Nationally, an estimated 20% of police calls involve a person experiencing a mental health or substance use crisis (11). While estimates vary, it is thought that 25% – 50% of all people killed by police officers are experiencing this type of crisis (12). As jurisdictions seek better strategies to address the needs of people experiencing mental health or substance use crises while reducing potential harm and respecting autonomy, many are developing multidisciplinary alternative response programs that do not involve law enforcement. Alternative Response Programs differ from co-response models, where a behavioral health specialist may accompany police to respond to a call, and may also be called Alternative Crisis Response Programs, Alternative First Response Programs, or Alternative Emergency Response Programs.

Alternative response program composition and dispatch models vary widely. Some programs have dedicated phone referral lines, respond during limited hours, and are staffed by trained community responders who can work with clients to identify social needs and refer and transport to agencies (e.g. Mobile Assistance Community Responders of Oakland (13) and Atlanta’s Policing Alternatives and Diversion Initiative (14). Other programs are dispatched through 911, have trained paramedics or EMTs and/or behavioral health clinicians, and respond to calls 24/7 (e.g. Rochester’s Person in Crisis Team (15) and Denver’s Support Team Assisted Response (16)). Many programs also include peer support specialists, behavioral health workers who have lived experience with homelessness, mental illness, and/or substance use who provide peer-based counseling services and support (e.g. Durham’s Holistic Empathetic Assistance Response Team (17) and Arlington, Virginia’s Mobile Outreach Support Team (18)).

In each jurisdiction that uses 911-dispatched alternative response programs, the specific types of calls that these teams respond to may vary. For example, many programs respond to calls for “wellness checks” or reports of a person experiencing a non-violent mental health crisis. Programs with paramedic first responders may also respond to calls for drug overdose. In responding to these new types of calls (e.g., behavioral health crises, wellness checks) that are not for medical emergencies requiring transport to a hospital, dispatched first responders working in these non-law enforcement crisis response programs are exposed to an entirely different set of moral dilemmas and pressures than encountered in traditional emergency medical services. For example, many unhoused clients experiencing substance use disorder decline offers of shelter or treatment. Although these engagements may support clients emotionally and foster increased trust in the system of care, the client is left in the community without improvement in their material circumstances, which can be distressing for some first responders. These exposures can put alternative response workers at risk of moral injury, which can not only impact the alternative response workers, but impedes their ability to effectively serve clients and can create structural challenges for municipalities, as moral injury can be associated with burnout and attrition (19).

Given the importance of having a framework to better understand how these work experiences impact first responders to better guide leadership support, interventions, and policy, this article presents a conceptual model developed through a four-year participatory action research (PAR) study with paramedics, EMTs, and peer support specialists working in non-law enforcement 911-dispatched alternative response programs in a large urban setting. We will describe the PAR approach and explain the research process used to develop and validate the conceptual model. We will then present the Alternative Response Program Moral Injury and Burnout Model and share the qualitative data that elaborate, explain, and support the model. We will also identify hypothesized and piloted critical interventions to mitigate and prevent moral distress, injury, and burnout based on the conceptual model and qualitative data. Finally, we will discuss how the model fits with existing models of moral injury and extends our understanding of moral injury in high-risk occupations. In keeping with the PAR approach, this manuscript was co-developed and is co-authored by program collaborators and academic partners.

This study employed an approach to research called participatory action research (PAR). PAR is an approach to research that values academic, experiential, and technical knowledge, engaging in a partnered process where decisions are made collectively about what problems deserve focused attention and what solutions may serve to mitigate these problems (20). Cornwall and Jewkes have argued that the primary difference between traditional research and PAR “lies not in methods, but in the attitudes of researchers, which in turn determine how, by and for whom research is conceptualized and conducted” (20, p. 1667). Baum, MacDougall, and Smith describe PAR as “collective, self-reflective inquiry that researchers and participants undertake, so they can understand and improve upon the practices in which they participate and the situations in which they find themselves” (21). In this study, we employed a cyclic process of problem identification, participatory research, intervention development, implementation, and evaluation aiming to create meaningful social change while producing generalizable knowledge. Consequently, intervention development, pilot testing, and evaluation occurs within the framework of the PAR cyclical model rather than a standalone experimental intervention.

The setting for this study was an alternative response program in a large city in northern California. The alternative response program was launched in 2020 and is run through the Fire Department, in collaboration with local public health and emergency management departments. The study began in 2021, less than a year after the alternative response program began, and at first the driving question behind the collaboration was to try and answer the question “How can this alternative response team work to improve the health and wellbeing of our city?” As the academic research partners and response team leadership in the Fire Department met and began to discuss the experiences of first responders in the program, we collaboratively identified moral injury as one of the challenges likely facing program personnel as they engaged in their work. From 2022 – 2025, one of the primary foci of the study was on understanding and mitigating moral injury among members of the alternative response program.

While team composition has varied, during most of the study period, the team included a paramedic member of the Fire Department and a peer support specialist who respond together to each 911 call. Teams have also included a licensed behavioral health clinician (e.g. social worker), an EMT (also a member of the Fire Department), or a Homeless Outreach Team member who can perform housing assessments and referral placements. Teams are dynamically deployed, meaning that for the duration of a shift, team members are either responding to a call or are at a designated spot in their vehicle ready to respond. The program has a high call volume, responding to over 50,000 calls between November 2020 and March 2025. Most calls are for non-violent behavioral health crises, though the teams also respond to overdoses and other problems facing community members experiencing homelessness (22).

The long term and strong partnership between the academic researchers, leadership of the program, and members of the response teams, as well as the cyclical nature of the PAR approach allowed for “methodological pluralism” (21), with a highly diverse set of data sources that contributed to the development and refinement of the conceptual model we present in this article. The conceptual model was developed through pairing the PAR approach with analysis using a modified constructivist grounded theory approach (23) and then, as described by Cole, returning to PAR partners to “test, revise, enrich, refine, and discuss what was emerging from the data” (24). The continual return to PAR partners “enabled testing for resonance, rigor, and applicability for practice” as the conceptual model was honed (24). Table 1 describes each data source, the timing of data collection and analysis, and the purpose of each type of data. We will briefly elaborate on these data sources.

Throughout the study period, Fire Department leadership and academic partners met weekly or bi-weekly. In 2022 – 2023, we conducted six focus groups with EMTs and paramedics working in the program and one focus group with peer support specialists working in the program. During each focus group, we defined moral injury and asked participants whether they thought their work put them in circumstances where they might experience moral injury and whether there were any practices in their workplace that help protect them from moral injury. Each focus group was audio recorded, transcribed, and thematically analyzed by two researchers. Focus group participants made several suggestions for interventions that they thought might reduce their exposure to potentially morally injurious events and help to mitigate moral distress and injury. In this period, we also examined de-identified encounter log data from over 10,000 calls to the alternative response program, reviewing the narrative descriptions of each encounter. Narrative review was aimed at understanding the breadth of call experiences, including both explanations of what actually happened as well as latent content about the intensity of the call.

From the initial thematic analysis of focus group data and encounter log review, we began to construct meaning diagrammatically through memoing and to visually depict sources of moral injury identified through the focus groups. This initial model was shared with department leadership and, separately, with members of the response program in early 2023. During these report-backs, leadership and members asked questions, provided experiential reflections, and weighed in on proposed interventions to consider to mitigate moral distress and injury.

In light of the strong support from members for the focus on moral injury, we added a researcher with extensive expertise in moral injury to the academic team. With this strengthened partnership, we designed a pilot study to test a series of interventions that we – meaning the academic partners, department leadership, and response team members – hypothesized might help to understand, prevent, and reduce moral distress and injury among team members.

The initial focus group discussions suggested that experiences of peer support specialists in the programs were distinct from that of EMTs and paramedics, largely because the organizations they work for have different work cultures and policies. Their different training, professional mandates, skills, and experience mean that they play different roles on the response team (e.g. peer support specialists self-disclose as a tool to instill hope and resiliency when connecting to clients and centering their needs). In some cases, teams work very well together, recognizing the complementarity of these distinct roles and in other cases there is friction in the team because of a lack of shared perspective and appreciation for what each team member contributes to the team. In order to have a separate space where academic partners on the study could learn from the experiences of peer support specialists, we established a peer support specialist advisory board, which brought together peer specialists from multiple programs, including but not limited to the alternative response program, and met monthly throughout the study.

In keeping with the PAR approach, we developed a series of efforts aimed simultaneously at understanding and intervening to improve moral distress, injury, and burnout for all members of the alternative response program. We will discuss some of these pilot interventions in more detail below. In brief, in 2024, with the support of the Fire Department, we conducted a department-wide assessment of moral distress and injury with a survey. While survey data were not central to the development of the conceptual model, they were used to validate the model as the survey was one of the few fully anonymous methods used during the study. Both the quantitative and qualitative results of the study have been previously published (7, 10). We then held a series of alternative response team conversations with team members (i.e. EMTs, paramedics, peer support specialists, and Homeless Outreach Team members). We held these conversations so that each shift had five sessions over five weeks. The conversations were aimed at both providing shared language and understanding around the impact of their everyday witnessing (25), moral injury and stress and also giving space to grapple with the ethical aspects of their work such as understanding their client population, issues around gender and race, and the long-term impacts of adverse childhood experiences. We also piloted holding weekly meals, where teams could come back to the station during their shift and eat a meal together, sometimes cooked by their supervisor and sometimes cooked by a team that was not in service. After these pilot interventions had been going for some time, we held reflective focus groups where we asked team members what they thought of the peer conversations and team meals.

During the study we had several opportunities to report our findings back to leadership of the Fire Department and the peer support specialist agency. We also reported findings back to the alternative response program membership. We connected with leaders of alternative response programs throughout California to share what we were learning about moral injury and hear more about the contexts in which other programs operated in order to try and identify common factors across programs.

For dispatched alternative response programs, teams are mobilized in response to calls to 911. While typically, calls to 911 are for help for the calling party, for these alternative response programs, the caller is often reaching out about a behavior that the caller finds inappropriate or unwelcome and which they would like the responders to address. As one team member put it:

Another participant reflected:

In the best of moments, team members are able to collaborate together, embracing their role in educating community members as well as supporting clients:

Team members are directly responding to the demands of the community, oftentimes with inadequate resources to support the clients they are serving, thereby being immersed in ongoing moral dilemmas.

Another participant chimed in, saying:

There was a consistent challenge with insufficient resources, ranging from psychiatric hospital beds to safe shelter facilities. One paramedic compared the challenge of trying to do alternative response work without sufficient resources to what it would be like if an ambulance couldn’t rely on the hospital:

At times, team members also felt burdened by having to decide which clients to offer resources to and which clients would not receive these resources. While these decisions were guided by policy, the team members felt implicated by their role in making these offers. One paramedic shared:

Alternative response programs rely on being able to refer or transport clients to resources to meet their needs, like a sobering center or shelter. While the community members’ demands and insufficient resources can produce dilemmas, when these negatively impact clients or undermine the work that team members do to build trust with clients so that they can assist a person over the long term, this produces distress.

For example, when follow up care or connection to services did not happen reliably because of short staffing within referral partner agencies, this produced moral distress.

One paramedic leader reflected:

Many people described the difficulty of working with clients with serious mental illness and substance use problems who decline services. One peer specialist team member said that it was “frustrating” when a resource is available but a client does not want to use it: “It’s raining outside. Like, why don’t you want to go inside, like, we have a shelter, right?” But the peer specialist also understood that “we may be offering them shelter but not knowing that the last time they were in a shelter they were assaulted, or their things were taken from them, or they were treated poorly by a staff member” (formative focus group discussion). In response to the moral injury assessment, several members shared about how encountering these dilemmas routinely produced distress:

Another participant wrote about the stress of witnessing “the slow burn of homeless. Seeing the same people who don’t want help and some who do but can’t get the help needed due to the broken system. Everyone takes energy. And the ones that don’t want help but cross our paths the most seem to take the most energy” (moral injury assessment). One respondent asked, “By maintaining their autonomy, are we perpetuating the problem or allowing them free will to live as they see fit?” (moral injury assessment). A peer support specialist said, “they need to do better to equip people in our position for what we are encountering” (peer support specialist advisory board meetings).

There were many ways that working conditions contributed to moral distress, injury, and burnout. Compared to other forms of peer work, social work, and emergency medical services, the goals of alternative response programs and metrics of success are more contested. Many people are involved in developing these programs and different entities and agencies may have different goals for the program and different ways of measuring success. Goals may include shifting calls from law enforcement to the alternative response program, connecting clients to resources, reducing transportation to hospitals, or relieving strain on ambulances in order to shorten response time for medical emergencies. At times, these goals are incompatible with each other. For example, in order to relieve strain on ambulances, teams need to spend a limited amount of time on each call so they can get back into service and respond to the next behavioral health crisis call instead of having that call dispatched to an ambulance. However, in order to connect a client to a resource, the team may need to spend a lengthy time on a call, building rapport with the client and carefully ascertaining what resource the client may be interested in accepting.

One team member said, “I think part of the problem is the goal isn’t defined. Everyone’s goals on these things are different” (formative focus group discussion). Another team member described how responding to crises but not doing follow-up felt incompatible with their internalized sense of how to accomplish the work:

In a context where the work is morally fraught, having a shared vision for the work and explicit communication about the indicators for success is particularly useful.

In addition, the alternative response program space is dynamic and policies that determine which calls team members are dispatched to or co-response protocols with law enforcement changed several times during study. Many team members expressed feeling “blindsided” by these changes and noted that they received no additional training on how to respond to new types of calls. For example, one team member found out during a shift that they would now be responding to juvenile calls (peer support specialist advisory board meetings).

There was also a challenge of working in an interdisciplinary and interprofessional team, where team members bring very different perspectives and skills to their work and lack a cohesive vocabulary. One peer specialist, for example, described how tiring it was to try to interpret to paramedic partners the emotional needs of clients with psychotic symptoms: “They’re like, ‘how did you even hear that? They’re just speaking gibberish.’” And described how this can produce tension with coworkers: “After we’re like fighting for the rights of this person, then we’re on the rig trying to repair and trust each other as a team again” (peer support specialist advisory board meetings). Debriefing after calls was helpful for strengthening communication. Facilitated program debriefing was a standard practice when the program first started, but after several months facilitated debriefing ended. One paramedic recalled that when the team stopped debriefing “our problems would occur within the crews because the social workers, the peers, the therapists, the various people we worked with, we all have different communication styles. And on an ambulance, you have to make it work, and if you don’t, you can just sit and be quiet. But … in this environment, it was very different. It’s hard to drive all day long with two other people who don’t want to talk” (formative focus group discussion).

Some team members also described experiencing a hostile work environment when they were partnered with a team member who was aggressive or who used racist or biased language:

The team member went on to observe that one of the contributors to moral injury was actually the “combination” of having to be in this hostile professional environment while simultaneously working to “provide space and safety for our clients and the people we’re supporting” (peer support specialist advisory board meetings).

In addition to varying goals, changing protocols, being unable to meet clients’ needs, and challenges with other team members, additional work conditions that were identified as contributing to moral distress, moral injury, and burnout were lack of communication from leadership to team members; a sense of disconnect between decision-makers’ understanding of the work and team members’ experiences of the work; and understaffing, difficulty with taking leave, and mandatory overtime requirements. Team members also experienced physical strain from the work and sometimes injury.

While members may experience either burnout or moral injury, these experiences frequently co-occur. In our conceptual model, we display these with bidirectional arrows, noting that experiencing one of the issues can also increase the likelihood of experiencing the other.

Team members described how much more “mentally taxing” alternative response was compared to other programs they have worked in (emergency medical services for EMTs and paramedics, other non-crisis peer support programs for peer support specialists). One paramedic said, “once you’re downtown, you’re just in it, you’re just in the water, just bathing in this human tragedy” (formative focus group discussions). Another shared: “I think for all of us, we just have that want and that need to help people, and we feel like if we can’t successfully do that, we take it personal and it eats away at us” (formative focus group discussions). One paramedic reflected that each call working in alternative response feels like “we’re either going to get verbally abused that we’re there, or they want something we don’t have, or we want to sell them something they don’t want … Maybe I should just pick up ambulance shift … [to] refresh my morale” (reflective focus group discussions).

The emotional exhaustion and reduced sense of professional efficacy experienced by team members was described as very different from what people experienced in other programs they have worked in as EMTs, paramedics, or peer support specialists.

A peer specialist who had recently moved to a non-crisis peer response program after several years on the dispatched alternative response team reflected:

The persistent exposure to events where team members face ethical dilemmas, do not consistently have the resources to support clients, and do not feel they have the professional skills that are needed to meet the challenges they face set team members up for both moral injury and burnout. Once a person is already experiencing hallmark symptoms of burnout such as emotional exhaustion and reduced sense of professional efficacy, they may be more likely to have cognitions associated with moral injury. And once a person feels symptoms of moral injury like guilt, shame, and withdrawal, they are more likely to experience burnout.

An important aspect of our model is recognizing that alternative response teams respond to people who are already experiencing inequities and social marginalization. For example, many responders talked about getting to know clients who experienced severe sexual abuse as children, leading to homelessness, which led to substance use and mental health problems, leading to further victimization. Witnessing these situations and being unable to help produces moral distress. One paramedic described his devastation at the death of a client whom he had spent the last year trying to help: he felt as though the world had failed her.

In addition, the program operates within a system that itself produces and reinforces existing inequities. Team members described their distress at feeling complicit in the perpetration of these inequities. Team members identified groups of people more likely to be victimized on the street or in a shelter or for whom there were no resources available. For example, shelter policy required elderly clients to use bottom bunks and mobility impaired clients to use bottom bunks in a corner. This meant that elderly or disabled clients were less likely to be provided shelter than younger, non-disabled clients, leading to further inequity. Team members felt distressed by their role in perpetuating these inequities:

In addition, because this work can cause burnout, sometimes team members work with colleagues who are cynical and numb. When team members become cynical about their ability to connect clients to services, they may then spend less time on scene trying to find a resource that may benefit the client. For example, one peer specialist described their distress about an incident where they perceived their colleague to be too cynical and impatient with a client’s situation. The person they were dispatched to help was very upset about not being able to get a safe place to sleep. The peer reported that their colleague said the client was “too elevated” and “tried to clear the call … I was like, I’m not leaving. That’s crazy.” The peer specialist was able to find a shelter bed for the client that night where he could stay with his dog, but felt exhausted, distressed, and angry at not having support within the team (peer support specialist advisory board meetings).

Several concerns about fellow team members’ actions that could perpetuate inequities were described in the moral injury survey, which was the only data sources where people could share their experiences anonymously and privately. For example, one person wrote that they found it troubling when “I witnessed others in the field acting judgmentally toward patients and it affected the quality of care that was provided. Every patient, regardless of background or circumstances, is entitled to the same quality of care” (moral injury assessment). In our conceptual model, we recognize that adverse outcomes for team members can lead to adverse outcomes for clients. Witnessing these adverse outcomes for clients then further harms the first responders who are seeking to help them, perpetuating this negative feedback cycle.

Several people identified personal and organizational practices that helped them manage their distress that arose from these dilemmas during the focus groups, peer support specialist advisory board meetings, and the alternative response team peer conversations. These strategies included the following:

Through the PAR study, we worked with team members and leadership to identify interventions hypothesized to prevent or mitigate moral injury, some of which we were able to pilot during the study. During formative focus group discussions and report back meetings, members brainstormed dozens of interventions that they thought could help reduce moral distress and prevent moral injury and burnout. These ranged from the aspirational (e.g. a month of paid vacation every three years) to the practical (e.g. cross-training with other partner agencies). Table 3 maps examples of the interventions that were proposed by program members onto each of the pathways in the conceptual model to illustrate team members’ practice-informed suggestions to reduce moral distress and burnout and protect against moral injury.

During the course of the PAR, we piloted some of the practice-informed suggestions for intervention and worked with leadership to support implementation of additional strategies to reduce moral distress, moral injury, and burnout. One of our key strategies was to acknowledge the moral and ethical challenges of the work and to create space for team members to talk with each other about these dimensions. This built on an observation that during the formative focus groups, team members valued the opportunity to be in dialogue with one another:

The piloted interventions focused on creating both structured and unstructured spaces and times to build this routine acknowledgement. During the facilitated peer conversations, we introduced language and explanatory frameworks to help team members reflect on their experiences and then held space for dialogue. As one paramedic reflected, these facilitated sessions helped with “putting a name to what it was that we’ve been looking at and witnessing. It’s sort of like these are things we recognize, these are there” (reflective focus group discussion). Learning frameworks to better understand the client population and the impact of stressors and moral injury on first responders was “confirming and validating … it provides an actual name for the feelings you’ve had” (reflective focus group discussion). One team member shared, “it helps to feel more empowered in your job when you know more information, you have more power, you have more knowledge to be able to execute what it is that we’re supposed to be doing out there” (reflective focus group discussions).

These conversations then continued into unstructured spaces, including weekly meals that teams could take together on shift. The meals provided an opportunity to connect with colleagues, share experiences, exchange ideas, and decompress. Because of the conversations that happened in facilitated spaces, relationships between colleagues also changed. In one reflective focus group, a participant shared:

In addition to having new language and deeper connections, members really appreciated the opportunity to reflect on the moral dilemmas that they experienced through their work with a deliberate invitation to talk with colleagues about the ethical dimensions of what they witnessed. For example, one person brought up a client whom they had known for the years:

Another participant reflected, “Even though you know you’re kind of doing the right thing for that person and the community, it still doesn’t feel good … It’s complicated work. It is the most complicated job I’ve had” (reflective focus group discussion). These sessions were so positively received that Fire Department leadership institutionalized them for the alternative response teams by contracting with the employee assistance program to provide a therapist to facilitate biweekly peer conversations on an ongoing basis.

An additional thrust of the interventions was to improve communication internally within the program, externally with partners like shelters and law enforcement agencies, and externally with the community. One illustration of this strategy is around communication with law enforcement during times that they co-respond to a client. Conflicting or vague policies dictating how to respond to a person in crisis can put alternative response team members at risk of physical harm (e.g. police policy on disengagement aimed at de-escalation to mitigate risk of harm may mean that alternative response team members need to physically subdue a client who is acting violently if that person is unarmed). After an incident where team members were injured by a client while police stood by, program leadership met with police leadership to carefully review the case and create a shared document on how different first responders should engage with each other on scene, which they laminated and gave to all alternative response team members. One of the program leaders reflected that witnessing their team be injured in this incident had been a source of moral distress for them as a leader. They were able to turn that distress into an action they could take that would help reduce the chance that this situation would be repeated. Team members appreciated this responsive and proactive strategy and the communication they received about the leader’s actions.