NIH investment in predoctoral physician-scientist training. The NRSA F30 awards (current NIH-funding opportunity announcement PA-23-260; formerly PA-21-049) provide up to six years of support for students in dual-degree programs receiving institutional support from the NIH’s National Institute of General Medical Sciences (NIGMS) through the Medical Scientist Training Program (MSTP). A separate F30 funding opportunity announcement (PA-23-261) exists to fund dual-degree predoctoral trainees at institutions that do not have MSTP support.

According to the NIH Research Portfolio Online Reporting Tools Expenditures and Results (RePORTER), the first predoctoral NRSA Individual Fellowships specifically for MD-PhD trainees (F30) were awarded in 1990 by the National Institutes of Alcohol Abuse & Alcoholism (NIAAA) and the National Institute of Mental Health (NIMH), with 7 fellowships totaling $157,091. Figure 1A describes data obtained from the NIH RePORTER Advanced Project Search (4) showing a steady increase in the number and funding of F30 projects between 2000 and 2022, culminating in 833 awards spanning 17 NIH Centers and Institutes, amounting to more than $37 million in total costs in fiscal year 2022 (FY2022). The corresponding number of F30 projects and total direct costs through the F30 mechanism awarded to Yale University in FY2000–FY2022 is shown in Figure 1B, with 24 active F30 projects totaling $860,791 in FY2022. Figure 1C shows that the steady increase in the number of F30 applications received by NIH tracks with the increase in national MD-PhD student enrollment (5–8), indicating that the proportion of students applying for these awards remains constant (between 10% and 12% of all enrolled students since 2015). The term “national enrollment” includes students in all years of an average 8-year MD-PhD training program, < 0.5% of whom are ineligible for federal funding mechanisms (9). In addition to the F30 award mechanism, eligible MD-PhD students may apply for F31 or F31-Diversity NRSA Predoctoral Individual Fellowships, which support PhD training. The F31-Diversity fellowship mechanism (PA 23-271) for MD-PhD students supports up to six years of both research and clinical training, whereas the F31 fellowship mechanism (PA 23-271) supports up to five years of mentored research training while conducting dissertation research. However, as the NIH RePORTER database does not specify which students supported by an F31 or F31-Diversity are dual-degree candidates; only F30 data are shown in Figure 1, A and C. For Figure 1, D and E, we combined the F30 and F31 (referred to as F30/F31) data to show that the steady increase in the number of NRSA applications follows the increase in total MD-PhD student enrollment at Yale, with a slightly higher proportion of students (14%) overall submitting applications compared with the national data in Figure 1C. Figure 1E presents the applications and awards data by matriculation cohort rather than by FY, showing the actual number of students who applied, who were funded, and who did not submit an NRSA application.

Comparing NIH F30 application and award numbers to Yale MD-PhD Program submission and award data. (A and B) Data from the NIH RePORTER using the advanced projects search “project details” category F30 for FY2000–FY2003 to plot (A) the number (black) and total annual expenditure (purple) for F30 projects supported across all participating NIH Institutes (A) and all F30 projects awarded to the “organization” Yale University (B). (C) Data from the NIH Data Book for the number of F30 applications received (green squares) and awards made (pink squares) across all participating NIH institutes; AAMC FACTS tables provided the Total MD-PhD Enrollment by U.S. Medical School. (D) The Yale University Research Enterprise Operations Sponsored Awards database data for all F30, F31-Diversity, and F31 (MD-PhD students only) applications submitted in FY2003–FY2023 follows a similar rate of growth as the increasing enrollment into the Yale MD-PhD Program. (E) The number of students in each matriculation cohort (MD-PhD students entering each year) who submitted (green circles) or were awarded (pink circles) an NRSA (F30 or F31). Since 2012, when the workshops were introduced, the goal has been to have all eligible students prepare and submit an NRSA fellowship application, minimizing the number who did not apply (black circles).

Figure 1D shows a steep increase in awards beginning in 2012 (2008 matriculation cohort in Figure 1E) that corresponds with the introduction of formalized grant-writing workshops for MD-PhD students. A drop in the number of submissions in FY2017 anecdotally corresponds to a one-year hiatus in workshop offerings, and a smaller decline in 2022 may reflect the lack of in-person workshops and coaching during the COVID-19 pandemic. These dips in submissions do not correspond to a drop in the number of eligible applicants, suggesting the possibility that the workshop offerings may drive or affect the number of student submissions. While the percentage of enrolled students applying for and being awarded F30 fellowships at Yale are slightly higher than national percentages, without a comprehensive survey of other MD-PhD programs to determine the availability of grant-writing workshops and extent of support, a causal relationship between fellowship success and grant-writing courses remains unclear. However, the adoption of a formalized curriculum to provide MD-PhD students with foundational grant-writing skills remains an essential component of our overall effort to provide rigorous training in leadership and research management for physician-scientists.

Extramural funding outcomes of the Yale MD-PhD program. As of academic year 2022–2023, 158 currently enrolled students in the Yale MD-PhD Program are supported by several sources, as shown in Figure 2A. The MSTP training grant and its supplements are primary sources of support for 43 MD-PhD students (27%) who are mostly in the first two years of training. Once students affiliate with their doctoral research department and advisor, the stipend and tuition support source shifts primarily to the dissertation sponsor’s federal and nonfederal research funding and university funds. A positive outcome of 10 years of grant-writing training is that over one-third of eligible students currently enrolled (54 students; 34%) have or have had extramural fellowship funding, and an additional 14 (9%) have submitted extramural funding applications with award decisions currently pending. As shown in Figure 2B, the majority of currently enrolled students with extramural funding received F30 (41 students; 75%) compared with F31 and F31-Diversity (7 students; 13%) awards. The remaining students (6 students; 11%) are supported by fellowships from foundations and public charities, such as the Paul & Daisy Soros Foundation, the Robert Wood Johnson Foundation, and the American Heart Association. A second positive outcome of support and rigorous training is that 72% (34 of 48) of the current NRSA (F30, F31, F31-Diversity) fellowship applications were awarded on the first submission, with the remaining applications requiring resubmission (Figure 2C).

Sources of support for Yale MD-PhD students. (A) 158 current (academic year 2022–2023) students in the MD-PhD Program are supported by funds from investigator research grants and university funds (30% + 9% students with pending NIH fellowship applications), the NIGMS MSTP training grant (27%), and from extramural fellowships (34%). (B) Current students who are or have been supported by an NRSA F30 award (75%), an NRSA F31 award (13%), or private foundation (Robert Wood Johnson, The Wenner-Gren Foundation for Anthropological Research, Inc., Daisy and Paul Soros Foundation) or public charity (American Heart Association) (11%). (C) Of the NRSA F30 and F31 fellowships awarded, the majority (72%) is successful with the original application whereas 28% are awarded upon resubmission.

Seventeen of 20 total students in the 2019 matriculation cohort have prepared and submitted F30 or F31 applications; of these, seven have been awarded fellowships thus far. Typical reasons for not applying include ineligibility due to citizenship, a thesis project in the humanities that is not aligned with NIH funding priorities, or delays in starting dissertation research. Students who miss the eligibility window of 48 months after matriculation for F30 submission are highly encouraged to prepare F31 applications within their fifth year. Students in matriculation years 2020–2022 are currently preparing to apply or have not yet initiated the application process. The MD-PhD Program’s explicit goal is that all eligible students submit F30/F31 fellowship applications. This goal will be achieved through the regular grant-writing workshops coinciding with NIH submission deadlines that are open to all students, through active encouragement and regular reminders about deadlines and requirements, through one-on-one advising, as well as the introduction of a required 4-week Proposal Development bootcamp in the summer before the beginning of a student’s fourth year in the program.

When to apply for NRSA fellowships. In FY2014 (PA-14-150), the NIH restricted the eligibility to apply for F30 funding to applicants who had “matriculated no more than 48 months before the due date of the initial application.” The Notice of Intent to Publish the Reissuance of the Ruth L. Kirschstein NRSA for Individual Predoctoral MD/PhD and Other Dual Doctoral Degree Fellows (Parent F30) Funding Opportunity Announcement (NOT-OD-14-056) justifies the 48-month eligibility period as an effort to encourage “applications from students early in the research training phase of their dual-degree training so that they can substantively benefit from the mentored research training opportunities of an individual fellowship award.” The 48-month eligibility cut-off does not apply to F31 or F31-Diversity applications. Figure 3 describes the distribution of the number of months after matriculation that F30/F31 applications were first submitted, showing that 175 F30 and F31 applications between 2003 and 2022 were submitted an average of 43 months after matriculation. There is no significant difference between when the successful and unsuccessful applications were submitted (Figure 3B; Kolmogarov-Smirnov test, P = 0.35). Submissions after 48 months include F30 applications submitted before FY2014 as well as F31 and F31-Diversity applications, which have no limit to submission eligibility.

Yale MD-PhD students matriculating 2003–2023 submit F30 applications within the NIH 48-month eligibility window. (A) Original submissions submitted after 48 months after matriculation include F31 applications, and F30 applications submitted before 2014, with an average of 43 ± 10 months after matriculation. (B) Funded applications were submitted 42 ± 11 months after matriculation; unfunded applications were submitted 46 ± 9 months after matriculation (Kolmogorov-Smirnov test, P = 0.32).

Submissions and awards by sex and underrepresented minority group. Between FY2003 and FY2023, 196 new F30/F31 applications were submitted by 175 unique individuals (some individuals submitted more than one original application) from the Yale MD-PhD Program. Of the 196 new applications, 78 (40%) have been awarded, while ten (5%) are currently pending. In contrast, only 69 applications by 53 unique individuals were submitted to foundations and private charities; of these, 15 (22%) have been awarded.

Figure 4 details the status of F30/F31 applications and awards by all eligible trainees in five-year cohorts (by year of matriculation into the MD-PhD Program), stratified by sex and underrepresented in medicine (URM; NIH-defined ethnic and racial categories) status. The last group (years 2018–2022) includes eligible students who have not yet prepared F30/F31 applications or whose applications were recently submitted and are pending (Figure 4). The proportion of students who submitted applications as well as the proportion of applications funded increased in all cohorts over the 15-year period between 2003 and 2017. Two-way ANOVA analyses comparing the proportions of NRSA-funded males versus females and the proportions of funded URM to non-URM trainees indicate no significant difference based on these demographic stratifications over that 15-year period. Table 1 summarizes the data depicted in Figure 4.

The proportion of students submitting and awarded F30 and F31 applications by sex and underrepresented minority groups. Since 2003, an increasing proportion of Yale MD-PhD students submit F30 and F31 applications; the last 5-year cohort of students who matriculated between 2018 and 2022 are just beginning to prepare and submit NRSA applications. The proportion of funded NRSA (green section of each bar) applications, NRSA applications that were submitted but not funded (black section of each bar), and students who did not submit an NRSA application (top section in light colors) are shown for all, male, female, and ethnic/racial underrepresented minority (E/R URM) students in each 5-year cohort. The patterned section for all the 2018–2022 cohorts include students who have not yet submitted or have submitted applications that are pending.

NRSA F30/F31 submissions and awards by sex and underrepresented minority groups

A hidden curriculum is a barrier to success wherein proven strategies for success are taught informally and are not equally accessible to all students (10). The playing field is leveled by requiring all eligible students to apply and providing the necessary resources and critical feedback. In particular, the opportunity to practice skills related to persistence and resilience to feedback and criticism encountered in the application writing and resubmission process is, in the authors’ opinion, invaluable to the future success of students pursuing a career in academic research. Additionally, professionalization and streamlined administration to support students’ fellowship applications are especially important when advancing diversity and inclusivity in research and training capacity.

Effect of predoctoral NRSA awards on physician-scientist training outcomes. For the 288 graduates (through the end of FY2023) who entered the program since 1990, there was no statistically significant effect of having had an F30 or F31 award on the median time to degree, which is 7.8 ± 1.1 years and 7.9 ± 1.2 years with and without an NRSA, respectively (Figure 5A). Six graduates who pursued postdoctoral fellowships rather than residency training did not have NRSA funding. Twenty-six percent (n = 76 of 288) were awarded F30 or F31 as MD-PhD trainees.

Short-term outcomes. (A) The total time to degree is the same for students who held an F30 or F31 during their MD-PhD training and those who did not. (B and C) Residency disciples of students with NRSA funding (B) and without NRSA funding (C) reveal slight differences, with funded students preferring highly competitive, procedural specialties such as anesthesiology, dermatology, ophthalmology, and surgery.

Of the above 288 graduates, 35% (102 of 288) remain in training as residents or fellows. Of these, 51% (52 of 102) held F30 or F31 awards as program trainees and an additional 8% held other non-NIH predoctoral fellowships. In contrast, only 9% (25 of 186) of all post-GME (all sectors) alumni held F30/ F31 awards; this finding is consistent with the data shown in Figure 1D, indicating that the increases in the number of F30/F31 applications submitted by trainees in the program is relatively recent (see the steep slope that begins in 2012). A comparison of residency choices made by the 102 alumni who remain in training stratified by those who held an NRSA award (Figure 5B) versus those who did not (Figure 5C) revealed that, while individuals within both groups were equally likely to pursue training in internal medicine, neurology, and pediatrics, which are disciplines traditionally chosen by physician-scientists (2), NRSA recipients were more likely to pursue residencies in anesthesiology, dermatology, ophthalmology, and surgery. Moreover, family medicine and pathology residencies were not pursued by NRSA recipients. It is not clear how receipt of an NRSA award affected students’ choices of residency and career interests, and these data do not suggest that NRSA receipt confers success in matching to one residency versus another.

Award of mentored career development (K-type) awards and RPGs are often used to evaluate the long-term outcomes of physician-scientist training (11). Fifty-five of the 288 alumni (19%) of the Yale MD-PhD Program have or have had K awards; of these, only 5 (7%) had been awarded F30/F31 fellowships. Fifty-three of the 288 alumni (18%) have or have had RPGs. Independent of K award success, only 1 alumnus with an RPG had an F30/F31 award. However, 34 alumni with K awards also had subsequent RPG funding. The numbers of alumni who have had both F30/F31 awards and K awards or RPGs are currently not sufficient to determine whether there is a correlation between F30/F31, K, and/or RPG funding success. As has been reported by several groups (11–13), we did find a positive correlation between K award success and subsequent RPG support using Spearman’s correlation (r[109] = 0.431, P < 0.0001; data not shown) among our alumni. As expected, and shown in Table 2, most of the K award and RPG recipients are currently employed in the academic sector.

Alumni matriculated 1990–2019 in post-GME sectors with F30/F31, K-awards, R-equivalent funding, and both K-award and R-equivalent (“RPG”) funding