Developing sustainable national infrastructure supporting high-impact research to improve the care of older adults with cancer: A Delphi investigation of geriatric oncology experts

1. Introduction

Due to the aging and cancer demographics of the US population, there is an increasing need for clinicians able to care for the growing population of older adults with cancer and a pressing need for research to inform the care for this population [1,2]. In response to these needs, the Cancer and Aging Research Group (CARG) was developed (www.mycarg.org) [2]. CARG is a coalition of investigators dedicated to: 1) linking existing interdisciplinary aging and cancer researchers together; 2) supporting the development of high quality aging and cancer research; and 3) fostering the careers of new researchers in aging and cancer. However, the field of geriatric oncology remains relatively young, evolving, and loosely organized; the development of a sustainable infrastructure is needed to help organize, support and accelerate research efforts in the field. In response to this, NIH funding was obtained to support the development of sustainable infrastructure to support the growing community of aging and cancer researchers in the United States (Geriatric Oncology Research Infrastructure to Improve Clinical Care, R21/R33 AG059206; MPIs Dale, Hurria, Mohile). The overall goal of this award is to establish a sustainable national research infrastructure to facilitate and support interdisciplinary collaborative research in aging and cancer. During the first year of funding, we evaluated, prioritized, and organized efforts and resources for research infrastructure development by conducting a formal needs assessment of current barriers in conducting geriatric oncology research. We utilized a modified-Delphi process with CARG investigators to assess the needs for a national infrastructure and the results are reported herein.

2. Delphi Procedures

The Delphi method is an iterative survey process for reaching consensus from the individual opinions of experts when the solution to a problem is unknown [3,4]. We conducted a two-phase modified-Delphi process to generate, refine and quantify the opinions of aging and cancer experts regarding the barriers to conducting geriatric oncology research and unmet needs for geriatric oncology researchers to be resolved by the R21/R33 infrastructure grant. Phase 1 was conducted with a broad panel of investigators in geriatric oncology, with the goal of understanding all barriers and facilitators to conducting geriatric oncology research. Phase 2 was conducted with a smaller panel of those experts directly involved with the development of the R21/R33 infrastructure grant and thus, with a better understanding of the capabilities and scope of the funding mechanism to address potential identified needs. Data from Phase 1 were collected and stored in Research Electronic Data Capture (REDCap), a software toolset for electronic collection and management of research data [5]. Data from Phase 2 were collected using Poll Everywhere, a web-based interface for live interactive audience participation. The University of Rochester provided IRB exempt status for the study.

2.1. Expert Panel Selection and Recruitment

Phase One (P1): Eligible participants for the first Delphi phase included researchers, oncologists, geriatricians, trainees, nurses, social workers, specialists in rehabilitation medicine, basic scientists, biostatisticians, and patient/caregiver advocates who expressed interest in being a member of CARG and had attended at least one of the bimonthly CARG calls. Participants had to have conducted – or intend to conduct in the future – research in geriatric oncology. P1 consisted of a single round of a web-based questionnaire where experts were contacted via email, with a link to a survey with three parts: a consent to participate, a demographics survey with screening questions regarding eligibility, and the first Delphi round questionnaire. Participants who met eligibility criteria were able to complete the first Delphi round questionnaire.

Phase Two (P2): The panel of experts for the second Delphi phase consisted of a convenience sample of members of the CARG Oversight Board and a committee of Organizational Liaisons who attended the 1st R21/R33 CARG Infrastructure Grant Conference (October 4–5, 2018 at City of Hope). Members of the Oversight Board were selected based on specific expertise and their demonstrated commitment to aging and cancer. Members of the Organizational Liaisons Committee included leaders from national organizations interested in aging and cancer research such as professional organizations, funders and clinical trial groups. Participants for P2 were made aware that the survey would be conducted during the CARG R21/R33 meeting using Poll Everywhere software prior to the meeting. Individuals were provided iPads during the conference meeting with the option of participating.

2.2. Delphi Procedures

The two phases of the modified Delphi process were conducted 2–4 weeks apart. P1 started on 9/10/2018 and participants were asked to complete the survey within 9 days. P2 was conducted on 10/4/2018 during an in-person meeting of the CARG Infrastructure Grant Conference. While all responses were anonymous in the sense that they could not be attached to any given participant, participants may know the group composition due to participation in calls together and attendance at a common conference. Data analysts were blinded to the identity of participants. A flow chart illustrating the Delphi process is shown in Fig. 1.

Fig. 1.

Flow chart of Delphi process to reach consensus.

In P1, participants answered both closed and open-ended questions related to identifying: 1) current barriers to conducting geriatric oncology research, 2) unmet needs for geriatric oncology researchers, 3) needed resources, 4) expertise not already available at their home institution or outside the institution, and 5) infrastructure needs that could be provided through a training grant. Following completion of P1, group responses were reviewed by investigators and themes were used to develop the questionnaire for P2. Group responses from P1 were presented to the P2 panelists, and they were asked to rank priorities in perceived barriers for the R21/R33 grant infrastructure to resolve. For each question, two successive rounds were conducted. Using Poll Everywhere's interactive interface, the question was posed, live responses were collected in real-time, and aggregated responses were instantly displayed graphically to the full P2 panel. The panel had the opportunity for open discussion of the aggregated responses before the question was posed a second time to obtain consensus. Data for all partial/complete responses were analyzed using MAXQDA 2018.1 for open-ended questions and Excel 2016 for close-ended questions.

3. Delphi Results

Of 275 invited, 67 participants completed P1; three individuals completed demographics only (did not meet eligibility criteria to proceed to P1 questionnaire). Participants in P1 were primarily White (69%), non-Hispanic (94%) and female (60%). Most were Assistant Professors (31%), Professors (25%), or Associate Professors (15%). Most had co-authored or authored a manuscript pertinent to the care of older adults with cancer or their caregivers (90%). More than half had been funded to do research in geriatric oncology (69%) (Table 1). The demographic characteristics of the individuals who completed demographics but did not meet eligibility criteria for P1 were very similar to the demographics characteristics of the expert panel who participated in P1.

Table 1.

Demographic characteristics of Delphi Phase 1 participants (N = 67).

N (%)
Age, years (mean, range)	46.4, (27–84)
Race
American Indian or Alaska Native	1 (1%)
Asian	13 (19%)
Black or African American	5 (7%)
White	46 (69%)
Unknown/not stated	2 (3%)
Ethnicity
Latino Hispanic	3 (4%)
Latino unknown	1 (1%)
Non-Hispanic	63 (94%)
Gender
Female	40 (60%)
Clinical title
Trainee (Fellow/PhD Student)	7 (10%)
Instructor	2 (3%)
Assistant Professor	21 (31%)
Associate Professor	10 (15%)
Professor	17 (25%)
Stakeholder	3 (4%)
Other clinician-researcher	2 (3%)
Other	5 (7%)
(Co)authored a manuscript pertinent to the care of older adults with cancer or their caregivers	60 (90%)
Ever funded to do research in geriatric oncology	46 (69%)

3.1. Phase One (P1)

The most prevalently cited barriers to conducting geriatric oncology research were the need for funding and protected time for research. Other frequently identified barriers included lack of access to data/lack of infrastructure for data sharing, difficulties with identifying and developing collaborations, challenges identifying the appropriate measures or tools to use in research of older adults with cancer, difficulty recruiting study participants, lack of mentorship and training, and lack of support/buy-in from their home institution and the wider research community (Fig. 2, Table 2).

Fig. 2.

Barriers to conducting geriatric oncology research.

Table 2.

Barriers to conducting research in geriatric oncology and unmet needs of geriatric oncology researchers.

Themes	Sample quotes
Need for funding	“Lack of geri-onc specific funding and career development award”
Protected time/balancing clinical load with time for research	“Isolating dedicated time” “…a key issue in the field is that it is largely MD/practitioner driven at the moment and therefore highly governed by available so--called ‘protected time’for research…”
Lack of access to data	“Lack of access to large datasets with geriatric measures”
Collaboration	“I am a basic, cancer and aging researcher. Thus the main challenges I encounter are finding the right way to interact with geriatric oncologists. What are the right questions to ask that will maximally take advantage of our basic science methods/resources and the geriatric oncologist's methods/resources?”
Identifying appropriate measures/tools	“Comparability due to the proliferation of Geriatric Assessments and post-treatment measures represents a challenge” “Appropriate PRO's validated for older adults with cancer” “We need infrastructure to facilitate data capture and endpoints most relevant to older adults”
Participant recruitment	“Enrolling true ‘older’ adults outside of the range of 65–75 years” “Recruiting patients and caregivers into research studies fast enough to meet recruitment targets.”
Mentorship and training	“coordinating cross disciplinary mentorship” “Locally as a junior faculty [in] training and in-house collaboration/mentorship is difficult.” “Mentorship in writing IRB protocols and grant writing.” “Training in qualitative research/mixed methods research” “I need help getting trained in GA and in training staff to conduct GA as part of clinical trials”
Institutional support/buy-in from wider community	“Geriatric oncology isn't as ‘sexy’ as genomic research”
Statistical support	“Need statisticians who are familiar with related measures.” “Expertise outside of the geriatric oncology ‘circle’ is difficult to obtain (e.g., data science, informatics)”
Access to administrative and research staff	“Geriatric oncology specific research team would enable enhanced activity. Currently our clinical trials unit performs other studies but geriatric oncology trials have traditionally required different skill sets e.g. knowledge of assessments etc.”
Involvement of patient advocates in the research process	“Have never had access to a patient or caregiver to discuss research methods with.” “We collected helpful information but recruitment was difficult.”

Specific unmet needs identified by researchers included mentorship, funding, statistical support, collaboration, time management/protected time (e.g. “isolating dedicated time”), access to administrative/research staff and access to training in specific areas related to research interests in aging. Within mentorship, a particular emphasis was placed on assistance with brainstorming, writing grants and protocol development (Table 2). In addition, participants expressed content-area-specific mentorship and training in both geriatric assessment and oncology, plus methodological mentorship including study design and training in specific research methodologies (e.g. qualitative/mixed methods).

Self-identified early-career researchers identified distinct needs specific to their career stage including feedback on scientific writing/grant writing, examples of successful proposals, opportunities for junior faculty in leadership roles, and access to pilot data. Later career-stage investigators identified specific needs around protected effort for mentors, administrative support for training and mentorship, and a need for training such as “train the trainer” courses or education related to burnout.

3.2. Phase Two (P2)

Results from P2 further supported the themes that emerged in P1. In P2, priorities for the six Cores of CARG Geriatric Oncology Research Infrastructure were ranked to understand how best to address the needs and barriers that emerged from P1 (Table 3). Recognizing finite resources of the R21/R33 grant that require prioritization, the panel of experts in P2 ranked funding support for special projects as the most important barrier to resolve. The prioritized unmet needs identified in the Delphi process were then used in the CARG Infrastructure development process for the grant. Table 3 details components of CARG infrastructure development and plans for addressing the identified specific needs areas.

Table 3.

Research infrastructure proposed to meet needs and overcome barriers.

Cores and other aspects of research infrastructure	Needs and barriers addressed
Leadership, Mentorship, and Training	Mentorship and training Collaboration Protected time/balancing clinical load with time for research Bi-weekly CARG conference calls; monthly Junior Faculty CARG conference calls; monthly CARG data-meeting calls Junior Faculty Advisory Board Pilot funding awards
Clinical and Biological Measures of Aging	Identifying appropriate measures/tools
Behavioral, Psychological, & Supportive Care Intervention	Mentorship and training Participant recruitment SCOREboard: Patient and caregiver advisory board
Care Delivery & Comparative Effectiveness Research	Mentorship and training Participant recruitment
Epidemiology, Biostatistics, & Informatics	Statistical support Research design
Dissemination and Communication	Institutional support/buy-in from wider community Lack of access to (pilot) data

4. Discussion and Conclusions

We report the results of a needs assessment and consensus-building Delphi process among experts to better understand the relevant barriers to research and needs of researchers in geriatric oncology at all career stages. The most frequently reported barriers to conducting geriatric oncology research were the need for additional funding and more protected time. Other barriers and unmet needs of researchers include mentorship, identification of appropriate measures and tools for use with older patients with cancer, content area specific mentorship and training (e.g. geriatric assessment, oncology, study design, and research methodologies), and statistics training/support. Junior faculty participant cited greater training-related themes, such as grantwriting and mentorship, whereas senior faculty needs were more related to protected time for mentoring and developing mentoring skills. The identified areas of need were used to guide the development of specific areas of support within the context of an R21/R33 funded infrastructure grant to support the national growth of the field of geriatric oncology.

Many of the barriers to conducting research and needs of researchers identified through the current Delphi process are consistent with those reported in other specialties. Challenges related to obtaining funding and balancing clinical time with time for research are well-described among clinician-scientists [6]. A focus group study examining mentee perspectives on early career mentoring among translational investigators with active mentored career development awards (e.g. K08, K23) cited network issues, such as finding appropriate mentors in the field and coordinating mentorship from a team when research interests bridged multiple fields, as a significant challenge [7]. A study of barriers and facilitators to mentoring early-career investigators and trainees in rheumatology cited the perception that having a network of internal and external collaborators was necessary for success, but that it is challenging to find mentors at other institutions and even within home institutions [8]. These challenges are also faced by researchers in geriatric oncology, who often need mentorship and training related to both geriatrics and oncology, and a need to bridge between two distinct, separate fields with a limited group of mentors from which to draw.

An additional challenge identified for researchers in geriatric oncology is the lack of mentorship specific to the field at their home institution, often due to the small number of senior people in the field. Existing programs aimed at supporting researchers from various medical specialties with research interests in aging have utilized a mix of traditional dyadic mentoring in addition to peer mentoring. For example, the CoMPAdRE program at Columbia University Medical Center was structured around periodic retreats for junior faculty cohorts to engage in peer-to-peer discussions as well as one-on-one meetings with mentors [9]. The mentoring infrastructure for the Grants for Early Medical/Surgical Specialists' Transition to Aging Research (GEMSSTAR) scholars program from NIA comprised of conferences that allowed peers to network with each other, participate in sessions led by mentors, and meet individually with mentors [10]. As noted by High and Kritchevsky, a critical aspect of the success of these programs was that “peers” were determined by a shared interest in aging research: “Demonstrating efficacy, determining the appropriate timing of anti-aging interventions, and assessing multiple disease outcomes will require a skilled workforce in aging within each specialty.” [11] Programs that bring together researchers with a common interest in aging are an important step, but there remains a need for infrastructure to connect and support researchers specifically interested in geriatric oncology research. The CARG Infrastructure development (i.e., CARinG) has a strong focus on mentorship, developing mentoring skills for senior faculty, peer mentoring, and supporting junior investigators in the identification of mentorship at their home institution. The CARG Geriatric Oncology Research Infrastructure to Improve Clinical Care initiative will address the specific unmet needs of geriatric oncology researchers through six Cores, each of which will require leadership: (1) Leadership, Mentorship, and Training, (2) Clinical and Biological Measures of Aging, (3) Behavioral, Psychological, & Supportive Care Intervention, (4) Care Delivery & Comparative Effectiveness Research, (5) Epidemiology, Biostatistics, & Informatics, and (6) Dissemination and Communication. In addition, two Advisory Boards (Junior Faculty Advisory Board and an older adult patient advisory board [SCOREboard [12]]) will provide opportunities for mentorship/training and input from patient and caregiver stakeholders.

In conclusion, this needs assessment and consensus process using a modified-Delphi process identified barriers to research in geriatric oncology and specific unmet needs of researchers in aging and cancer. The following specific priorities for research infrastructure were identified: providing opportunities for mentorship/training, identifying relevant tools and measures for use with older patients with cancer, training for specific research methodologies, and biostatistics training and support. These priorities will shape the development of the CARG Geriatric Oncology Research Infrastructure in the form of Cores to coordinate and advance a national agenda in aging and cancer research.