Using photovoice to define “community” in substance use disorder research engaging service users, providers, and policymakers

Abstract

(197 words)The ideals of community-engaged health research are challenging to realize ethically and methodologically, particularly when focused on substance use disorder (SUD), in which the power differentials inherent to stigma marginalize people with lived experience (LE) as full community participants. This study provides a typology and exploratory data generated through Photovoice to understand relationships among community members with different roles related to LE of SUD (i.e., SUD service providers, service users, and the public) in the context of a large-scale community-engaged opioid overdose reduction intervention. A team of 14 participants and researchers used an adapted thematic analysis method to interpret Photovoice data from 65 total participants in Kentucky, Massachusetts, and Ohio. Photovoice data indicated the influence of stigmatization on perceptions of service users as problems to be solved by service providers without the significant investment of the rest of the community. Service provider burnout is exacerbated by competition for scarce public investments in addressing the dire need for SUD services amid the opioid epidemic. Community-engaged SUD research can be improved by defining the parameters of communities in explicit, participatory, and reflexive ways and recognizing how disinvestment and stigmatization influence relationships among community members occupying different roles in SUD services.

1. Introduction

There are significant but under-explored implications of how we use the term “community” in health services research, especially considering the power differences central to stigmatized behavioral health conditions (Clouston & Link, 2021; Frederick et al., 2017; Hatzenbuehler et al., 2013; Link & Phelan, 2001). Communities may be categorized based on identity, geography, and/or affinity, but they are not naturally occurring, apolitical spaces (Bradshaw, 2008; Duea et al., 2022; Hill Collins, 2010). Initial decisions about what constitutes a community in participatory health research have critical consequences throughout a project (Duea et al., 2022; Zimmerman, 2020). These considerations are crucial as community engagement frameworks proliferate in the literature and become “inclusion imperatives” for funders in response to service user demands for “nothing about us without us” and arguments that democratization of knowledge production improves external validity (Green & Johns, 2019; Greer et al., 2018; MacKinnon et al., 2021; Ortiz et al., 2020; Renzetti, 1997; Vaughn & Jacquez, 2020; Voronka, 2017, 2019).

Yet, there is a gap in health services research between the valorization of community engagement in principle and actual involvement in practice, which too often amounts to tokenistic representation, co-optation, and false empowerment (Arnstein, 1969; Brown & Jones, 2021; Colder Carras et al., 2023; Damon et al., 2017; Daya et al., 2020; Egid et al., 2021; Green & Johns, 2019; Greer et al., 2018; Groot et al., 2022; Happell et al., 2018; Jones et al., 2023; Marshall et al., 2015; Simon et al., 2021; Souleymanov et al., 2016; Williams et al., 2020; Zimmerman, 2020). Although increasingly prioritized by funders, input from the public and service users in a consumer-based healthcare marketplace entails contradictory imperatives. These conflicting priorities further muddy definitions of “community” with notions of citizens exercising their democratic rights, consumers optimizing their personal choices for care, and lay experts using experiential knowledge to inform health improvements (Knaapen & Lehoux, 2016; Madden & Speed, 2017; Williams et al., 2020). When defining communities to be engaged in health services research, investigators are often imprecise in distinguishing among interest groups referred to as 1) the public (a.k.a., citizens, taxpayers), 2) health services providers (a.k.a., professionals, practitioners), and 3) health service users (a.k.a., consumers, patients) (Brown & Jones, 2021). Typically, only community members in the provider category have the compensated time and professionalization to participate in research activities (Blue et al., 2019). These unclear classifications of community undermine our understanding of which members are amplified and excluded in community-engaged research.

The discrepancy between participatory research ideals and realities may be exacerbated by service provider and public mistrust and stigmatization of service users rooted in power differentials among these roles (Hatzenbuehler et al., 2013; Link & Phelan, 2001; Valdez et al., 2024). Enacted stigma involves individuals in positions of power – such as behavioral health providers and criminal-legal system (CLS) professionals who have been shown to hold stigmatizing attitudes (King et al., 2020; Moore et al., 2023; Sufrin et al., 2024) – excluding people judged to be socially problematic from accessing health and social services (Addison et al., 2023; Clouston & Link, 2021; Link & Phelan, 2001). Stereotypical beliefs driving stigma comprise the extent to which health conditions are perceived as being dangerous, socially undesirable, visible, controllable (i.e., caused by the individual), and/or changeable (e.g., recovery is perceived as possible) (Krendl & Perry, 2023). Internalized, anticipated, and perceived stigma can result in losses of social identity and social harm that definitionally exclude stigmatized people from belonging to a community (Addison et al., 2023; Ghane et al., 2021; Moore et al., 2023). Loss of social identity and social harm is compounded by criminalization of substance use disorder (SUD) symptoms (e.g., substance possession), imperatives of self-help groups for people in recovery to adopt “diseased identities” that may reinforce internalized stigma, and the disproportionate lack of access to SUD services among non-Hispanic Black people in the U.S. (Khan et al., 2023; Larochelle et al., 2021; McGovern et al., 2024; Rosales et al., 2022; Timmer et al., 2022).

Service user advocacy for greater inclusion has led to the development of peer support specialist (PSS) roles in behavioral health prevention and treatment as system navigators with lived experience (LE) (Voronka, 2017). PSS legitimacy involves “experiential” expertise, which requires that PSS code-switch their comportment, language, and dress to perform “authentic” identities as people with LE to clients, and “professional” and “respectable” identities sometimes without compensation and benefits commensurate with other team members (Greer et al., 2018; MacKinnon et al., 2021; Voronka, 2019). Within the broader context of confusing definitions of community in health services research, PSS occupy particularly complex roles in SUD service provision with conflicting demands and positions in professional hierarchies.

Thus, imprecise definitions, tokenization of LE, and stigmatization represent major challenges to rigorous community-engaged SUD research. The current project responds to recent calls (Addison et al., 2023; Jones et al., 2023) for clearer distinctions among the roles of service users, practitioners, and the public based on varying positions of power. We use a participatory action research (PAR) method, Photovoice, to explore whether and how participants in differing roles perceive SUD service users as equal members of communities. Photovoice is an appropriate method to address this question because of its design to empower participants to assess, explain, and change their communities’ health concerns (Evans-Agnew et al., 2022; Macdonald et al., 2022; Seitz & Orsini, 2022; Strack et al., 2022; Wang & Burris, 1997). Additionally, Photovoice has been widely used to engage people with SUD in understanding factors influencing recovery (De Seranno & Colman, 2022; Flanagan et al., 2016; Muroff et al., 2023; Smith et al., 2022). This study also reflects on what Photovoice participants learned from each other’s photos and stories and the value of Photovoice as a non-tokenistic research involvement mechanism (Jones et al., 2023). The goals of this study are to 1) report on Photovoice findings regarding differences in power and community belonging across roles and 2) reflect on the utility of Photovoice in facilitating meaningful inclusion of people with LE in SUD research.

2. Methods

2.1. Context, data collection, and sampling strategy

This Photovoice project was a supplement to the HEALing Communities Study (HCS) funded by the National Institute on Drug Abuse and Substance Abuse and Mental Health Services Administration to evaluate the impact of implementing evidence-based practices across healthcare, behavioral health, and CLS sectors, paired with community engagement and anti-stigma communication campaigns, on reducing opioid overdose deaths (Walsh et al., 2020). This four-site, parallel-group, cluster-randomized wait-list controlled trial involved partnering with local coalitions within geographically defined communities (i.e., county, towns) and assigning each community to either the intervention (Wave 1) or the wait-list comparison group (Wave 2) (Walsh et al., 2020). HCS comprised 67 communities in Kentucky, Massachusetts, New York, and Ohio that, on average in 2016–2017, had twice the national rate opioid-related fatalities (El-Bassel et al., 2020). Of these communities, nearly half were rural and 17 are considered Appalachian, where there are high concentrations of areas facing federally designated Health Care Professional Shortages and significant health disparities (El-Bassel et al., 2020). The 67 HCS communities are home to 8.67 million people, with the average population of an HCS community totaling approximately 129,000 (El-Bassel et al., 2020).

The Kentucky, Massachusetts, and Ohio sites developed and executed a shared Photovoice protocol as an optional supplement to understand localized study contexts that allowed for cross-site comparison, participatory analysis, member-checking, and presentation to HCS coalitions and advisory boards (Balvanz, Olvera, and McGladrey et al., 2024). In spring 2022, we worked with Wave 1 HCS local coalitions and partner agencies to recruit people with current or previous personal SUD (defined as “SUD-LE” in this paper) and SUD service providers. In Kentucky, there were 10 Photovoice groups with a total of 45 participants: eight community-based groups (four rural, four urban) comprising people with LE of SUD and SUD service providers and two groups of PSS employed by HCS (one urban, one rural). In the Massachusetts and Ohio sites, participants were people with LE of SUD, with eight participants across two community-based groups in Massachusetts (one urban, one rural) and 12 participants across two community-based groups in Ohio (one urban, one rural). The majority of participants across sites identified as female and white, with ages ranging from 21 to 74 and roughly equivalent percentages of participants with education levels from high school to graduate education. Most knew of a family member or close friend who has SUD (see self-reported demographics in Table 1). Participants attended up to six sessions including an orientation, 2–4 photo-discussion sessions, and a dissemination planning session. Photo-discussion sessions consisted of 1) photo-topic development (e.g., what stigma looks like in my community), 2) presentation of photography, and 3) critical discussion of photo meanings and interpretations. Participants received between $25 and $50 per session. HCS team members created displays of photos/caption combinations and topics developed by participants into postcards, posters, PowerPoint slides, and videos to convey SUD-related barriers to and opportunities for community change and facilitated dissemination among local and state HCS partners. All Photovoice data collection with 65 total participants occurred between March and August 2022.

Table 1.

Demographics by HCS Photovoice Site (n = 65 total participants).

	KY (n = 45)	OH (n = 12)a	MA (n = 8)
Gender (%)
Male	37.8%	9%	50%
Female	62.2%	91%	50%
Race and ethnicity (%)b
Hispanic/Latinx	0%	0%	0%
Non-Hispanic White	93.3%	100%	100%
Non-Hispanic Black	6.6%	0%	0%
American Indian/Alaska Native	2.2%	0%	0%
Average age in years (range)	42 (21–74)	36 (25–52)	36 (31–52)
Education level (%)
High school degree or equivalent	4.4%	45.4%	12.5%
Some college, no degree	20.0%	18.2%	25.0%
Associate degree	8.8%	18.2%	25.0%
Bachelor’s degree	13.3%	18.2%	0%
Graduate education	22.2%	0%	12.5%
Did not respond	31.3%	0%	25.0%
Knew of a family member or close friend who has a substance use disorder (% yes)	62.2%	100%	100%
Personal connection to SUDc (n)
Service provider	36	1	2
Person in recovery	18	12	8

^a12 Ohio participants contributed to Photovoice data collection, but one did not provide demographic information, so these percentages are based on 11 participants. The twelfth participant self-identified as someone in recovery.

^bThese percentages do not add up to 100% because participants were allowed to select multiple race/ethnicity categories for self-identification.

^cAt this stage in the project, this question was framed as a qualitative prompt to assess the range of community roles represented among participants, which is why the categories do not directly correspond to the typology. Roles are not mutually exclusive (i.e., a person can be a service provider and a person in recovery); because of this, numbers of respondents are provided rather than percentages.

This study protocol (Pro00038088) was approved by Advarra Inc., the HCS single Institutional Review Board. As detailed in the HCS Photovoice protocol (Balvanz, Olvera, and McGladrey et al., 2024), the facilitators oriented participants to the ethics of photography, including how to avoid taking identifiable photos, aligned with recommendations for informed consent in Photovoice projects with people in SUD recovery (Smith et al., 2024). Participants also determined whether and how they wanted to be publicly involved in the dissemination of Photovoice projects and could maintain total anonymity if desired.

2.2. Qualitative approach, research paradigm, and positionality

Based upon the literature and ideas about potential PSS role conflicts from data collection and initial review of Photovoice transcripts, we developed a conceptual model, Community Roles Typology (Fig. 1). Categorization on the horizontal axis is based on an individual’s SUD-LE as a service user. Categorization on the vertical axis is based on an individual’s primary role affiliated with service provision: what pays their bills (i.e., service provider); motivates them to volunteer time (i.e., advocate); or involves them directly or indirectly in funding SUD services and research (i.e., researcher, policymaker, and/or taxpayer). PSS are categorized as service providers with SUD-LE.

Fig. 1.

Community roles in relationship to community-engaged substance use disorder research

*Footnote: The asterisked roles are more theoretical than realistic because of employment challenges associated with actively receiving services, but all the other categories were represented by HCS Photovoice group members.

The Community Roles Typology informed subsequent coding and thematic analysis, which was utilized for this study because the method accounts for differences in positionality and social location (Clarke & Braun, 2021). Photovoice researchers re-contacted participants and facilitators about their interest in co-authoring this paper, and 14 people responded (see Table 2 for their self-reported role affiliations). Contributors were asked to write about prompts reflecting on the roles illustrated in Fig. 1, what community means to them, their motivations for involvement in opioid overdose reduction efforts, their experiences with the Photovoice process, and how their skills and interests align with Photovoice.

Table 2.

Co-author community roles related to the opioid overdose epidemic.

	Advocate	Provider	Researcher
Person actively receiving services and/or involved with the CLS
Person with previous history receiving services and/or CLS	2	1	2
Person without previous history receiving services and/or the CLS	1	1	7

The first author completed an initial review of the aggregated co-author responses and noted prominent topics that were used to guide the coding framework. The second author open-coded the aggregated co-author responses based on the coding framework. A group of trained staff and students applied codes to Photovoice transcripts (45 transcripts across the three sites) and photos/captions. Staff and student coders used the comment feature in Microsoft Word and PowerPoint. The second author reviewed coding for consistency and transferred coded documents into NVivo 12 (Lumivero, 2023) for further analysis. We analyzed Photovoice transcripts, photo data, and co-author responses labeled with the “role-relations” code. Participants are located by role in Fig. 1 using italics (e.g., provider with SUD-LE), and participants who later provided co-author responses are noted for example as follows: provider with SUD-LE/co-author.

3. Results

Results are presented as three themes with underlying subthemes (see Table 3). The first two themes address the study’s aim to report on Photovoice findings regarding differences in power and community belonging among roles by describing participants’ perceptions of whether and how people are treated as equal members of their communities within and across service user roles. We then consider participants’ frustration with the disinvestment in SUD services and competition for limited funding, which exacerbate provider burnout and compassion fatigue. Finally, the third theme suggests the potential of Photovoice as a PAR method to facilitate reflection on intersecting role positionalities in SUD research. The final theme aligns with the second goal of the study to explore how Photovoice can meaningfully include people with LE in SUD research.

Table 3.

Themes and sub-themes represented in the Photovoice data.

Theme	Sub-themes
Differences in perceived status as community members across service user roles	Stigmatization causes a loss of social identity. PSS are needed to buffer against discrimination. People without LE have important roles in addressing stigma.
Conflicts and community disinvestment in SUD service provider roles	Community power players do not take ownership of addressing in SUD and invest in change. Lack of investment commensurate with demand for services foments provider competition and burnout. Burnout and compassion fatigue is heightened by competition for limited SUD service funding.
Value of Photovoice in reflecting intersecting roles in community-engaged SUD research	Photovoice facilitates perspective-taking across service user, provider, and policymaker/taxpayer roles. Photovoice helps researchers create a common language and develop shared understanding across roles.

3.1. Differences in perceived status as community members across service user roles

Several advocates with SUD-LE shared that people without SUD-LE viewed people seeking help for SUD in their communities with mistrust, anger, and fear, which diminished help-seeking persons’ social identities. An advocate with SUD-LE described shame in asking for help in a society prizing independence:

There’re so many situations where we could make so much progress if a couple people were working together on it. But we’re taught as a society: Independent … And that was part of my using. In my head, I was just supposed to know how to live … And that pride in me took over, and I couldn’t say, ‘I don’t know what the f**k I’m doing. I’m bottoming out here. Can somebody just give me a hand?’ _… Everybody else appears to be getting along just fine. Something must be wrong with me. And there’s stigma in saying that.

Particularly, many advocates with SUD-LE mentioned the shame that accompanied CLS-involved experiences of SUD, such as the hopelessness of being “doomed from the start” by juvenile justice involvement; the culture shock of reentering communities with the expectation of independence after experiencing the constant surveillance of incarceration; being labeled by a judge as a “menace to society” for selling substances to fund their addiction; and being put down by people in authority in front of their children. One advocate with SUD-LE stated that shame is compounded by threadbare social supports: “We’re already struggling, and we’re already living pretty much to pay for childcare to barely get to work. And when we’re trying to get above that entry level job to break through the glass ceiling to get to that next organizational level, childcare can be such a barrier. Transportation can be such a barrier. It keeps us in that lower class altogether.” Several advocates and providers with SUD-LE justified this treatment as deserved for their behaviors while in active use but also recognized the effect of stigmatization on their loss of social identity as community members, doubting that their voices matter because they have been made to feel small by their communities.

An advocate with SUD-LE said that being discriminated against by “important people” necessitates the “buffer” of empathetic PSS. Advocates and providers with SUD-LE from all three sites stated that belonging to a recovery community with common goals, interests, and thought patterns allows them to meet others where they are and work together through shared struggles with stigma, recovery, and system navigation barriers. However, advocates and providers with SUD-LE noted the heterogeneity of recovery communities and stigmatization of some recovery paths. A provider with LE compared these differences to the theological disagreements among Christian denominations (see Fig. 2). Another provider with SUD-LE described judgement in the Anonymous community (AA/NA), saying, “You’re interacting most of the time with people whose only experience is being in recovery, and their only experience is AA, and they will sit and tell you, like I was told over and over, ‘You are going to fail because you don’t want to do AA.’ … No matter how many times I asked for alternatives, nobody ever had any.” An advocate with SUD-LE said that sharing about medication treatment and mental health struggles was perceived as a moral failing in their experience with Anonymous groups. Comparisons of the relative severity of individuals’ experiences (e.g., CLS involvement, drug type/administration route, poverty) create division and separation antithetical to peer camaraderie, according to one provider with SUD-LE.

Fig. 2.

Example of photo and caption on perceived differences in status as community members across service user roles.

Participants with SUD-LE indicated that people without SUD-LE, especially those in positions of political power and/or authority as service providers, also have important roles to play in shifting stigma toward service users. An advocate with SUD-LE/co-author said, “Not all professionals have to be people in recovery, but they should require some training in how to coach, guide, speak to [people], ‘hands-on’ training to help provide services.” Another advocate with SUD-LE/co-author described the importance of having powerful social network connections to the recovery process:

People can work their way up, build a network, and get to know some of those people in positions of power, but the sad reality is that the process of networking is daunting for a person with SUD who already lacks a variety of competencies in areas like social skills, economic status, education, and transportation. One woman in our Photovoice group expressed that it was sobriety-threatening at times to watch others continue to grow in their recovery because of who they know, as she continued to encounter hurdles because she didn’t know the ‘right’ people.

A provider with SUD-LE/co-author said that destigmatization of service users can happen when they speak loudly to “the ones at the top, those ones in the front, those ones that have that power.” These findings demonstrate that common interests and understandings of LE among community members within and across service user roles cannot be assumed and must be deliberately cultivated when engaging participants in community-based SUD research.

3.2. Conflicts and community disinvestment in SUD service provider roles

The data showed that participants perceived policymakers and taxpayers as uninterested or unwilling to take ownership of a “problem” that only affects lower-value members of their communities. As illustrated in Fig. 3, even personal SUD-LE did not necessarily reduce a provider with SUD-LE’s mistrust of others with SUD-LE as professional colleagues. A provider without SUD-LE explained prevailing notions of service provision and how communities help people with SUD:

Fig. 3.

Example of photo and caption of conflict and community disinvestment in SUD service provider roles.

Drug addiction, struggles, that is not a problem, it’s a condition. You’re not going to solve it ….It’s a person, a family at a time … We make incremental changes and incremental improvements … The people who can provide help, who have the capacity, they’re just not going to get that shot of dopamine like you do by handing money to somebody who you think needs it … But you do that to make yourself feel good …

A provider without SUD-LE/co-author suggested that communities use “the language of investment” to shift community attitudes toward people with SUD-LE, “because investments don’t always yield immediate returns. It takes a while.” Unfortunately, participants did not see power players in their communities taking ownership and investing in change. For example, a provider with SUD-LE who runs a recovery nonprofit described a vicious cycle in which “I’m trying to fundraise so that I can pay my grant-writer so that I can write the grants to receive the awards to pay for the services that everybody needs and wants … because we can’t charge the people who need it the most to provide the service.” Another provider without SUD-LE noted:

… the disconnect between those who make policy and those who actually carry it out … Depending on where you are in the country, it’s ‘let’s spend more money on law enforcement,’ rather than spending money on the back end for … providing resources for treatment and … to help bring them out of poverty.

Advocates and providers with and without SUD-LE in at least five of the Photovoice groups described rampant Not In My BackYard (NIMBY) attitudes among policymakers and taxpayers without SUD-LE against shelters, treatment and harm reduction facilities, and affordable housing developments.

Several providers with and without SUD-LE suggested that the lack of policymaker/taxpayer investment in SUD services foments burnout and competition among service providers for funding that is not commensurate to the skyrocketing demand for services. They discussed strained CLS, social services, and behavioral health workforces with persistent vacancies because of insufficient compensation, especially in the wake of the COVID-19 pandemic, with one provider without SUD-LE describing how “we’ve had them just leave the industry all together … I’ve had staff that were like, ‘You know what? I’m going to go get a Monday through Friday, whatever time to whatever time, factory job.” A provider without SUD-LE observed that burnout among his colleagues breeds defeatism, fatalism, and apathy.

Burnout and compassion fatigue are heightened by competition for limited SUD service funding. One provider without SUD-LE said competitors poach her behavioral health staff with “‘blood money,’ saying, ‘I will pay you $10,000, but you have to leave now.’ Like, you don’t care about patient care?.The people that she’s serving? You’re just like, ‘Leave now and come here.’” Another provider without SUD-LE detailed a cross-sector meeting of CLS, behavioral health, and policy-maker groups that became “very heated about who gets to do what, and that possession mentality of I’m the only one that can help this group of people, I’m not going to work with this provider.” Although some providers with and without SUD-LE described improved coordination across agencies and sectors in recent years in response to greater overdose mortality from heroin and fentanyl, they underscored that inadequate funding and staffing for SUD and related social services reduced their capacity to address service users’ complex concerns and spend time required to build trust. These mismatches between resource needs and public investments intensify stigma against SUD service users as costly social problems.

3.3. The value of Photovoice in reflecting on intersecting roles in community-engaged SUD research

Co-authors with different service user and provider roles found the Photovoice project impactful in taking the perspectives of team members with varying points of view on the opioid epidemic in their communities through sharing their photos and stories. A provider without SUD-LE/co-author said her biggest takeaway was learning from others who work in a different field and felt empowered by other community members asking important questions about her role and purpose. A likeminded provider with SUD-LE/co-author said:

The other great element of Photovoice was not only the community aspect of literally getting out in my own community to interact with people, but also the interaction of a community of other Photovoice volunteers … It was a community of people coming together with a creative cause … There was a fellowship where you felt like you were making a difference even if a small one. We made a connection of friendship and workship. I know it may not be a word, but it should be.

A provider with SUD-LE/co-author wrote: “What we saw matters. How we see it matters. How we choose to respond matters … The epidemic affects the whole community either directly or indirectly. There is no neutral zone.” An advocate with SUD-LE/co-author found it meaningful to express herself and be heard by and involved with people in positions to influence change in her community.

HCS researchers without SUD-LE/co-authors who facilitated groups reflected on their efforts to speak a common language with participants and to level-set expectations in developing shared understanding across roles. A researcher without SUD-LE/co-author who facilitated Photovoice groups was struck by the vividness of participants’ stories and photos in conveying LE intersecting with community roles, observing, “It’s one thing to listen to the stories of those who are living in active addiction or in recovery; it’s quite another to look through their eyes via their photographic representations, to see the world from their perspectives, and to encounter their self-stigma on such a personal level.” One researcher without SUD-LE/co-author asked the following reflective questions:

Particularly for facilitators who do not have a background in mental health or addiction services, how might we better prepare for an instance where a coresearcher might struggle during or after the sessions? Additionally, how do we better prepare ourselves as facilitators for the emotional weight of discussions during these projects and the impact(s) that may have on us personally?

Another researcher without SUD-LE/co-author shared her sense of obligation to accurately yet concisely summarize interim analyses presented to participants to ensure her synthesis of individual stories into collective reckonings did justice to participants’ LE. The protocol’s requirement of participatory analysis/photo-captioning and exhibitions to HCS coalitions and advisory boards elevated researchers’ accountability to preserving the integrity of participants’ voices, which was evident when one participant who missed the final session was upset with the caption selected by others until researchers explained the sessions and volunteered to redesign the photo/caption combination. This project’s application of Photovoice illustrated the method’s usefulness in inspiring reflection not only among participants occupying different roles but also among researchers in their responsibility to amplify participants’ creative expressions of their SUD-LE and experiences as providers to be heard by community members in positions of power.

4. Discussion

This study confirms how stigmatization processes and imprecise specification of roles in SUD service provision have the potential to invalidate solely demographic- or geography-based definitions of communities in SUD research because they do not account for role-based differences in which individuals are perceived as full community members. SUD service users, providers, and policymakers/taxpayers have unequal levels of power, and stigmatizing attitudes position service users as problems that service providers are charged with solving out of sight of community members not directly involved with SUD services (Hoffman et al., 2023). This study’s findings suggest that shortcomings in public investments distort help-seeking and provider relationships as providers suffer burnout competing for limited resources to meet the great need for SUD services amid the opioid epidemic. Staffing scarcity reproduces compassion fatigue and inhibits the individualized, dignified treatment that people with SUD-LE and providers perceive as vital in overcoming the loss of social identity and social harm that accompanies SUD stigma. Photovoice offered a compelling research method for both illustrating the influence of power differentials among roles related to SUD services and creating opportunities for people in varying roles to see the opioid epidemic from others’ perspectives.

This study implies that common interests and priorities across roles cannot be assumed in community-engaged SUD research based only on shared areas of residence, given how social harms perpetuated by relational stigma disenfranchise people with SUD-LE and degrade public investments in SUD service provision (Addison et al., 2023). PAR entails building consensus among participants on study design, conduct, and interpretation, which is difficult in the face of conflicting roles in SUD service delivery and funding. A first step in fostering PAR team development is collecting data from researchers, advisory board members, and research participants on not only their sociodemographic characteristics (e.g., race/ethnicity, gender, age) but also their LE of SUD service use and role related to service provision. Early reporting to PAR teams on their composition across roles and other differences followed by strategic recruitment of people in less-represented roles could mitigate the tokenism of including people with SUD-LE only in the recruitment and community outreach stages of research projects. Participatory SUD research teams might be strengthened with processes for members in non-PSS roles to explore disclosure of SUD-LE, against ableist professional cultures where it is perceived as a liability (King et al., 2020). To sustain consensus on research questions, methods, and analysis across roles, participatory SUD research teams could incorporate opportunities for continual communication, support, and reflexivity around team dynamics, emotions, logistics, mutual learning, expectations, and adjustments to project involvement around changing personal and professional demands (Arnold et al., 2022; Egid et al., 2021; Greer et al., 2018; Steketee et al., 2020; Williams et al., 2020). Perspective-taking across roles through narrative PAR methods like Photovoice may help address stigmatization, role strain, and tokenization of SUD-LE so they are not replicated in community-engaged SUD research design and team composition.

4.1. Limitations and next steps

The study’s transferability is limited by the HCS research context, which provided structured support for the nonprobability sampling recruitment of motivated Photovoice team members involved with SUD services as providers and consumers via funded local coalitions and partner agencies with the expressed goal of reducing opioid overdose deaths. The racial and ethnic composition of Photovoice participant groups was less diverse than the broader population of the 67 communities enrolled in the HCS during the study period (Larochelle et al., 2021). Additional research accounting for SUD service user, provider, and policymaker/researcher roles in practice contexts with greater racial and ethnic diversity is needed to further investigate the influence of various intersections of research team member and participant identities and contexts on study design and interpretation. This research could integrate Photovoice and other storytelling PAR methods into regular team meetings and data collection as reflection and accountability check-ins.

4.2. Conclusion

This study affirms other literature indicating the utility of critical reflections around Photovoice projects in facilitating explicit and inclusive discussions of research team members’ assumptions, understandings, and differences in power and capacities related to their intersectional positionalities and roles (Arnold et al., 2022; Cioffi et al., 2023; Egid et al., 2021; Evans-Agnew et al., 2022; Jones et al., 2023). In community-engaged SUD research, the construct of “communities” and who can authentically represent them requires careful characterization beyond residency status and continual reflexivity regarding who is at the table, who is missing or muted, and why. For stigmatized service users and overtaxed service providers, optimistic presumptions that communities are research assets and sources of belonging may ring hollow in the face of poverty, disinvestment, exclusion, racism, and discrimination. Therefore, it is incumbent on community-engaged SUD researchers to afford participants mechanisms to define their communities and their respective roles in them (Frederick et al., 2017).