Worldwide, there has been a trend away from primary care by both allopathic and osteopathic medical school graduates in recent years.1–2 This is a cause for concern, because the availability of primary care directly and favorably affects the health status of local communities, socioeconomic regions, and countries.3–4 We carried out the research reported below to identify reliable predictors of students’ choosing primary care as their specialty of choice.
In the 2003–2004 academic year, we asked 2,345 fourth-year osteopathic medical students from 21 colleges of osteopathic medicine and two branch campuses to fill out and return a questionnaire on their plans for and opinions about their graduate medical education. The number of graduating fourth-year osteopathic medical students was reported in the American Osteopathic Association (AOA) Fact Sheet as 2,713.5 Our sample was formed from the records submitted by each college to the American Association of Colleges of Osteopathic Medicine at the time of the survey. Our sampling frame represented 86% of the graduating seniors in the 2003–2004 academic year. The questionnaires were sent to each college of osteopathic medicine with directions to distribute them to their graduating seniors. The students returned their questionnaires to the schools, which forwarded the responses to us. The number of students responding to each question varied across the survey. As it will be shown, questions regarding debt and, in particular, amount owed, was a sensitive question for the medical students.
The questionnaire was 14 pages long and covered many aspects of the students’ experiences, including financial aspects of medical school, future career plans, opinions on their educational experience, and their thoughts about osteopathic medicine. The present report centers on the factors that distinguish those osteopathic medical students who chose primary care specialties (PCSs) from those who chose non-primary-care, nonsurgical specialties (NPCSs). For this report, we define primary care specialties as family practice, pediatrics, and general internal medicine.
To determine the influence of debt load on specialty choice, in our questionnaire we asked all 2,345 seniors the following set of questions.
* To the best of your knowledge, what was your total education-related debt (undergraduate debt and medical school debt) after completing your medical school education?
* How much of your education-related debt do you still owe?
* What impact did your debt load have on your choice of a specialty to pursue?
This last question was answered using a Likert scale to classify responses: no impact, minor impact, moderate impact, and major impact.
Questionnaires were returned from 1,882 students from 21 schools, an 80% return rate. Respondents were from 15 private colleges (1,353 students) and six public colleges (529 students). A total of 576 (approximately 30%) of the responding students planned to enter one of the primary care areas of practice, whereas NPCSs were elected by 1,030 (55%) of the group. Responses from the 174 students going into surgical-related specialties, 9% of the respondents, were substantially different from those of the other students, and were not combined with the data from either group. Data from the 66 students who were undecided were also dropped from further analysis. The following analyses were based on responses of the 1,642 remaining students, who made up 87% of the respondents and 70% of all students who were surveyed. Totals for each question may vary, because all students did not answer all questions.
For the set of questions about debt, the number of students who answered each question also varied question by question. Only 850 students (51%) answered the question on total debt, only 930 (56%) students answered the question on education-related debt still owed, and only 943 (57%) students answered the question about the impact of debt on specialty choice. Results of the analysis are based on the number of students who actually responded to the question; no attempt was made to impute values for missing data.
We performed a straightforward analysis of the students’ responses to the three questions about debt.
* Regarding the question about total education-related debt, the average debt of the 850 students who responded was $152,000, with a median debt of $160,000. The minimum debt was $8,200, and the maximum was $400,000.
* Regarding the question about education-related debt still owed, the average debt of the 930 students who responded was $149,000, with a median debt of $160,000. The minimum debt was $5,000, and the maximum was $400,000
* Regarding the question about the impact of debt on specialty choice, 592 of the 943 respondents (62.8%) reported no impact, 187 (19.7%) said minor impact, 128 (13.6%) reported moderate impact, and the remaining 36 (3.8%) said there was a major impact. The average debt of the respondents was $152,000, with a median debt of $160,000. The minimum debt was $5,000, and the maximum was $400,000.
However, a clear progression in the average amount of remaining debt in each of the four impact categories was demonstrated:
* No impact: $125,811
* Minor impact: $153,578
* Moderate impact: $166,400
* Major impact: $170,722
In addition, we looked at the proportion of respondents in each impact category who fell above or below the median reported value ($160,000) of debt still owed. Figure 1 illustrates that relationship.
For respondents reporting no impact, the majority (331; 59%) owed less than the overall median debt amount; for those reporting a major impact, 10 (fewer than 30%) owed less than $160,000. The data show that for these respondents, there was a clear relationship between remaining debt and their views of debt’s impact on specialty choice, but was there a link between impact and PCSs versus NPCSs? Figure 2 demonstrates that there was. Although it might be argued that a very small minority of physicians in practice report a major impact, the link between these students’ views of a high debt load and choice of an NPCS as a career is evident.
Information and conclusions about the role of debt and its influence on primary care choice are conflicting. Rosenblatt and Andrilla6 conclude that debt does influence choice of primary care after looking at the 2002 data from the Association of American Medical Colleges. Kahn et al7 conclude that debt has no influence on primary care choice after looking at three medical schools located in the South. For the present study, we surveyed a large percentage of all U.S. fourth-year osteopathic medical, and our findings strongly suggest that there is an influence of debt on PCS.
As Table 1 shows, substantially fewer students, both men and women, were entering PCSs than NPCSs. That being said, a PCS was of interest to a higher percentage of women. Table 1 also shows a similar pattern among those who are married and those who are not. Only those who chose a PCS were more likely to be married than those who chose an NPCS. Table 1 also shows other characteristics of those respondents who chose a PCS and who chose an NPCS.
There was also a relationship between the medical schools students attended and their career choices. As mentioned earlier, this study surveyed 21 colleges of osteopathic medicine, and, of these, 6 were state supported and 15 were private colleges (and 2 branch campuses which are affiliated with private colleges of osteopathic medicine). The major divisions were private osteopathic medical schools and public osteopathic medical schools. The percentage of students entering PCSs ranged from 17% to 61% across the osteopathic medical colleges surveyed (data not shown). There was a 6% increase in the percentage of students entering PCS from private colleges relative to those from public colleges. The data are shown in Table 1.
Other factors influencing PCS or NPCS choice
Other factors that influenced students’ specialty choices are shown in Table 2. Dealing with people more than techniques was a major or strong influence for 508 (88%) of those going into PCSs, and this was followed by intellectual content of specialty and possessing the skills now. Prestige and income ranked very low. Dealing with people more than techniques was also deemed important by almost two thirds of those planning an NPCS rather than a PCS, but these students ranked intellectual content of specialty and possessing the skills now as having a greater impact on their choice than dealing with people more than techniques. Technical skills, lifestyle, academic environment, research, and even prestige and income were cited as being highly influential more frequently by this NPCS group than by those interested in a PCS.
Related studies on allopathic students
Senf et al8 looked at factors influencing allopathic students’ choosing family practice as a specialty. Their study found a small but persistent association between rural background, lower socioeconomic status, and the level of parents’ education and students’ choosing family medicine. Faculty role models and school mission were also positively related to choosing family practice. Student’ perceptions of family practice and nonprimary care were distinctive in that those choosing nonprimary care had concerns regarding level of prestige, low income, and breadth of knowledge associated with primary care. The role of debt was unclear in the study.
Compton et al9 studied allopathic medical students from the class of 2003 at 15 schools and found that as students progressed through their medical school training, women were more interested in primary care than were men, and those interested in nonprimary care were influenced by prestige and income.
There is also international evidence on factors influencing choice of nonprimary care versus primary care (family practice). Dikici et al10 found that the reasons given by first-year medical students from Turkey choosing nonprimary care were better financial opportunities, prestige, personal development, more benefits for the patient, and wanting to work in an urban area. It should be noted that pediatrics was a preferred specialty in this group.
Dorsey et al11 looked at factors influencing specialty choice among allopathic students and showed that lifestyle (classified as controllable or noncontrollable) accounted for 55% of the variability in specialty preference from 1996 to 2002 after controlling for income, work hours, and years of graduate medical education required to complete specialty training. It is worth emphasizing that primary care specialties were classified as having an uncontrollable lifestyle, whereas emergency medicine and dermatology were classified as controllable.
In our study, there were few differences in short-term or long-range plans between osteopathic students choosing a PCS or an NPCS. The option for the osteopathic medical student to pursue postgraduate training is, at this time, available either through residency programs approved by the AOA or the Accreditation Council for Graduate Medical Education (ACGME). Significantly fewer of those students planning a primary care practice planned to do so via a residency approved by the AOA (157; 31%) versus an ACGME-approved residency (260; 52%). If the program is dually accredited, meaning that the program has the approval of both the AOA and the ACGME, another (82; 16%) of those planning to enter a PCS said they would opt for it. These data are shown in Table 1.
After completing the postgraduate portion of training, the resident is eligible to take a set of examinations administered by experienced individuals in the same area of concentration. This set of experiences is referred to as the “boards.” Passing these examinations attests to the public and peers that competency in the particular specialty area has been achieved. These board examinations are administered by two organizations, the AOA, which is responsible for assessing those individuals who graduate residency programs approved by it and become doctors of osteopathic medicine, and the American Board of Medical Specialties (ABMS), which evaluates, through various specialty groups, those residents who graduate from programs approved by the ACGME. The ABMS concerns itself with allopathic residency program graduates. Although both groups of students preferred the option of sitting for both the AOA and ABMS boards to that of sitting for either individual board, fewer of those pursuing careers in primary care medicine desired this dual option. PCS respondents were fairly evenly divided between sitting for the AOA boards and sitting for both AOA and ABMS boards, with the ABMS board alone a distant third option. In contrast, 610 (60%) of the students in the NPCS group wanted to take both boards, compared with 257 (45%) of the PCS group. Fewer of the NPCS group—only 212 (about a fifth)—planned to sit for the AOA only, compared with 250 (44%) of the PCS group. In addition, almost as many of these NPCS students wanted to sit only for the ABMS boards as for the AOA boards. These data are shown in Table 1. To further separate these responses, we asked whether dually accredited (AOA-ACGME) residency programs were more appealing than residency programs accredited by the AOA only. Table 1 shows that this was the case in both PCS and NPCS groups. The PCS students favored the dual option by a 2:1 ratio, and the NPCS students favored it by a 3:1 ratio.
As Table 1 shows, income and prestige were seen as less important by PCS students in determining a specialty than by their NPCS counterparts. Consistent with this lower ranking of income on their choice to pursue primary care, students chose this career anticipating lower initial incomes with a smaller rate of increase than the rate of increase that they assumed their NPCS colleagues would enjoy. By their 10th year of medical practice, the PCS group projected making $90,000 less than their NPCS counterparts expected to make at that time. This finding is shown in Table 3. In comparing anticipated locations for practice after residency, those opting for a PCS were more likely to select a small city or town than those choosing an NPCS. About 290 (50%) of the PCS group planned on living in cities with populations of 100,000 or fewer compared with 275 (27%) of the NPCS group. The data are presented in Table 1.
The responding PCS students were twice as likely as their NPCS colleagues to anticipate setting up practice in towns or cities with a population of 100,000 or fewer. The school the student attended was a factor, with the private schools having a slightly greater percentage of their graduates opting for PCSs, although the majority of students chose NPCSs. Those entering PCSs did so because their interest in dealing with people superseded interest in prestige, income, and an academic environment. Those in the PCS group were fully aware that their incomes would probably be less than those of their NPCS colleagues. Those who chose PCSs tended to be married with dependents. The majority of respondents in this study preferred dually accredited programs, whether they chose a PCS or an NPCS. There are modifiable factors (e.g., debt) and nonmodifiable factors (e.g., gender) that influenced choice.
We suggest that positively influencing the modifiable risk factors will increase the probability but not the certainty of students’ choosing primary care as a medical specialty among osteopathic medical students.