Pediatric Resident Career Intentions
Comments by Robert C. Bowman, M.D. [email protected]

Peds traditionally has had significant component of public health and service to the underserved. Our society does invest huge sums of money to encourage underserved locations, but these sums (1 billion since 1970 to family medicine for example) are insignificant compared to billions each year to NIH and GME funding. The pressures to subspecialize and do research are enormous in internal medicine and pediatrics. The institution needs you for many things, especially getting research dollars, supporting residency programs, and doing procedures.

One of the key motivations for the study is to understand how pediatrics can graduate more into research and subspecialties. Be aware of how the data are interpreted. PEDS Journal policy comment: The study, "Pediatric Residents' Career Intentions: Data From the Leading Edge of the Pediatrician Workforce", suggests an increasing number of pediatric residents intend to practice primary care, rather than becoming specialists. Residents are putting more importance on factors such as spouse/family considerations, job security and geographic location than they are on factors such as interest in research. The study concludes that with a projected increase in the number of female pediatricians, and a decline in international medical graduates (who have traditionally been more likely to choose a specialization), pediatric care may continue to shift toward an increased proportion of general pediatricians.

Inclusion of ethnicity of physicians is a very key factor in this study. Few studies include this. AMA Masterfile is the primary source of most workforce studies. In the most detailed AMA database only 70% of physicians have ethnicity data. In my study of FP programs that graduated rural physicians, residencies with a higher % of minority residents (hispanic and black) were less likely to choose rural practice. This does not mean that minorities do not choose underserved locations. URMs are 4 times more likely to do so, 40% vs 10% according to AAMC figures.

Very nice use of methods and description of use. Not a good thing to do regressions and then report bivariate data. Why do a regression? It is very easy to cheat in regressions to put in what you want to find, but they seem to be fair in data entry. Don’t do a Likert Scale and then change the data reported as Essential or Very Important. Also terms are not consistent with any other studies, usually the categories would be Very Important, Important, Somewhat Important, Not Important. Some would consider Essential and Very Important to be similar terms, which may explain their inclusion together. Don't like the debt figures, would prefer final estimates of educational debt for a continuous variable rather than over or under 50k.

Also in my study of rural grad rates of FP programs, other key factors were programs having a rural mission and being in a more rural state. Another was having fewer residencies at the main location other than family medicine. Also having a program director who was the rural contact was a key factor. I suspect in many pediatrics programs, there is a general tone of the program and director that leads to more or less generalists.

A generalist graduation rate study in pediatrics would be most revealing. This study actually has the dependent variable. Just need to add independent variables such as generalist mission, program director interest toward generalism (pro, con, neutral), number of other residency programs, private vs public hospital, private vs public medical school, number of subspecialty faculty, number of generalist faculty (use a ratio between), policy on vacation (do residents have to take vacation on ambulatory rotations), total research funding, etc.)

Did not like the variable for urban inner city practice. There is a big difference in inner city poverty or low income practice and inner city practice. The IMG data might be much different with IMGs staying inner city because of their cultural connections, rather than service to the underserved. The URM data would also be more revealing with this inner city poverty category. Similarly the rural figures are better expressed as rural underserved areas rather than just rural.

Also many ways to express rural. Re: IMG studies - great variability in where IMGs locate. In states like West Virginia they make a great contribution. In the midwest there is not much contribution. The IMG data on rural is likely rural underserved, as they have taken underserved practice commitments to get their Green Card and to be able to stay in the US, using the Agriculture department waiver or some other federal agency.

Recruitment (getting there) vs retention (staying in practice)
• UNMC primary care stats from Wigton study of cohorts around 1980 graduation from UNMC - 33% started rural, 50% did rural, 33% in rural at 12 years later same number went rural to urban as did urban to rural, don't know pediatric individual stats on this.

Many do not understand the concept that rural people want physicians who stay and it may take 4 or more years for significant numbers to come your way in a primary care practice, unless they are desperate. Concept of minimal numbers needed, such as 5000 for peds and IM, 2500 for FP, etc. 4 pediatricians going to Scottsbluff in a short period of time is nice for call but the financial and competitive situation is not good - all trying to start a practice is difficult and did not work. I wrote an article on rural family practitioners noting the dilemma between sharing call and competing by trying to build a practice.

Concept of sharing call, having at least 3 or 4 to share with. Unless your are in a very large city, there may not be enough subspecialists to share call. In Omaha this has been a problem with perinatologists, UMA neurologists, etc. Going from every 4th night to every 3rd or 2nd , usually without a change in salary and with the likelihood of it taking months to a year to replace the lost colleagues makes one strongly consider leaving. Domino effects do happen. Loss of a primary care physician costs $225,000. These are particularly difficult for underserved or academic practices to replace. Subspecialists are even more costly to replace.

Generalists and specialist/researchers are different breeds – concept foreign to authors

Primary Care is about externals such as service to others. Research and subspecialization are about focus and internal issues. To do good research over the long term and be anything less than a selfish whatever is very difficult. A review of social styles may be helpful. Primary care people tend to be amiables and expressives or a mix. Researchers, program directors, chairs, etc. tend to be drivers and analytics. These are different personality types. Understanding how the various types fit together may be helpful in resolving differences and narrowing gaps. Doug Wheatley at our department does social styles testing and orientation with new faculty and residents. This helped me understand how I, as an amiable-expressive, could work with Jim Stageman, our program director, who is an analytic, demanding details and info rather than great concepts and doing good.

The dollars invested in research efforts in family medicine have been enormous and have taken a good chunk of the Title VII dollars that should have gone to increase the distribution of physicians, but the funding has not resulted in the expected numbers of researchers. Research consistently comes up with competition with primary care in family medicine, rural medicine, pediatrics, etc. AAMC studies reveal that some senior students interested in rural practice were interested in research, but the research was more into social issues. National studies of medical schools show that the higher the NIH funding, the less likely a rural choice by graduates of the school (Rosentblatt, JAMA). This study also notes this finding

Marital factors - Another important advance in this study is a division into married, unmarried, and married to a physician. My main suggestion here is that it would be a more useful category to note whether they were married to a professional rather than just a physician. The findings may be clearer. Dual professional families make primary care difficult and rural practice nearly impossible as both have to find a job in a much smaller job market with fewer professional openings.

Gender Issues – there is often an assumption that women do not choose rural practice for various reasons. Single female medical students, and to some degree males, are fed the party line that it is difficult or impossible to find a suitable spouse in rural areas. Since few ever go to rural areas, they have no way to believe otherwise. In FP studies there is some difference in who goes rural, but those who go often stay as long or longer. This is similar to the findings that fewer medical students from private medical schools choose rural, but when they do, they are really committed. In my studies of FP programs, those with more women did tend to graduate slightly fewer into rural practice but this factor did not contribute to the final regression model, the minority variable in particular took it out as a significant factor. The real problem for women in rural practice or in subspecialty practice may be the lack of appropriate role models during medical training to show them that they can balance gender roles and practice. There are some small offices in urban and rural locations where doctors turn their offices into a day care for their smaller children with staff being a part of the family. Increasing business emphasis may be taking this away more and more.

Day care situations, and more colleagues for call may be more important than part time opportunities in enticing more females to do subspecialty fellowships.

Research involving social issues, work environment, public health, or community medicine may induce more females (and other pediatric residents) to become researchers. This allows them to see patients and do research that is more connected to making a difference in practice and in the health of their patients.

Control issues – every physician faces these and all need to develop the ability to say no, or later (subtle form of no) and to be able to delegate. To think that this is a key factor is to deny reality. Some personality types, amiables, have trouble saying no. Some types of practices, underserved, rural, etc. allow more access to patients and there are higher expectations of community duties. Highest on the control issue demands are solo doctors and academic doctors. Flexible time (get away for family events, ability to decide where to make an impact in clinic or research or community) may be a big issue for those interested in academics and recent changes and demands on academic departments may mean perceived threats to this quality may be making it difficult for some to consider academic careers.