with the previous year's GQ.[17] Another follow-up of the GQ in 1994 indicated that medical student residency choices were not strongly influenced by educational indebtedness.[5] A 1995 study on focus groups of California medical students found no association between residency choice and student debt,[4] and in a retrospective survey of practicing young physicians, only 3.2% believed that debt had influenced their specialty choice.[10] However, certain subgroups (those with children, those with parents of lower income or education) were more likely to be influenced by their debt. Marci and Roberts[9] suggested that a threshold of debt ($75,000) existed, above which students were more likely to be influenced in their residency decisions. None of these later studies used actual debt figures for graduating medical students. Only Spar and colleagues'[3] data, which reflected students graduating between 1988 and 1990, compared actual student debt and residency choice.

As with previous investigations examining PC, our study is limited by the difficulty in determining whether those who choose careers in internal medicine or pediatrics actually practice PC. If current trends continue, approximately 40% of internal medicine residents will practice a subspecialty, and nearly one third of pediatrics residents will choose a subspecialty.[18] These internal medicine and pediatrics subspecialists may practice little PC medicine. Because these career decisions are not made until at least 3 years after medical school graduation, it is difficult to factor this information into an analysis of debt and residency.

Despite conflicting reports on the relationship between student debt and residency choice, experts agree that choosing a residency is a challenging process and that debt is only one factor in a making this complex decision. Previous reviews have examined other factors and concluded that work hours, lifestyle, and prestige are important in the decision-making process.[19–21] Other social and economic considerations, such as leisure time, work schedules, and controllable lifestyle, may also contribute to residency choice.[22,23]

Several studies have focused specifically on factors related to choosing careers in PC. Some have suggested that experiences during third-year clerkships influence the choice of PC residencies.[24,25] Students report conflicting opinions in regard to the role of projected income in their decision to enter PC.[26,27] Other factors, such as sex,[28] having an interest in technology, and coming from a rural background, may be confounding variables in trying to establish the role of debt in choosing a PC career.[29–31] One recent study even suggested that among students who enter PC residencies, increased student borrowing can have a positive effect on physician-shortage areas through debt repayment programs that require practice in an underserved setting.[32]

During the study period (2001–2005), the GQ showed that nationwide about two thirds of graduating students indicated that educational debt had no influence on their career choices. Further, a small minority (4.1% to 4.6%) of graduating students indicated that educational debt had a strong influence on their residency choices. It is possible that previous studies, which relied on survey data, underestimated student concerns with regard to debt. Although medical school tuition has risen by over 10% during the past 2 years,[33] recent declines in interest rates may make student loans more attractive. Students may be less concerned about debt when questioned directly, but may behave differently when choosing a specialty to support a lifestyle. Finally, the data obtained for this study were more precise than survey responses in that actual financial data obtained from offices of financial aid were compared directly with student residency choices from graduate medical education match lists. However, our study involved 3 Southern medical schools, and the results may not be generalizable to all schools or regions in the United States.

An important factor for medical students in making their career choices may be perceived ability to handle educational debt. During the study period 2001–2005, the number of US medical school graduates entering the fields of internal medicine, internal medicine/pediatrics, family medicine, and pediatrics decreased by 11%.[34] During this same time period, average medical student debt increased by almost 2% per year, and in 2003, the average debt for all medical students was $109,457.[33] Assuming an interest rate of 5.25%, the monthly payment on this amount would be $1609 for 10 years.[35] Further, there is great disparity in physician incomes on the basis of specialty choice, location of practice, type of practice, practice structure, and other variables.

Despite rising education costs, we found no association between residency choice and medical student debt. It has been suggested that many medical students have a poor understanding of debt management and are not aware of the specifics of their own future loan repayment schedules.[4,36,37] Lack of knowledge about debt repayment, unfamiliarity with earning potential, and an inadequate appreciation of financial strategies may lead students to make career decisions independent of their level of indebtedness. Education on debt management and an increase in loan repayment programs providing financial incentives for students interested in PC may help to increase the attractiveness of PC as a career regardless of indebtedness.

Conclusion

We examined actual student debt obtained from the offices of financial aid at 3 medical schools and found no relationship between debt and residency choice, as determined by graduate medical education match results. We conclude that medical students make career decisions on the basis of a complex set of factors. Although it may be a consideration for some students, indebtedness was not an independent predictor of a PC residency for a large sample of graduates from 3 medical schools.