INTRODUCTION

Primary care is an integral part of patient care, emphasizing prevention, screening, and maintenance of well-being. Communities served by primary care physicians (PCPs) are linked to overall improved health.1 , 2 Concerns exist, however, about an increasing shortage of adult PCPs to care for an aging population.3 – 5 Fewer physicians now enter into outpatient primary care,6 and there is a looming attrition rate of PCPs, with almost one-quarter of practicing PCPs in 2011 being 56 years of age or older.7 Current strategies to address the primary care shortage include payment structure reform,8 , 9 implementation of team-based primary care with patient centered medical homes,10 , 11 as well as increasing the number of primary care based graduate medical education residency positions.12 , 13

Many internal medicine residency programs have emphasized ambulatory medicine by creating primary care tracks that are designed to highlight outpatient medicine. Professional and governmental organizations, including the Health Resources and Services Administration (HRSA) and the Council on Graduate Medical Education (COGME), endorse and support primary care residency programs in order to help promote a workforce that consists of at least 40 % outpatient primary care physicians as compared to the 2010 level of 32 %.7

While several studies have examined which aspects of medical school curricula might pique medical students’ interest in primary care careers,14 – 16 the effect of primary care curricula in residency training on resident career choice is less well documented. Recent studies of primary care programs reveal that participation in a primary care track significantly increased intent to practice primary care at graduation from residency, but importantly, these cohorts were not followed into practice to determine whether stated career choices persisted.17 , 18 Other studies evaluating small primary care tracks focused on training PCPs for underserved patient populations have demonstrated that the majority of these individuals pursue a primary care career. These studies, however, lacked comparison groups.19 , 20

Given the limitations of existing data, the effectiveness of primary care residencies in supporting an outpatient generalist career choice compared to categorical programs remains uncertain. Our study examines whether participation in a primary care internal medicine residency is associated with a subsequent career in outpatient primary care amongst 10 years of graduates from the University of California, San Francisco (UCSF) primary care residency cohort compared to their categorical colleagues.

METHODS

Design

We used survey data to perform a cross-sectional study comparing the careers of primary care and categorical alumni from the UCSF internal medicine residency program between 2001 and 2010. After 2010, there was widespread implementation of a pathways to discovery program, which enabled residents to participate in immersion curricula in various areas, including international health, health systems leadership, medical education, and underserved medicine, as well as the already existing pathways of molecular medicine and clinical outcomes research. We consequently chose to study the years before 2011 when all categorical residents at UCSF received an internal medicine curriculum that emphasized clinical skills and academic development without the addition of a focused area of specialty interest.

Setting and Participants

The UCSF internal medicine residency program includes a traditional categorical program, two primary care residencies, and a focused research track in Molecular Medicine. Each program has a separate match number, and residents choose between these tracks based on their training preference. The UCSF Primary Care General Internal Medicine track (UCPC) is designed to develop leaders in general medicine through dedicated training in ambulatory medicine. The UCSF San Francisco General Primary Care track (SFPC) trains PCPs who will care for urban underserved patient populations. Lastly, the Molecular Medicine program is primarily composed of MD-PhD bench researchers who ultimately fast-track into basic science research careers.

From 2001–2010, 513 residents graduated from the UCSF internal medicine residency program, of which 137 completed UCPC, 55 completed SFPC, and 321 completed the categorical program. All alumni included in this analysis were American medical school graduates.

Procedures

We collected residency alumni email addresses through direct contact, social media, professional websites and word of mouth, then surveyed 481 alumni whose contact information we acquired. A senior faculty member (JK) emailed alumni a personal link to a 27-question survey using Qualtrics software, a web-based survey program. We emailed our survey weekly for three consecutive weeks in February 2013 or in September 2013, depending on when alumni contact information was obtained. All alumni received identical emails and follow-up communication. Participation in the survey was voluntary. Responses were de-identified for analysis. This study received approval from the UCSF human subjects institutional review board and the San Francisco Veterans Affairs Research and Development Committee.

We developed our survey based on a review of previously utilized residency alumni surveys, input from residency faculty and local survey experts, feedback from current residents, and input from senior advisors from the UCSF Academy of Medical Educators. The survey was then piloted at two separate residency research “work-in-progress” sessions with non-participants to ensure question clarity. This survey was nested within a larger questionnaire examining alumni careers, mentorship, scholarship and interest in underserved medicine.

Alumni were asked to provide demographic information such as age, gender, race/ethnicity, as well as the training program each alumnus participated in during residency, namely, categorical, primary care, or Molecular Medicine. Alumni were also asked to report whether the majority of their current clinical work is in outpatient primary care, to state their interest in an outpatient primary care career pre-residency and post-residency, and to identify whether the ambulatory experience during residency influenced career choice (Appendix 1 available online). All items were rated on a five-point Likert scale (1 = strongly disagree to 5 = strongly agree).

Analysis

We used training program, namely, categorical versus primary care, as our primary predictor variable. Primary care residents were defined as graduates from UCPC or SFPC programs, while categorical residents were defined as all other respondents. Our primary outcome variable was whether the majority of a respondent’s current clinical work is in outpatient primary care. Responses were dichotomized such that alumni who answered “strongly agree” and “agree” were characterized as having a current career in primary care, while alumni who answered “strongly disagree,” “disagree,” and “neutral” were characterized as not having a current clinical career in primary care. Secondary outcome variables included change of interest in a primary care career during residency, and whether the ambulatory medicine experience during residency influenced career choice. Secondary outcome responses were similarly dichotomized.

We ran descriptive statistics for all variables and performed Chi-Square analysis to compare the career outcomes of primary care and categorical alumni. We did not run multivariate analysis due to small sample size, but we used a Bonferroni correction (alpha = 0.0125) to account for multiple comparisons. We used SPSS version 21 for all analyses.

RESULTS

Of the 481 alumni emailed, 322 alumni responded to our survey for a 67 % response rate. A slightly higher percentage of responses were from categorical alumni with a 70 % response rate as compared to a 64 % response rate from primary care graduates. From the 322 respondents, we excluded 42 graduates who self-identified as current fellows, as they had not yet started their faculty careers and we could not be sure of their final practice type. We also excluded the five graduates from the Molecular Medicine residency, as their research-oriented training made them poor candidates for our control group. Six graduates completed the survey twice, and of those, we analyzed only the first completed survey. For this analysis, 275 participants were included in order to compare 106 responses from former primary care residents to 169 responses from former categorical residents. Two of the respondents did not answer all questions, resulting in numbers less than 275 in the denominator for some questions. Graduates from the categorical and primary care programs were similar with respect to year of graduation, age, and race/ethnicity. A higher percentage of the primary care alumni are female (Table 1). The two primary care residencies were similar demographically, except the UCPC cohort had a slightly higher percentage of Caucasians and Asian-Indians while the SFPC cohort had a slightly higher percentage of Asian-Other. We conducted a regression analysis across the years of graduation comparing primary care and categorical residents for their subsequent career choice and found the slope of the line across the years to be essentially 0 for both groups. These findings suggest no effect of time on the outcomes of interest.

Table 1 Demographic Characteristics of Categorical vs. Primary Care Track Alumni
Full size table

Our primary outcome revealed that 57/106 (54 %) of graduates from the primary care track agreed that the majority of their current clinical work is in outpatient primary care medicine as compared to 34/160 (20 %) of categorical residency graduates; p < 0.001. The difference in subsequent primary care careers between the two UCSF primary care programs’ alumni was small and not statistically significant (50 % for UCPC and 63.3 % for SFPC, p = 0.22).

Our secondary outcomes revealed that while 98/106 (93 %) of primary care alumni agreed or strongly agreed that they were interested in a career in primary care prior to starting residency, 67/106 (63 %) agreed or strongly agreed that they were interested in a career in primary care at the end of residency (Table 2).

Table 2 Primary Care Career Interest Pre-Residency to Post-Residency and Influence of the Ambulatory Medicine Experience
Full size table

This decline in interest in a primary care career may be partially or fully explained by the finding that 88/106 (83 %) of primary care alumni felt that their residency ambulatory clinic experience influenced their subsequent career choice. Notably, 30 of the 34 (88 %) primary care residents who lost interest in a primary care career during residency agreed or strongly agreed that their ambulatory experience during residency influenced their career choice (Table 2).

When focusing on the particular subgroup of individuals who had an initial interest in an outpatient primary care career pre-residency, 23/38 (61 %) of these categorical alumni lost interest in an outpatient primary care career during residency as compared to 34/98 (35 %) of primary care residents; p = 0.006 (Table 3).

Table 3 Change in Interest in Primary Care During Residency in a Subgroup of Individuals with an Initial Interest in Primary Care Pre-Residency
Full size table

DISCUSSION

Fifty-four percent of primary care residency alumni continued on to a career in outpatient primary care compared to only 20 % of categorical alumni, demonstrating that training in a primary care residency is more frequently associated with an outpatient generalist career. It is striking, however, that 35 % of alumni in primary care programs who were interested in a primary care career before residency lost their initially stated interest during residency. A majority of these alumni agreed that the ambulatory experience during residency influenced their career choice. At the same time, very few categorical alumni were convinced to pursue a primary care career by the end of residency. Our study demonstrates that while primary care residency programs better foster careers in primary care than categorical programs, there is a significant atrophy of interest in primary care that occurs amongst GIM trainees during residency.

Several potential reasons may account for this loss of interest in primary care during residency. Foremost amongst them is the ongoing inpatient focus of IM residency training. While most of medicine is practiced in the outpatient realm, the majority of resident training occurs in inpatient settings,21 even within primary care residencies. In many programs, residents only practice part-time in clinic and are not adequately protected from concurrent inpatient duties.22 , 23 Given the noted strain between competing inpatient and outpatient responsibilities,24 the continuity clinic experience for IM residents can become stressful and there is variable program compliance with ACGME ambulatory training requirements.25 In an attempt to uncouple inpatient and outpatient training, a few new program models have developed, which include a year long ambulatory block26 and an alternating 4:1 week inpatient to ambulatory schedule27 – 29 that might help to support ambulatory education. Emphasizing outpatient training and education during residency while simultaneously providing protected time to practice ambulatory medicine may help retain resident interest in primary care.

In addition to maximizing protected outpatient training time, there have been multiple appeals for a reform of the clinic experience itself24 , 30 – 34 as satisfaction with IM residency continuity clinic directly influences career choice toward primary care.31 , 32 A recent study compared internal medicine residents’ likelihood of entering GIM as a result of clinic to likelihood of entering a career in GIM before clinic, and demonstrated that 28 % were less likely to enter GIM as a result of their clinic experience, 59 % had no difference in likelihood, and only 11 % were more likely to enter general IM as a result of clinic.34 Current attempts to improve the outpatient clinic experience include the development of patient-centered medical homes and accountable care organizations,35 , 36 efforts to decrease panel size, and the creation of multidisciplinary, inter-professional teams.10 These changes should help to train residents in team environments that allow physicians to work at the peak of their training levels and in more satisfying work conditions.36 Additionally, increased support of non-face to face patient care encounters as well as ambulatory payment reforms37 might entice debt-ridden residents38 , 39 to enter the primary care field.

As insurance coverage broadens in the era of the Affordable Care Act, more PCPs are required to care for an aging population. It is therefore essential to determine which educational efforts ultimately promote careers in primary care amongst future physicians. Calls have been made for a re-evaluation of Graduate Medical Education (GME) funding,40 and accountability measures will likely grow increasingly important.7 , 41 – 43 Our study has attempted to evaluate the outcomes of primary care residencies, and our results suggest that more work may be needed to improve the existing primary care training experiences. Identified outcomes including retention of primary care interest throughout residency and an ultimate career choice in primary care are likely to be important factors to consider in any future education reform.

This study has several limitations. First, it examined alumni from a single academic center and may not be generalizable to other primary care residencies. Second, we collected cross-sectional data and consequently were able to generate associations, but we cannot prove any causal relationships. In addition, our modest sample size precluded multivariate analyses to adjust for covariates potentially impacting career plans. Third, our data covers a 10-year time frame during which external innovations and changes occurred within GIM that may have influenced our results. For example, the advent of hospitalist medicine may affect the number of residents entering outpatient primary care. While there may also have been changes within the UCSF training programs during the study time, when we analyzed outcomes by year, we found no major trends or differences between groups. Fourth, our outcomes were self-reported, so it is possible that those individuals who participated in primary care tracks might overemphasize their interest in a primary care career before or after residency. The responses of primary care alumni might also be influenced by current attitudes towards primary care. Additionally, those individuals interested in primary care may have had undergraduate medical education experiences that prompted an interest in primary care prior to residency. These residents likely self-select into primary care tracks, thereby potentially minimizing the effects of the program itself. Our results, however, reveal a higher retention of interest in a primary care career amongst graduates of primary care residencies as compared to graduates of categorical programs, suggesting that primary care residencies may reinforce interest in a primary care career better than categorical programs.

In this survey, we did not address whether the clinic experience affected career choice positively or negatively, differences in qualitative and quantitative aspects of ambulatory training between primary care and categorical residencies, or whether the growth of hospitalist medicine might have affected outcomes. These are important areas to explore in future studies.

CONCLUSION

Our study surveyed practicing physicians after graduation from residency to determine whether there might be an association between primary care residency training and a consequent career in primary care. Our results suggest that resident interest in primary care declines during residency, even when residents are in dedicated primary care tracks. We also found that ambulatory clinic experience influenced eventual career choice amongst those who decided not to pursue a career in ambulatory general medicine. Improving and emphasizing the outpatient clinic experience during residency may be an important factor to further interest in primary care and warrants further research.