‘We need transparency and communication to build trust’: exploring access to primary care services for young adults through community-based youth participatory action research and group concept mapping

Young adults’ access to care

Young adulthood is a critical period for establishing lifelong health behaviors, presenting an important opportunity for healthcare providers to engage and support young people through routine care [1], 2]. Recommended preventive health care for young adults (approximately 18–26 years of age) includes screening for non-communicable and infectious diseases, mental health care, vaccinations, family planning, and health education [3]. However, the transition from pediatric to adult healthcare services is also a point of vulnerability for access, resulting in high rates of discontinuity in care, especially for primary care and preventive health services. Upon aging out of usual sources of care (i.e. pediatric practices or school-based health centers), many young adults face barriers to accessing adult-centered health services. As the final stage of the transition process, transfer to adult health services is critical to continuity of care. There is a need to better understand and address primary care access for young adults.

A growing body of literature has documented a drop off in numerous health indicators for young adults as compared to adolescents, including health risk behaviors and access to care [4], [5], [6], [7], [8], [9], [10], [11]. Young adults have higher emergency department utilization rates with disparities along gender and racial lines; nearly half of the care delivered to young Black men takes place in emergency departments [12], 13]. Young adults face structural barriers in initiating adult healthcare services, including disruptions in insurance coverage, housing, employment and educational status. Further challenges arise from learning to navigate and initiate care independently from parents and caregivers [1]. Existing health disparities are perpetuated, if not amplified, along racial and economic lines in this transition to adult health care, with particularly low rates of preventive health services delivered to young Black and Latino men [14], 15].

Conceptual framework: youth-engaged research

Community-based participatory research (CBPR) is an approach to scientific investigation and program development that is based on partnership and collaboration between researchers and communities at every stage of the research process [16]. A subset of CBPR, youth participatory action research (YPAR), centers youth as experts in their own lived experiences and integrates youth-led advocacy based on study findings [17]. Both CBPR and YPAR provide avenues for capacity building and positive youth development through community-academic partnerships [18]. In the “action” stage of YPAR, youth partners leverage the study outcomes to create transformative change in their community. Through the process of YPAR, youth participants gain skills and experience in academic inquiry and civic engagement, while building social connections and developing self-efficacy.

Group concept mapping

Group concept mapping (GCM) is a mixed-methods participatory methodology that has been previously applied in CBPR and YPAR studies [19], 20]. GCM inherently centers the lived experiences and priorities of participant-researchers through a series of two participatory activities: (1) group idea generation (i.e. brainstorming) and (2) individual data structuring and rating. The resulting “concept map” is a graphical representation of the themes that emerged during the group discussions, as interpreted by the participants. GCM has been used to engage adolescents and young adults in studies focused on community health, health equity, program design, and health risk behaviors [20], [21], [22], [23].

Data analysis

The GCM methodology combines qualitative data collection with quantitative data analysis. Multi-dimensional scaling (MDS) and hierarchical cluster analysis (HCA) were conducted using the groupwisdom™ software. Responses from the sorting activity are aggregated and converted into quantitative data in the form of a similarity matrix, which assigns a binary value based on whether any two given statements were sorted together in each potential instance. MDS is used to assign each point a coordinate on an x-y graph to spatially represent the “closeness” of each statement to each other statement based on the similarity matrix. Then, HCA is applied to the basic point map to create clusters of the most closely related points, typically resulting in several cluster models. A best-fit cluster map is a graphical representation of the themes that emerged in discussion, as interpreted by each participant [4], 5]. The stress value is the metric used to determine the appropriateness of fit of the multi-dimensional scaling solution to the original similarity matrix. Values typically fall between 0.20 and 0.35, with a lower value indicating a closer fit, i.e. less stress [26].

As opposed to the cluster map, which graphically represents the themes that emerged in the brainstorming session, the go-zone map assigns each statement a spatial coordinate based on its average value for two rating variables along the X-Y axes, which in this study were “importance” and perceived “feasibility” for intervention. The go-zone map provides valuable information about the priorities that the participant group places on each statement, as a product of its importance and feasibility ratings. A pattern match analysis can be used to compare ratings between two participant groups.

Concept map

A 4-cluster arrangement was determined to be the best fit concept map (Figures 1A and 1B). The stress value for the chosen map was 0.21, within the range of interpretable values, indicating an acceptable fit [26], 27]. Four topic clusters emerged: Acceptance & Etiquette; Communication & Comfort; Availability & Accessibility; and Knowledge & Resources.

1. Acceptance & Etiquette: This theme captured intangible characteristics of the patient experience, including interpersonal interactions with clinical staff and healthcare providers, and cues that contributed to participants feeling respected and welcome within the clinical space, particularly with regards to identity (i.e. gender identity or race/ethnicity). Included in this theme was also the patient’s comfort seeking care in general, and medical mistrust based on negative personal experience or accounts from family, friends, or in the media (e.g. news coverage of physician abuse cases).

2. Communication & Comfort: This theme encompassed the operations of the clinical office as perceived by the patient, from a sense that the clinical office was adequately staffed, to language services and the physical environment (e.g. clean, well-furnished).

3. Availability & Accessibility: This theme encompassed statements related to clinical care access, including the process of making appointments (including same day/urgent care appointments, follow up appointments, and referral appointments within the system), availability of appointments, and contacting the clinical office. Also included in this theme were transportation to clinical space, having a medical home (multiple specialties in one place), and having agency over provider choice.

4. Knowledge & Resources: This theme reflected the need for ancillary services directed towards young adults, including education around available services, healthcare navigation support, and insurance navigation. Many participants referred to their assigned “advocate,” a service provided through The Hub that pairs youth members with a staff member who are equipped to answer questions regarding healthcare navigation. The item with the highest importance rating in this theme was “financial transparency,” referring to an expressed fear of surprise billing and a lack of information about cost of services, which participants stated was a barrier to accessing care.

Figure 1A:

Cluster map. Each cluster represents a theme that emerged from group discussion and analysis. Each point represents a statement. Stress value of 0.21 indicates an acceptable fit for this particular solution. Points are geographically located on the map based on multi-dimensional scaling analysis that averages all valid cluster activity responses. Points that are closer together were more frequently clustered together in the participant responses.

Figure 1B:

Statements by cluster. Statements are listed in full, organized by cluster topic, with statements ordered by average importance rating (on 1–5 scale), listed to the right of each statement.

Go-zone map

The go-zone map (Figures 2A and 2B) positions each statement as a coordinate on a spatial plane based on the average participant ratings for “feasibility for intervention” (X axis) and “importance for intervention” (Y axis). The plane is then divided into quadrants “low feasibility/low importance,” “low feasibility/high importance,” “high feasibility/low importance,” and “high feasibility/high importance.” The statements in the “high feasibility/high importance” category can be considered to be the highest priority for intervention.

Figure 2A:

Go-zone map. Coordinates are based on each statement’s average rating for “importance for intervention” (Y-axis) and “feasibility for intervention” (X-axis), on a 5 point scale. Statements are numbered for reference in Figure 2B. The correlation value is 0.07, indicating no predictable correlation between the two variables.

Figure 2B:

Statements by go-zone quadrant. Statements (numbered) are listed in full, grouped by quadrant of the go-zone map.

Pattern match

The pattern match graph (Figure 3) compares cluster ratings between two groups. Here, we compare importance ratings between youth and staff participants. The two main findings are that the youth ratings are consistently higher (on a 1–5 scale) than the staff ratings, and that the most discordant importance rating between the two groups is for the “Communication & Comfort” cluster, which contains the items “ease of communicating with my provider,” “physical clinical space (i.e. nice furniture, clean bathrooms, murals on the wall),” “adequate clinic staffing,” “language accessibility,” and “personal barriers.”

Figure 3:

Pattern match for importance, youth vs. staff. The pattern match graph compares youth and staff importance ratings for each cluster.

Youth-led interpretation

The study results were presented to The Hub youth and staff in March 2024 for further discussion and interpretation. In discussion groups led by youth researcher leaders, The Hub youth members who attended (8 in total, ages 19–24 years, 3 of whom participated in the prior study methods) generated labels for the concept map clusters, and discussed the overarching themes that emerged from the data, which one participant summarized as “we need transparency and communication to build trust.” Youth then generated several recommendations for interventions to present to hospital system stakeholders, which were subsequently presented to clinician and administration stakeholders to advocate for changes to improve access to care for young adults. These proposals include a text-based notification system that would allow patients to know when they are next in line to see the clinician when appointments are delayed, an “office hours” feature for sending messages to their provider during which they can expect an immediate response, and a web-based educational repository to learn the basics of healthcare navigation (including how to make an appointment, find a provider, manage prescriptions, and navigate insurance benefits).

Youth participatory action research

Youth participatory action research (YPAR), as a variant of community-based participatory research (CBPR), has utility beyond the descriptive research outcomes. The process of involving youth participants as leaders and co-researchers builds capacity through several avenues. In our project, one youth leader (JS) was trained in research ethics and methods, co-led participant activities, was active in data analysis, and presented findings at a research symposium. A second youth leader (AM) was actively involved in preparation for participant sessions and data analysis. Both youth leaders facilitated the youth participatory interpretation sessions, presenting results to participants and eliciting their perspectives and interpretation of results. The participants themselves were exposed to the process of academic inquiry through transparent and participatory study processes and were active in interpreting the results through a youth-led participatory interpretation session. While we did not evaluate the impacts of study participation, previous studies have shown YPAR to increase self-efficacy, empowerment, and civic participation among participants [28].

Strengths and limitations

A major strength of the study is that participants represented a group of young adults from demographic backgrounds that historically face numerous barriers to health access; nearly all participants identified as Latinx or Black-Latinx and live in Spanish-speaking households in low-income urban neighborhoods. All study activities took place at the Uptown Hub facility where participants were familiar and comfortable in the setting. One limitation, however, is that by partnering with a youth-serving program affiliated with a large healthcare system, we may have selected for participants who are more connected to care than demographically similar youth in New York City. While The Hub does not provide clinical services aside from mental health care, they do provide healthcare navigation support including assistance with insurance enrollment. Since all participants reported having health insurance coverage, this study did not capture the experiences of uninsured young adults, a group who face significant barriers to healthcare access. We hypothesize that non-Hub connected youth in the community experience at least these barriers, and likely more.

Limitations to the methodologic design included the observed cognitive difficulty of the sorting and rating activity, which was a barrier to collecting valid data from several participants, as evidenced by several invalid responses to the sorting activity. In future studies with a similar population, we would spend more time reviewing the instructions, sharing examples, and having participants practice the activity before completing it. In order to mitigate anticipated task fatigue, we limited the number of statements for the sorting and rating task to 34, which is smaller than other studies with similar populations. Some of the items that were rated lower for “importance”, including language access, accessibility, and clinical environment for non-binary and transgender individuals, were likely affected by low representation among the participant group, of which there were fewer gender expansive participants, with all participants being fluent in English. While this study did include the perspectives of adult staff members who work closely with youth on matters related to health system navigation, future studies could explore including the perspectives of caregivers and healthcare providers who care for young adults.