Shared mission, vision, and goals

Shared goals are a critical component of interdisciplinary teamwork, and gaining team consensus on shared goals is one of the first objectives that should be accomplished [Reference Young8,Reference Choi and Pak9]. While high-level research goals are often predetermined in the funding announcement, developing a mission, vision, and specific goals with the study team from the start creates a solid foundation to assess deliverables and scope over the course of a project. To ensure team members had awareness of and agreement with the PROSPR-Lung mission, vision, and goals, a virtual “Project Launch” meeting was held for all study team members at each site (e.g., Investigators, Project Managers, Programmers, Biostatisticians, Research Assistants, and more). The launch meeting provided a space for team members to discuss, critique, and refine the mission, vision, and goals, which were further solidified at PROSPR-Lung’s initial in-person meeting. Subsequently, PROSPR-Lung leadership took steps to guarantee the mission, vision, and goals remain a focal point in meetings, communications, and project deliverables. For example, when an individual requests the use of PROSPR-Lung data for an abstract, manuscript, or new funding application proposal, the requestor must specify how the research contributes to PROSPR-Lung’s mission.

Early in year 2, PROSPR-Lung created a logo, website [10], and Twitter account [11] to further unite around a team identity and to provide cohesion for the team [Reference Rolland, Smith and Potter12]. To promote consistency in dissemination efforts, PROSPR-Lung created branded posters and presentation templates and made them available, for optional use, to all team members on the consortium’s web-based document-sharing platform.

Keeping PROSPR-Lung’s shared mission, vision, and goals at the center of our work provides the study team with guidance and insight into how their work contributes to study results and to the scientific community. It has allowed project leadership to maintain a delicate but critical balance between achieving individual and research team goals and ensuring that the main results advance together.

Leadership

Strong leadership is noted throughout SciTS literature [Reference Rolland, Burnside, Voils, Shah and Brasier3,Reference Bennett and Gadlin13–Reference Aarons, Reeder, Miller and Stadnick15] as a necessity for building and maintaining a successful, productive research team. Study leadership is determined during the application development phase when roles and structures are strategically designed to set the study up for success. PROSPR-Lung designed a leadership structure (Appendix 1) that includes study team members from across sites and roles consisting of a small governing body (Executive Committee) responsible for overall PROSPR-Lung guidance and oversight, a larger governing body (Steering Committee) responsible for making strategic decisions about the overall scientific direction of PROSPR-Lung, and three operational cores (Administrative Core, Data Acquisition Unit [DAU], and Dissemination and Engagement Core). The research center’s governing bodies are composed of representatives from each site, providing diversity in knowledge, specialization, and experience. Specific research project leadership within PROSPR-Lung is divided among Site Investigators in alignment with their expertise. Throughout PROSPR-Lung, successful collaboration is promoted through strengths-based division of labor, with responsibilities and accountability assumed not simply by role but also by expertise, interest, passion, and career goals.

In addition to shared leadership and accountability, Rolland et al. describe high-functioning teams as those with written authorship and operational policies [Reference Rolland, Burnside, Voils, Shah and Brasier3]. Explicit, written governance, and authorship guidelines were among the early priorities of the PROSPR-Lung Administrative Core and are adhered to at all levels of leadership within the study team. The leadership structure also fostered responsible, compliant, transparent data use and analyses through our Proposal Access Review Committee (PARC) (Fig. 2). The PARC, detailed below, solicits lead author presentations to the Steering Committee and acts as the hub for review, tracking, and dissemination of PROSPR-Lung intellectual products.

Figure 2.

PROSPR-Lung’s Proposal Access Review Committee (PARC) intellectual product proposal review, administrative core compliance review, data request, and tracking process. DUA = Data Use Agreement, IRB = Institutional Review Board, PARC = Proposal Access Review Committee.

PROSPR-Lung’s leadership strategy placed inclusion, strengths-based leadership, and transparency at its core. In addition to fostering productivity, these PROSPR-Lung leadership elements help to establish and maintain a culture of trust that is essential for effective teams.

Research support systems

A research support system is the infrastructure in which the project and its staff, data, communications, and science can be coordinated, managed, and disseminated. Rolland et al. highlight research support systems as the singular best practice that “strongly influence[s]” all other best practices [Reference Rolland, Burnside, Voils, Shah and Brasier3]. As with the study leadership plan, the formulation of PROSPR-Lung’s research support system began at the proposal development phase and was reviewed and reworked to ensure efficiency and necessity throughout the award period. PROSPR-Lung’s key research support system elements and implementation prioritization strategy is detailed below, summarized in Fig. 3, and presented chronologically in Appendix 2.

Figure 3.

Project management tools utilized by PROSPR-Lung to drive collaboration, efficiency, and scientific progress. Refer to Fig. 1 for specific SciTS best practices for which these tools were used to implement. CoT = culture of trust, DM = data management, ICs = interdisciplinary conversations, L = leadership, MVGs = mission, vision, and goals, RSS = research support systems.

Data management

The efficient data management system implemented and improved throughout the study complements the key elements of PROSPR-Lung’s research support system. A strong research data management system requires a distinct project management and leadership structure, a clear and thorough Data Use Agreement, programming code transparency, detailed workplans, documentation of variable definitions and other important details within a data dictionary, consistent file nomenclature, and excellent communication.

The PROSPR-Lung DAU is responsible for the creation and maintenance of an accessible and transparent data management system for all team members. It aims to identify data elements needed for PROSPR-Lung analyses, develop and refine a data repository/common data model (CDM) with data from all sites, and provide a limited subset of PROSPR-Lung data for external use [26]. In line with best practices, the PROSPR-Lung team identifies the questions the data are expected to answer and the data needed to answer the questions, assesses availability of the data, develops common data elements, maps and transforms individual data points to common data elements, and performs quality checks (Appendix 3) [Reference Rolland, Reid and Stelling27]. Further, PROSPR-Lung’s data management system includes an analytic dataset request process with built-in compliance checks (Fig. 2). In the first 2 years of the project, approximately 4.0 FTE was directed to the DAU’s Project Manager, Data Specialists/Programmers, and Biostatisticians at the lead site as infrastructure was developed and implemented.

Data extraction, programmer codes, and workplans

Keeping its mission, vision, and goals at the center of decision-making, PROSPR-Lung executes the iterative process of determining which variables need to be captured and made available to fulfill the project’s goals. At the same time, the DAU leads collaboration with study sites to create and validate the data collection and database creation process (Fig. 4). Focusing on flexibility, the DAU built on the existing infrastructure of HCSRN’s Virtual Data Warehouse (VDW) [Reference Ross, Ng and Brown19,Reference Hornbrook, Hart and Ellis28] to create a relational database with standardized data elements. This allows PROSPR-Lung to meet project aims while also setting up its database to readily answer future research questions.

Figure 4.

PROSPR-Lung’s methodology to capture, harmonize, and provide limited and deidentified datasets. EHR = electronic health record, HFH = Henry Ford Health, KPHI = Kaiser Permanente Hawaii, KPCO = Kaiser Permanente Colorado, MCHS = Marshfield Clinical Health System, NCI = National Cancer Institute, UPHS = University of Pennsylvania Health System.

To extract the data elements, the DAU develops programming code centrally, tests the code at one subsite, and develops workplans with programming code. These workplans are distributed to all subsites and recorded in SmartSheet®. On bimonthly conference calls, the DAU facilitates communication across site programmers to address challenges associated with current and future workplans, update site progress, identify issues, and work toward resolution. Code version control is maintained using a distributed version control system for tracking changes in source code (Git™) [29] and pushed to a remote repository (GitLab™) [Reference Hixon30] to improve transparency and collaboration with sites during code distribution. A rigorous three-round quality assurance process is administered by the DAU before finalizing each CDM (Appendix 3).

Interdisciplinary conversations on approaches, methods, and results

As defined by Choi and Pak, “interdisciplinarity analyzes, synthesizes, and harmonizes links between disciplines into a coordinated and coherent whole [Reference Choi and Pak9].” PROSPR-Lung cultivated an interdisciplinary team that includes expertise in pulmonology, radiology, health economics, epidemiology, biostatistics, statistical modeling, bioinformatics, health equity research, behavioral health, and dissemination and implementation research. Study team members are strategically included in meetings based on specialty, task, and interest. Biostatisticians meet monthly to share methods and obtain feedback on analytic approaches. Leaders to move forward publications, presentations, and other intellectual products are determined based on strengths and interests. Monthly Steering Committee meetings provide opportunities to review current manuscript proposals, analysis plans, and results, with the benefit of multidisciplinary perspectives united in a common goal. Semiannual in-person or virtual meetings provide focused time for study team discussions on science, methods, results, and learnings. Steering Committee and semiannual meetings also provide the opportunity to collaborate with scientists from NCI, other Trans-PROSPR members, and subject matter experts invited as guest speakers.

Prioritizing interdisciplinary discussions facilitates study team engagement and valuation, and results in better research products by establishing accurate definitions and employing data quality processes early in development [Reference Rolland, Burnside, Voils, Shah and Brasier3].

Culture of trust

Building and maintaining trust that spans time and distance requires psychological safety, which includes transparency, visibility, accountability, and inclusivity [Reference Bennett and Gadlin13]. Teams that trust each other feel comfortable questioning each other’s data, share mistakes and misunderstandings, encourage open discussions, meet commitments, and trust that other team members will do the same. A team that has cultivated a culture of psychological safety also encourages critical review of research proposals and analytic conclusions [Reference Rolland, Burnside, Voils, Shah and Brasier3]. To facilitate these imperative team assets, PROSPR-Lung prioritized the development of the PARC early in the project timeline (Appendix 2). The PARC establishes and oversees fair, clear, and efficient policies for all PROSPR-Lung intellectual products. These products are defined as publications, meeting abstracts, presentations, or new funding proposals that involve PROSPR-Lung investigators or staff who use funds and/or data from one or more PROSPR-Lung sites. PROSPR-Lung’s detailed governance document, modeled after the PROSPR I governance document, specifies policies and procedures for the review and approval process. This governance document also serves as a central point of reference for authorship requirements and responsibilities, adapted from the International Committee of Medical Journal Editors’ recommendations [32]. PROSPR-Lung’s guidelines require that at least one person from each data-contributing site be included as a coauthor to ensure accurate interpretation and provide key insight into the idiosyncrasies of their site’s data. Rotating PROSPR-Lung site members are responsible for reviewing all proposals for significance, design, feasibility, and potential overlap with other projects, as well as for making recommendations about all proposals. A Project Manager oversees the PARC process (Fig. 2), manages communications between proposal authors and PARC members, and facilitates meetings to discuss procedural issues and resolve disputes. REDCap® [Reference Harris, Taylor, Thielke, Payne, Gonzalez and Conde31] is used to manage intellectual product submission (Appendix 5), review, approval (Appendix 6), and tracking forms.

Cultivating a culture of trust is a central thread woven throughout the SciTS best practices that PROSPR-Lung applies to its operations and demonstrates in essential ways, including: (1) thoughtfully planning and strategically executing meetings that include salient people and topics; (2) distributing detailed meeting minutes in a timely manner with action items, deadlines, and assigned personnel to ensure accountability; (3) establishing data delivery timelines and workplans with web-based centralized progress tracking in close collaboration between the DAU and the data-contributing sites; (4) uploading final products to (i) the public PROSPR-Lung website [10], (ii) a secure, web-based document management platform [24] provided by HCSRN [33], and (iii) our internal product tracking system in REDCap® [Reference Harris, Taylor, Thielke, Payne, Gonzalez and Conde31]; and (5) formally soliciting site feedback through annual surveys (Appendix 7 and 8), presenting results to the Steering Committee, and incorporating feedback to improve study operations. Propelled by suggestions in annual survey responses, PROSPR-Lung implemented role-specific meetings to ensure all voices were heard, improved workplan clarity and consistency of code repository use, incorporated site programmer feedback of draft workplans prior to release, and improved action item communications by including action item tables in emails as well as in meeting minutes. These strategies to establish a culture of trust helped PROSPR-Lung build a consortium that values transparency and encourages sharing new ideas.