Study inclusion
From four electronic databases, we identified 10,956 citations. An additional 8 citations were retrieved from manual searching of reference lists. We identified 29 papers that met our inclusion criteria. Reasons for exclusion are reported in our PRISMA flow chart (Additional File 3). A detailed list of all citations excluded on full-text examination and reasons for exclusion can be found in Additional File 4. Agreement between reviewers for title and abstract screening was between 0.84 and 1.
Characteristics of included studies
Included studies were published between 2000 and 2021, with 82% of the studies published since 2016. Multiple countries of origin were represented across included studies such as the United Kingdom (17%), United States (14%), and the Netherlands (10%). Most studies (90%) were set in the community, followed by long-term care or retirement homes (7%), and then hospital (3%). In terms of study design, 76% were mixed methods, 10% randomized controlled trials, 7% protocols, 3% case studies, and 3% cross-sectional design. Additional File 5 summarizes the characteristics of our included studies.
Terminology and Operational Definitions of Co-Design
We identified 12 different terms to describe the process of co-design. Figure 1 shows the frequency of each term. Twenty-six (90%) studies provided operational definitions (Additional File 6), and 3 studies did not. Thematic analysis of the 26 operational definitions revealed repetitive concepts that were categorized into 10 themes that were further grouped into 5 proposed components of co-design as shown in Table 1. From this analysis, our proposed standardized definition for co-design is a user-centered approach involving collaboration between researchers, end-users, and other relevant stakeholders who are actively engaged throughout a process of iteration and continuous reflection to create an intervention tailored to the specific needs of the target population.
We systematically identified whether the proposed components of our co-design operational definition were carried out in each study’s methodology. Twelve studies included a user-centered approach, 17 demonstrated collaboration between researchers, older adults, and other relevant stakeholders, 11 tailored their interventions to the specific needs of the target population, 10 actively involved end-users throughout the study, and 5 demonstrated iteration or continuous reflection.
Figure 1. Use of Co-Design Terminology in Included Studies
Table 1
Thematic Analysis of Co-Design Operational Definitions
Terminology | Proposed Components of Co-Design |
User-centered approach (38–51) | Collaboration between researchers, older adults, and other relevant stakeholders (39,41,42,44–46,48,50–61) | Tailored to specific needs (38,39,46,47,50,53,55,57,58,62,63) | Active involvement throughout (39,43,45,48–50,58–61,64) | Iteration and continuous reflection (38,49,50,56,58) |
Co-Design (39,40,50,65) | (39,40,50) | (39,50) | (39,50) | (39,50) | (50) |
Co-Creation (13,28,30–33, 37,39,42,43,54) | (39,41–45,49,51) | (39,41,42,44,45,51,54,55) | (39,55) | (39,43,45,49) | (49) |
Action Research (41,42,51,58,64,67,68) | | (41,42,51,58) | (58) | (58,64) | (58) |
Participatory Design (38,45,48,57) | (38,45,48) | (45,48,57) | (38,57) | (45,48) | |
Community-Based Participatory Research/Participatory Research (46,59,60,63) | (46) | (46) | (46,59,60) | (59,60) | |
User-Centered Design (38,47,49) | (38,47,49) | | (38,47) | | (38,49) |
Participatory Action Research (53,56) | | (53,56) | (53) | | (56) |
Integrated Knowledge Translation (52) | | (52) | | | |
Qualitative Methodology (62) | | | (62) | | |
User-involvement (43) | (43) | | | (43) | |
Co-learning (64) | | | | (64) | |
Stakeholder involvement (61) | | | | (61) | |
Total | 12 | 17 | 11 | 10 | 5 |
Use of Co-Design in the Phases of Research
Figure 2a shows the distribution of studies according to the phase of research in which end-users were involved. Nine studies utilized co-design in more than one phase, such as both planning and conducting.
Figure 2a. Distribution of studies according to phase of research in which end-users were involved.
Determine the levels of involvement of the end-users in the included studies
The levels of involvement of end-users progress from consultation to collaboration to consumer control the highest level involvement. Most end-users were involved at either the consultation (52%) or collaboration (45%) level in the included studies. Figure 2b summarizes the different levels of involvement, the number of studies within each level, and the methods used within each level.
Figure 2b. Levels of involvement of end-users in the co-design process.
Measuring Success and/or Effectiveness of Co-Design
None of the included studies described any type of process to evaluate the success and/or effectiveness of the co-design process itself. However, some studies evaluated the success and/or effectiveness of the co-design process through indirect methods, such as measuring participant satisfaction of the PA intervention itself (n=4), intervention adherence measures (n=6), and by assessing changes in PA performance and levels of PA (n=5) (Table 2). Of the 15 studies that indirectly evaluated co-design, all reported positive trial outcomes that they attribute to the use of a co-design process.
Table 2
Outcome Measures used to Examine the Success and/or Effectiveness of Co-Design
Indirect Methods |
Participant Satisfaction | Adherence |
-Gaming Experience Questionnaire (40) -Fall Prevention Program Satisfaction Questionnaire (67) -Program and Engagement Satisfaction Surveys (54,61) | -Physical Activity Adherence Questionnaire (40) -Attendance recorded (54,58,61) -Recorded the number of days or time the product was used (66,69) |
PA Measures |
-Functional Capacity measured by The Senior Fitness Test (40) -Fall Prevention Behaviors Questionnaire to assess five areas: 1) fall prevention practices, 2) regular vision assessment, 3) medication use, 4) exercise and 5) home environment (67) -Physical Performance Test (PPT) to assess upper body muscle strength, lower body muscle strength, balance, and balance and gait (67) -Recorded total steps/day (66) -Amount of PA via accelerometer measurements (59,60) -Measured physical fitness using handgrip strength, chair stand, 2-min step, back scratch, sit and reach, and flamingo balance test (59,60) -Short Physical Performance Battery (SPPB) measures physical functioning using gait speed, standing balance, and lower leg strength (69) |
Barriers and Facilitators Associated with Co-Design
We grouped author-reported barriers and facilitators into 5 categories, including framework and methodologies (any theoretical frameworks or principles used to formulate the co-design process and/or study methodology), logistics (details surrounding how the co-design process was organized and executed), relationships (dynamics among participants, and between participants and researchers), participation (participant engagement in the co-design process and the efforts made by researchers to increase engagement), and generalizability (aspects of the co-design process that help to make the intervention more applicable to the target population) (Table 3).
Table 3
Barriers and Facilitators Associated with Co-Design
Categories | Barriers | Facilitators |
Frameworks & Methodologies | • Lack of literature on co-creation governance and frameworks (55) • Requiring participants to complete interventions before participating in co-design (70) • Complexity of data collection measures (38) | • Combining appreciative action and reflection, or an integrated knowledge translation approach with normalization process theory (52,64) • User experience honeycomb model (45) • Participatory action research (53) • Photo-elicitation (71) • Training workshop and focus group facilitators in co-creation (54) • Fieldwork tasks (55) • Involving participants early, frequently, and throughout various stages (57,61) |
Logistics | • Open-ended questions (62) • Novice facilitators (62) • Hypothetical scenarios (41,62) • Strategies to ensure members who belong to minority or socially disadvantaged groups are time intensive (59) | • Role and workshop aim clarification (55,57,64) • Debrief sessions for facilitators (64) • Sharing workshop summaries with participants (45,54) • Utilizing community organizations or creating community advisory boards to assist with recruitment (53,59) • Formal committee name and constitution (58) |
Relationships | • Time and resources required to build trust within community of interest (59) • Bias arises from a desire to please and maintain group dynamics (38,48) • Participatory Action Research approach may create a power differential (53) • Seniors require longer interviews (38) | • Building group dynamics (meeting in person, allowing time to socialize, and demonstrating appreciation for participation) (40,57,61) |
Participation | • Fatigue/loss of concentration (41) • Participants who are more physically active may develop a louder voice than inactive participants taking away the perspective of the target population (39) • Cognitive, sensory, or physical disabilities may hinder participation (48) • Administrative costs associated with ongoing involvement (58) • Unfamiliarity with technology (49) | • Short sessions to prevent fatigue (38) • Homework tasks (64) • Small groups (63) • Comfortable location (59) • Active facilitator involvement (38,39) • Assisting older adults to fill out questionnaires (38) • Increasing ownership of project (55,67) |
Generalizability | • Small sample size (52,63) • Volunteers are more outspoken and active members of the community (59) • Variation in resources between communities (68) | • Purposive sampling (45) • Recruiting both experienced and novice technology users (65) |