Abstract
This scoping review provides key trends in the use of the Delphi method applied to geriatrics and gerontology research, and documents key information that can be used in the planning of future Delphi studies, like recruitment rates, number of Delphi rounds and panelists, attrition, consensus definitions, and stakeholders engaged.
Keywords: consensus methods, guideline development, stakeholder engagement
1. Introduction
The Delphi method is a structured process for obtaining expert consensus to develop evidence‐based resources (e.g., quality indicators, clinical guidelines) [1]. Despite an estimated 328% increase in the number of Delphi studies since 2014 per PubMed, no reviews have explored its use in aging research. In this scoping review, we characterize applications of the Delphi method in aging research, identifying methodological gaps in its application and providing recommendations for future Delphi studies focused on older adults.
2. Methods
This study followed the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses extension for Scoping Reviews (PRISMA‐ScR) [2] and was registered with Open Science Framework (https://osf.io/wfqh6).
A medical librarian (MRD) performed comprehensive searches to identify relevant articles. Searches were run in December 2022 (updated May 2025) using five databases. The search included all appropriate controlled vocabulary and keywords for the concepts of “Delphi technique” and “older adults” for articles published from 2018 to date (Table S1).
We used Covidence systematic review software for screening and extraction. First, two independent reviewers screened titles and abstracts using predefined inclusion and exclusion criteria (Text S1). We used consensus to resolve coder discrepancies. Next, we conducted a full‐text review of articles that met inclusion at the title and abstract screening phase to confirm eligibility in the final sample.
Following two rounds of extraction training with adjudication, our team performed independent data extraction using predefined, standardized templates with extraction domains piloted and refined prior to study inception (Text S2). When extraction ambiguities were encountered, a second reviewer resolved and finalized extraction.
3. Results
Figure 1 presents the PRISMA‐ScR flow diagram for 102 articles (References in Supporting Information) comprising the final analytic sample. Table 1 presents selected characteristics; Table S2 presents characteristics for all extraction domains, and Table S3 presents the completed PRISMA‐ScR Checklist.
FIGURE 1.
PRISMA‐ScR flow diagram.
TABLE 1.
Selected characteristics of aging‐focused Delphi studies (N = 102).
| Characteristics | N a | % |
|---|---|---|
| Focal age group b | ||
| Middle‐aged (50–64) | 1 | 1% |
| Youngest‐old (65–74) | 4 | 4% |
| Middle‐old (75–84) | 3 | 3% |
| Oldest‐old (85+) | 3 | 3% |
| All Older Adults (Focal Group NS) | 98 | 96% |
| Focal sex/gender c | ||
| Men only | 1 | 1% |
| Both men and women (focal group NS) | 101 | 99% |
| Focal racial/ethnic groups c | ||
| All racial/ethnic groups (focal group NS) | 102 | 100% |
| Focal populations with health disparities c | ||
| Underserved rural populations | 1 | 1% |
| Focal population with health disparities NS | 101 | 99% |
| Focal population setting b | ||
| Home | 6 | 6% |
| Nursing home | 19 | 19% |
| Adult day health/adult day service | 1 | 1% |
| Hospital | 20 | 20% |
| Senior center | 1 | 1% |
| Assisted living | 2 | 2% |
| Other | 16 | 16% |
| All settings (focal setting NS) | 49 | 48% |
| Delphi panelist backgrounds b | ||
| Older adults/patients | 13 | 13% |
| Unpaid caregivers | 8 | 8% |
| Clinicians/care providers | 91 | 89% |
| Researchers | 64 | 63% |
| Policymakers | 10 | 10% |
| Advocacy and education groups | 8 | 8% |
| Other | 20 | 20% |
| Delphi panelist inclusion criteria b | ||
| Authored Publications | 26 | 25% |
| Clinical/scientific expertise | 69 | 68% |
| Employment with study sites or projects | 3 | 3% |
| Existing society or group membership | 25 | 25% |
| Personal experience (patients/caregivers) | 11 | 11% |
| Primary investigators of past/current research efforts | 4 | 4% |
| Previous involvement with similar work | 3 | 3% |
| Professional background | 3 | 3% |
| Professional networks of other experts | 8 | 8% |
| Years of experience | 16 | 16% |
| NS | 4 | 4% |
| Total number of Delphi rounds | ||
| 1 | 5 | 5% |
| 2 | 33 | 32% |
| 3 | 52 | 51% |
| 4 | 8 | 8% |
| 7 | 1 | 1% |
| NS or unclear | 3 | 3% |
| Number of Delphi panelists in round 1 | ||
| 6 to 10 | 5 | 5% |
| 11 to 20 | 16 | 16% |
| 21 to 30 | 20 | 20% |
| 31 to 40 | 11 | 11% |
| 41 to 100 | 13 | 13% |
| More than 100 | 8 | 8% |
| NS or unclear | 29 | 28% |
| Adherence to CREDES guidance for Delphi study conduct | ||
| Yes | 11 | 11% |
| Not reported | 91 | 89% |
| Rationale for using Delphi method b | ||
| Time/convenience | 5 | 5% |
| Blinded aspect | 9 | 9% |
| Lack of established guidelines | 47 | 46% |
| Lack of established tools or measures | 7 | 7% |
| Other | 36 | 35% |
| Rationale not provided | 22 | 22% |
| Level of agreement required for consensus | ||
| < 70% | 9 | 9% |
| 70% | 29 | 28% |
| 75% | 10 | 10% |
| 80% | 17 | 17% |
| 100% | 2 | 2% |
| Varied | 6 | 6% |
| Other | 9 | 9% |
| NS | 20 | 20% |
| Study duration | ||
| < 6 months | 32 | 31% |
| 7–12 months | 16 | 16% |
| > 12 months | 6 | 6% |
| NS or unclear | 48 | 47% |
Full study characteristics along with corresponding citations can be found in supplemental materials.
Articles can be included in multiple categories so summed percentages may exceed 100%.
Predetermined categories with n = 0 are excluded from presentation in this table. NS = Not Stated.
Most materials developed in these Delphi studies were sex and gender agnostic (99%) and without a focal age group (96%). None developed materials that were targeted to specific racial/ethnic groups. One study targeted a population with health disparities—rural veterans. Almost half of the reviewed articles lacked a focal setting for the developed materials (48%).
Clinicians/Care Providers and Researchers were the most commonly engaged Delphi expert panelists (89% and 63%, respectively). Persons with lived experience, including Older Adults/Patients and Unpaid Caregivers, were less represented (13% and 8%, respectively). Over half of the reviewed articles utilized three Delphi rounds (51%), while one‐third utilized two Delphi rounds (32%). The average number of participating Delphi panelists in round one was 45 (ranging from 6 to 242), and the average recruitment rate to round one was 67%. Figure S1 depicts panelist attrition across rounds; rates ranged from 5% to 8% between rounds one and two, 7% to 16% between rounds two and three, and 16% between rounds three and four.
A small portion of articles in this review (11%) reported adhering to CREDES guidance [3] for Delphi study conduct. Level of agreement required for consensus included thresholds of 70% (n = 29), 75% (n = 10), or 80% (n = 17); 20 articles did not specify the level of agreement required for consensus.
4. Discussion
Our analysis is the first to comprehensively document key features of Delphi studies in aging research. We found that 67% of invited panelists engaged in the first round of Delphi review. Over half of our sample utilized three rounds of Delphi review, and among these, the number of participants averaged 51 at round one, 43 at round two, and 36 at round three. Because rates of attrition ranged from 5% to 16% across rounds, we conservatively suggest future investigators plan for no less than 16% attrition between rounds. Regarding the level of agreement required for consensus, most articles (57%) used 70% agreement or greater.
Engagement of older adults and caregivers as Delphi panelists was low. Also rare was the development of materials for populations experiencing health disparities, which include racial/ethnic minority groups, individuals with low socioeconomic status, rural populations, sexual minority groups, and persons with disabilities. Future Delphi studies should better integrate persons with lived experience, including developing culturally tailored materials for populations experiencing health disparities. Martin‐Kerry and colleagues provide recommendations to support meaningful engagement of older adults using the Delphi method [4].
Adherence to CREDES reporting standards for Delphi studies [3], which have been cited in over 1600 publications as of June 2025, was low in this sample (11%). We encourage investigators to conduct Delphi studies with this guidance and recommend journals require the completed CREDES checklist with Delphi study manuscripts.
Author Contributions
A.M.O. study concept and design (co‐lead), analysis and interpretation of data, manuscript drafting (lead), and final approval of manuscript; S.E.J. study concept and design (co‐lead), analysis and interpretation of data, manuscript drafting, and final approval of manuscript; S.G.T. study concept and design, analysis and interpretation of data, manuscript drafting, and final approval of manuscript; M.R.D. data acquisition, manuscript drafting, and final approval of manuscript.
Funding
This work was supported by National Institutes of Health/National Institute on Aging: K99 AG086523.
Disclosure
There are no sponsors for this study.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Table S1: Search strategy.
Text S1: Inclusion and exclusion criteria.
Text S2: Extraction domains.
Table S2: Full characteristics of all studies.
Table S3: PRISMA‐ScR checklist.
Figure S1: Panelists attrition across rounds by total number of rounds.
Acknowledgments
The authors confirm that all individuals who have contributed significantly to the development of this manuscript have been named as coauthors.
References
- 1. Diamond I. R., Grant R. C., Feldman B. M., et al., “Defining Consensus: A Systematic Review Recommends Methodologic Criteria for Reporting of Delphi Studies,” Journal of Clinical Epidemiology 67, no. 4 (2014): 401–409, 10.1016/j.jclinepi.2013.12.002. [DOI] [PubMed] [Google Scholar]
- 2. Tricco A. C., Lillie E., Zarin W., et al., “PRISMA Extension for Scoping Reviews (PRISMA‐ScR): Checklist and Explanation,” Annals of Internal Medicine 169, no. 7 (2018): 467–473, 10.7326/m18-0850. [DOI] [PubMed] [Google Scholar]
- 3. Jünger S., Payne S. A., Brine J., Radbruch L., and Brearley S. G., “Guidance on Conducting and REporting DElphi Studies (CREDES) in Palliative Care: Recommendations Based on a Methodological Systematic Review,” Palliative Medicine 31, no. 8 (2017): 684–706, 10.1177/0269216317690685. [DOI] [PubMed] [Google Scholar]
- 4. Martin‐Kerry J., Scott S., Taylor J., et al., “Supporting Meaningful Participation of Older People in Core Outcome Set Development,” Journal of the American Geriatrics Society 73, no. 2 (2025): 661–665, 10.1111/jgs.19179. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1: Search strategy.
Text S1: Inclusion and exclusion criteria.
Text S2: Extraction domains.
Table S2: Full characteristics of all studies.
Table S3: PRISMA‐ScR checklist.
Figure S1: Panelists attrition across rounds by total number of rounds.