Doing More With Less: A Centralized Model for Fellowship Program Coordination Open Access

Pediatrics subspecialty fellowship programs are typically small, vary in their approach to training, and have unique administrative structures and cultures. The Accreditation Council for Graduate Medical Education (ACGME) recommendations for fellowship administrative support vary by specialty, but none requires a full-time coordinator.^1,2  Thus, fellowship coordination is usually assigned to an administrator who has various other duties. Coordinators function in silos and often focus on the unique features of individual programs, rather than capitalizing on commonalities, such as common requirements from the ACGME and the American Board of Pediatrics. These operational silos lead to redundancy and inefficiency. Centralizing program coordinators in “communities of practice” can have positive effects, as shown in a model that brought together several coordinators from similar disciplines.³ 

To date, no studies have assessed a centralized model of support for small fellowship programs with a coordinator serving multiple programs and with dedicated, specialized administrators. The goal of this study was to examine the feasibility, acceptability, and impact of a centralized administrative model for pediatrics fellowship support at a large university-based institution.

The University of California, San Francisco (UCSF), Department of Pediatrics has 18 subspecialty divisions with 19 fellowship programs. Prior to 2014, each division had complete autonomy over fellowship administration, with part-time coordinators who had a variety of other administrative duties.

In 2014, the Department of Pediatrics restructured fellowship administrative support in response to complaints about redundancy, turnover, and inadequate communication. To design the model, a small group conducted a task analysis of coordinator support, and developed the centralized model. The model was reviewed and approved by fellowship program directors and the institution's Office of Graduate Medical Education prior to implementation. As the focus of this study is an administrative program, no Institutional Review Board approval was required.

In the new model, all coordinator tasks were reassigned to a team of 6 full-time staff members, including a data coordinator, a human resources coordinator, and 4 program coordinators. One program coordinator was assigned a dual role as team manager. The figure illustrates the team's organizational structure, and the table summarizes roles and responsibilities for each team member. Each coordinator was assigned a program cluster based on the number of fellows, similarities in program requirements, and recruitment cycles.

The fellowship administrative team currently supports 19 pediatrics fellowship programs, with a total of 85 fellows. One of the program directors functions as the team's faculty advisor. The team meets twice a month to discuss processes, concerns, and ideas; quarterly meetings with program coordinators and directors serve to discuss general issues and foster further collaboration. Additionally, coordinators meet regularly with individual program directors to ensure that program-specific goals are met.

Eighteen months after implementation of the new model, we surveyed all program directors to determine their perceptions of the model, and reviewed the financial impact as well as the effect on staff turnover. We estimated the average annual cost savings based on the change in full-time equivalents (FTEs) between the traditional and new model.

A total of 15 of 19 (79%) program directors completed the survey. Among respondents, the majority preferred the new model and rated it as better or much better in efficiency, effectiveness, productivity, knowledge, and helpfulness. Total FTEs allocated to program administration decreased from 10 FTEs in the traditional model to 6 FTEs in the new model, with a reduction of approximately $250,000 in administrative costs annually. In addition, we noted a decrease in staff turnover. Between 2010 and 2013, 5 coordinators left their positions versus none between 2014 and early 2017.

The centralized model for fellowship administrative support was received with enthusiasm by faculty, fellows, staff, and departmental and institutional leadership. Not only did the model lead to high satisfaction among fellowship program directors, but it also resulted in significant cost savings and zero staff turnover. While some program directors had expressed apprehension when first presented with the proposal, comments on the survey indicated how much they appreciated the change. One program director commented that she was in “program director heaven,” and another stated, “There is great improvement from prior [years], and the change has been amazing.” The fellowship administrative team received a UCSF Great Team Award in 2016 and an invitation to present at the 2016 Macy Regional Conference on Innovations in GME.

Team members report that they can use the expertise of other team members to improve strategies and processes, which helps them manage fellowship programs more effectively. The approach creates team knowledge, which allows flexible distribution of tasks if a program temporarily has a higher workload. The sense of team identity and better access to resources is similar to the findings by Bing-You and Varaklis,³  who created a “community of practice” model for residency program coordinators. Our model went a step further by assigning several programs to a single coordinator and centralizing certain tasks, such as data management and human resources.

Limitations of our intervention include the single institution, single specialty nature, which may limit generalizability. We also did not conduct formal evaluations of fellows or staff. Of the 19 fellowship program directors, 4 did not complete the survey, and it is possible that nonrespondents were less enthusiastic about the new model. Finally, cost savings with the new model are estimated, since in the traditional model, the amount of time allocated to fellowship coordination was hard to calculate as many program coordinators had several other tasks.

We believe this model can be adapted in other institutions to centralize administrative support for multiple small fellowship programs. We strongly recommend conducting a careful task analysis and soliciting input from affected stakeholders before finalizing a central administrative support structure. At our institution, we continue to seek input from program directors, coordinators, and fellows to optimize the functioning of the team and make adjustments to accommodate changes in program size and requirements.

The restructuring of fellowship administration from individual divisional silos to a unified departmental team was feasible, and produced numerous positive results, including program director satisfaction and reduced costs and staff turnover.