Suicide attempt survivors’ recovery-related agency in the relational context of services: a qualitative analysis

Abstract

Background

Suicidal persons’ contacts with services present a key opportunity for suicide prevention. However, interventions by services are not always effective. A deeper understanding of suicidal service users’ agency and its implications may facilitate the provision of meaningful responses to help-seeking during suicidal crises. This abductive study explores the recovery-related agency of suicide attempt survivors and the perceived role of interactions with services in facilitating or hindering it.

Methods

Fourteen Finnish suicide attempt survivors were interviewed in-depth on their experiences of interacting with services during a recent suicidal episode. An operationalization of recovery-related agency as the expressed ability to take (mental or physical) action in a direction perceived as aiding recovery from suicidality (i.e., the coupling of recovery-related intentionality and power) was used to explore transcribed interviews through directed content analysis. Data were further categorized based on whether the service context was perceived as helpful or unhelpful to recovery efforts.

Results

All participants expressed both agency and non-agency in relation to their recovery process. The relational context provided by services was presented as highly relevant for the achievement and sustainability of recovery-related agency as well as for participants’ experience of safety in instances when agency was lacking. The results are presented as a typology of recovery-related agency in its perceived relational context, with the categories of sustained agency, strained agency, contained non-agency and uncontained non-agency.

Conclusions

The concept of agency helped capture important aspects of suicidal individuals’ recovery-related efforts and the role of services in facilitating or hindering them. The findings illuminate the value of viewing suicidal service users as agents of their own recovery process as well as the potential costs of ignoring this perspective in service delivery and design.

Background

Suicidal persons’ contacts with services present a key opportunity for suicide prevention. In Finland, 66% of those who died by suicide during 2016–2018 had visited healthcare within a month of their death, 46% within the week and 21% on the day of their death [1]. Yet finding meaningful ways to respond to service users in suicidal crises presents an ongoing challenge for health services internationally [2–4]. Evidence-based models of suicide prevention are based on an acknowledgement of the importance of a collaborative alliance between service users and professionals [4, 5]. However, healthcare service design and provision too often relies on practices that ignore or minimize the basic interpersonal aspects of care and thus fail to capitalize on their potential [2, 3]. The relative scarcity of research on suicide prevention practices from the perspective of service users and/or as an interpersonal process may contribute to this challenge.

While models of healthcare interactions that emphasise the role of service users’ subjectivity in shaping their process and outcome have emerged in the last decades, the influence of these ideas on healthcare responses to suicidality remains limited [6–8]. This is perhaps due to the dominance of medicine’s perspectives in both suicide research and healthcare systems; while applying the logic of medicine to address suicidal behavior arguably has many benefits, the medical lens is not optimally suited for considering or capitalising on either service users’ subjectivity or the interpersonal aspects of service interactions (for a discussion on these issues, see [9].

While the perspectives of medicine persist in guiding healthcare responses to suicidality, alternative approaches to conceptualizing and optimizing the role of services have also emerged or been adopted in the field of mental health. These include the self-determination theory (SDT) [10], a theory of human motivation and behavior that is “centrally concerned with the social conditions that facilitate or hinder human flourishing”, p. 3. SDT focuses on individual motivations and the satisfaction of basic needs. It highlights how social contexts shape these motivations and contribute to the effectiveness of health-promoting interventions. Empirical findings have supported SDT’s claim that the provision of effective need support predicts treatment engagement and outcomes in healthcare [10, 11] and psychotherapy [12].

Other models of service-assisted change that explicitly acknowledge the role of service users’ subjectivity and intentionality include the contextual model of psychotherapy [13] and a variety of recovery-oriented models for mental health services, e.g [14]. These models construe service users as active meaning makers who choose and use different aspects of the help available to them in creative and often unexpected ways, leading to outcomes that reflect the unique relational process of each therapy [15–17] or, more broadly, each process of recovery [18, 19]. While these theories do not explicitly use the concept of agency, they imply its usefulness for understanding how individuals in suicidal crises use services and, in turn, how services may be of better use to these individuals. Thus, we set out to explore expressions of agency in suicide attempt survivors’ accounts of their interactions with services during a suicidal crisis. For this purpose, we will propose a definition of recovery-related agency. We begin by defining what is meant by recovery in the current context.

Recovery as an idiosyncratic process

Recovery is a concept used widely in the medical, health and psychological sciences. Medicine defines recovery as the reduction of clinical symptoms below a threshold set by medical experts. However, models conceptualising personal recovery as a process of strengthening experiences of agency, hope and meaning irrespective of mental health struggles have begun to gain prominence also in the field of suicidology. Recently, Sokol et al. [18] presented a theoretical model of recovery from a suicidal episode based on a literature review, and Ropaj et al. [19] a Delphi consensus on what recovery from suicidal behavior entails from the perspective of those with lived experience. Both studies describe the recovery process as unique to each individual and emphasize the importance of service users being able to define recovery for themselves rather than being pressured to meet standards set by services. Drawing on these studies, we define recovery from a suicidal episode as a transformative process bringing about life-affirming change. This definition gives center stage to participants’ own understanding of a process that would lead them to perceive life as worth living and empower them to keep safe even when suicidal urges resurface or persist.

A definition of recovery-related agency

Agency is a concept used in all fields of science concerned with humans as intentional beings (e.g., philosophy, social sciences, psychology, and neuroscience). Its specific definitions and their philosophical underpinnings vary widely across scientific contexts, and no single definition of agency can meaningfully be proposed for more than context-specific purposes [20]. Our definition is informed by previous conceptualizations and discussions of agency in the context of pursuing (therapeutic) change [15, 16, 21–23], and aims to summarize the aspects of agency most relevant in the current context.

For purposes of the current study, we defined agency as having four essential attributes. The first is intentionality. Although the agent’s intentions may be more or less clearly formed, agentic action is necessarily guided by both reasons and goals, i.e., an idea of why a specific action should be taken or actively avoided, e.g [20, 21, 24]. The second is power, i.e., an agent must have the (potential) power to affect other entities, although this potential may not be realized in all (or any) of the ways intended, e.g [25, 26]. The exertion of such power may be any action (including an intentional omission of action), either mental or physical, with the potential to move one nearer to achieving one’s goal [20]. The third is an object, i.e., the concept describes a relationship between a subject with intentionality and power and an object of the intentionality that is (potentially) affected by the exercise of power. Again, this object may take the form of a mental, social, or physical “thing”, i.e., agentic action can be directed at objects in one’s own mind, in the social environment, or in the physical world. The object-related nature of agency is typically implicit in its definitions, but it is relevant for the current context and thus included as an attribute here. Fourth, agency is necessarily bounded, i.e., agency does not imply or require omnipotence, e.g [21, 26].

In summary, we define service user agency in relation to recovery from a suicidal episode as having (some) intentionality and (some) power in bringing about a transformative process resulting in life-affirming change. The aim of this abductive study is to explore, describe and interpret expressions of recovery-related agency in suicide attempt survivors’ accounts of their interactions with services during their suicidal crisis.

Methods

This study is part of a broader qualitative research project exploring suicide attempt survivors’ experiences of their interactions with health and crisis services during their recent suicidal crisis [9]. The project primarily applied a constructivist-interpretivist framework [9, 27] for an explorative study in a naturalistic setting. Our primary data for analysis consist of transcribed research interviews focusing on the participants’ service experiences. We also had access to the participants’ narratives of their index suicide attempt as documented in the Attempted Suicide Short Intervention Program (ASSIP) [28], in which all had participated.

The previous two studies exploring the same data have focused on describing the participants’ experiences of helpful and hindering aspects of healthcare services [29] and the subjective impact of ASSIP [30]. In this study, we chose an abductive approach to pursue a dialogue between theory and empirical data; the former suggesting that the concept of agency would provide a useful lens for interpreting service users’ accounts of service interactions, and the latter consisting of said accounts.

Participants

Participants were recruited through the non-governmental organisation (NGO) MIELI Suicide Prevention Center (MIELI), where they had received ASSIP in relation to a recent suicide attempt (i.e., the index attempt). All 104 eligible clients (i.e., all those over age 18 and residing in the Helsinki Metropolitan Area) entering ASSIP during the study period were invited to participate, and 14 chose to do so. Participant characteristics are presented in Table 1. The self-selection of the participants has been discussed in length elsewhere [9]; while these participants are representative of ASSIP completers in many respects, clients with somewhat better baseline and/or current functioning are likely over-represented in the sample.

Table 1.

Participant characteristics

	n	%
Registered sex
Male	7	50
Female	7	50
Age
18–29	5	36
30–45	4	29
46–59	3	21
60+	2	14
Current occupation
Employed	7	50
Student	3	21
Pensioner	2	14
Unemployed	2	14
Previous suicide attempts (before the index attempt)
Yes	8	57
No	6	43
Services used during current episode
ASSIP	14	100
Emergency services	14	100
Outpatient psychiatric care	12	86
Inpatient psychiatric care	4	29

Services

Here ”services” refers to the variety of service providers reported by participants as intervening in their crisis or considered by participants as potential sources of help. These providers included public and private healthcare providers (e.g., psychiatric inpatient and outpatient services, emergency services, occupational or student health services) and NGOs (all participants had used MIELI services, and some had experience of other NGOs).

Study interviews

Each participant took part in one semi-structured research interview conducted by the first author, a psychologist experienced at working with suicidal clients in healthcare but not involved in the participants’ care in any way. Semi-structured interviews were used because they allow for the elicitation of the participants’ own meaning-making on the broad topics introduced by the interview topic guide [31]. The interviewer’s positioning and its role in the co-construction of the interview data has been discussed in-depth elsewhere [9].

The study interviews took place at the MIELI Suicide Prevention Center (where participants had also received ASSIP) 4–10 weeks after the participants’ last ASSIP session and 3–6 months after their index suicide attempt. The interviews lasted 45–120 min and were video recorded. Participants were invited to narrate the experiences and interactions they found important in detail. In addition to the participants’ general experience of each service they had received, the interview topic guide explored, e.g., which aspects of these services participants perceived as helpful, unhelpful, or even hurtful, surprising elements, suggestions for improvement, and participants’ subjective assessment of whether each service had been helpful to them.

Data analysis

Directed content analysis [32] was used to explore expressions of recovery-related agency in the current data. In line with the definition presented in the introduction, we operationalized “recovery-related agency” as the participant’s expressed ability to take (mental or physical) action in a direction they perceived as aiding recovery, i.e., the coupling of recovery-related intentionality and power. “Recovery-related non-agency” was operationalized as the expressed inability to take such action or being confused as to what such a direction might be, i.e., their expressed lack of recovery-related intentionality and/or power. In our operationalization, we chose to take into account both descriptions of experiencing oneself as capable or incapable of taking a desired action (i.e., reported experiences of having recovery-related power) and reported behavioral expressions of this capability or lack thereof (i.e., reported exercises of recovery-related power) as expressions of participants’ agency. We use the term “expressed agency” to account for both forms of expression.

Because our research question concerned participants’ recovery-related agency in the context of interactions with services, we limited our analysis to excerpts in which participants described their agency specifically in relation to the context of services or professionals. We included both reports of actual interactions with professionals (e.g., an emergency room visit) and imagined (anticipated) interactions that participants described as relevant for their agency (e.g., described expectations of what an emergency room visit would be like based on stories from peers or the media or on one’s own previous experiences). Further, we focused on recovery-related agency and thus did not explore expressions of agency toward other goals (e.g., intentionality and power directed at taking one’s life).

In a previous analysis of the current data, we found that participants evaluated the helpfulness of each service in relation to how well that service recognized and responded to their personal recovery goals and tasks [29]. Thus, for the purposes of this research, the helpfulness of each service context was evaluated simply on whether participants reported perceiving it as aiding work on personal recovery task(s) and goal(s) they had found relevant in that specific context at that specific time (helpful relational contexts) or as unsupportive of or even detrimental to such pursuits (unhelpful relational contexts).

In our analysis, we first worked through the transcripts to identify excerpts in which participants discussed their recovery-related agency (i.e., expressed an ability or lack thereof to take action they considered meaningful for their recovery). Once we had identified all such excerpts, we proceeded to sort them into two categories based on our operationalization of recovery-related agency. After this, we re-categorized each excerpt based on the reported helpfulness of the current context, i.e., on whether the relational context was perceived as helpful or unhelpful in relation to the current recovery task/goal. The resulting two-by-two matrix thus represented a typology of expressed recovery-related agency in the perceived relational context of services.

In the following presentation of the results, data quotes have been translated from the original Finnish and edited for readability, while preserving the original meaning as closely as possible. Brackets in quotes indicate where text has been altered or added for clarity and an ellipsis indicates where text has been removed to shorten a quote.

Results

Participants’ accounts varied in the relative frequency of agentic and non-agentic expressions, but each account included examples of both. Within individual accounts, participants’ expressed agency varied from task to task (e.g., being able to take steps toward safety planning but at a loss for ways to try repairing a valued relationship) and from situation to situation (e.g., being unable to ask for help in one situation but able to do so in another). Recovery-related agency was thus described as both temporally and contextually fluid.

Personal recovery tasks as expressions of intentionality

Participants’ recovery-related intentionality was expressed in their personal recovery goals and tasks (see Table 2). These goals and tasks represented the participants’ understanding of what recovery meant for them (goals) and what actions would serve this end (tasks). The identification of relevant recovery tasks was, in itself, a commonly cited recovery task, meaning that participants’ recovery-related agency could be directed at clarifying intentions (goals and tasks) as well as at gaining power to act toward an existent intention.

Table 2.

Examples of participants’ recovery tasks and goals

Examples of personal recovery goals	Examples of personal recovery tasks
Ridding myself of the wish to die	Finding an apartment
Not being overwhelmed by negative feelings	Re-enrolling in school
Having/finding hope	Strengthening my sense of self-worth
(Re)discovering an interest in working or the ability to work	Opening up about difficult issues
Being able to meet the demands of daily life	Forming an understanding of the suicidal process
Having/finding a reason to stay alive	Learning to manage recurrent suicidal impulses without acting on them
Getting back to my own life	Asking for help when needed
	Finding the right medication
	Learning to talk about what’s bothering me
	Finding or returning to meaningful activities and/or relationships.
	Getting traumatic experiences “off my chest”
	Identifying personal recovery tasks
	Maintaining a reasonable rhythm of daily activities
	Getting out of the house and socializing

While recovery tasks were presented as actions necessary for recovery, they were also typically perceived as being beyond the participant’s independent power. Thus, interactions with services were presented as highly relevant for participants’ recovery-related agency.

Services as the context of recovery-related agency

All participants reported interactions with services that had enhanced their ability to identify, pursue and/or complete a recovery task, i.e., supported their recovery-related agency. Most also described interactions that had left them without support or even directly hampered their efforts. The context of a specific service or relationship was often presented as providing resources that had facilitated the pursuit or completion of some recovery tasks, while overlooking others. Thus, the same service or professional could be viewed as providing a helpful context at one moment (when support coincided with the recovery task currently perceived as relevant by the participant) and unhelpful at another moment (when support was not available for another recovery task emerging as relevant).

Professionals’ recognition of and support for participants’ pursuit of agentic power was appreciated and often also reported as leading to empowerment, whereas the lack of such recognition and support left participants feeling frustrated and often also powerless. Participants resented interactions in which their intentionality was overlooked, and often responded with either covert or direct forms of resignation or rebellion when they felt that these intentions were not recognized or respected. Perceived threats to the participants’ autonomy often led to a form of protective retreat, e.g., refusing an offered form of treatment, dropping out, or more subtly disengaging and deciding to withhold information from professionals.

Many participants reported that they had found it very difficult to communicate their struggles with agency, even though they wished for these struggles to be recognized and responded to. Participants associated this lack of power with both situational issues with trust (e.g., fearing an unwanted response from a specific professional) and more general difficulties in displaying vulnerability. One participant described an experience of being seen by professionals as either fully powerless or all-powerful and their lacking in the power required to correct the latter assumption:

I feel that either it is assumed that you are superhuman, like you can do everything … or then the opposite is assumed, like you can’t do anything … it’s a bit annoying because you don’t really know how you should behave, whether you should behave the way they expect or whether you should behave in a completely different way, so it’s difficult… [I’ve noticed that] it’s much easier to show that you’re stronger [than they think], harder to show that you’re weaker I guess.

Participants’ recovery-related agency was thus presented as multifaceted and in complex interplay with the relational context provided by services. Next, we present our categorization of this interplay.

Recovery-related agency and relational context

In the participants’ accounts, both agency and non-agency was reported in both helpful and unhelpful contexts. Thus, we present our results as a two-by-two matrix of recovery-related agency and relational context. The four categories in the matrix are labeled sustained agency, strained agency, contained non-agency and uncontained non-agency (see Table 3).

Table 3.

Recovery-related agency in its perceived relational context

	Context perceived as helpful	Context perceived as unhelpful
Expressed agency
	SUSTAINED AGENCY	STRAINED AGENCY
	The context is experienced as helpful, and the participant is able to work on a recovery task or complete it. Recovery-related action is possible and the accompanying emotional tone is positive.	The context is experienced as unhelpful, but the participant takes it upon him- or herself to work on a recovery task or complete it. Recovery-related action is possible, but the accompanying emotional tone is negative.
Expressed non-agency
	CONTAINED NON-AGENCY	UNCONTAINED NON-AGENCY
	The context is experienced as helpful, but the participant is not able to work on a recovery task. Recovery-related action is not possible, but the accompanying emotional tone is positive.	The context is experienced as unhelpful, and the participant is not able to work on a recovery task. Recovery-related action is not possible, and the accompanying emotional tone is negative.

Sustained agency

All participants reported on instances where they had been able to identify and engage in meaningful recovery tasks and goals and felt supported in this work, experiencing emotional tones of, e.g., confidence, safety, pride, and hopefulness. In these instances, agency was achieved through or nurtured by joint efforts, support received and/or acquisition of new resources. These were labeled as cases of “sustained agency”, as the accumulative effect on recovery-related resources (intentionality and power) was implied to be positive.

Sustained agency was often reported in relation to ASSIP’s facilitation of thorough cognitive and emotional exploration of the suicide attempt, a recovery task that participants generally agreed was important. One participant described it thus:

It can’t really be anything other than ASSIP that, well, it made me process [the suicide attempt], or it made me- it didn’t force me to do anything but got me do it and that means it really hit the spot.

In some cases of sustained agency, participants had entered into a relational context with agency (e.g., feeling prepared and able to delve into a difficult issue) and found the context to further support it. In other cases, non-agency was transformed into agency by a sustaining context. An example of the latter was given by a participant who entered ASSIP with a very fragmentary understanding of the suicide attempt and cited understanding the suicidal process as a critical recovery task:

I thought I would look crazy on the video, but it turned out that it was really clear what had led up to the suicide attempt … So it all kind of fell into place, because I hadn’t realized what [the suicide attempt] was all about.

Another common case of sustained agency entailed becoming able to complete or work on the recovery task of asking for help in a crisis because the relational context recognized the difficulty of this task and supported overcoming it. One participant gave an example:

I called [the outpatient clinic] a couple of times when I had questions … [it’s really difficult for me to] bother anybody, so it was good that I learned a little bit, I learned how to contact them … [it helped] that they showed me that they were worried about my wellbeing and we kind of rehearsed it every time that I should stay in contact so I wouldn’t be left on my own to think about things.

When participants perceived the relational context to support their autonomy, they reported being better able to utilize services in a way that effectively aided their recovery (e.g., by being honest about their situation and engaging in treatment). In many cases being explicitly allowed to regulate one’s engagement (e.g., to take breaks, to decide how much to disclose or to not be required to commit long-term) was cited as having made it easier to participate in treatment.

Strained agency

Some participants had been able to pursue or accomplish a recovery task in the perceived absence of any meaningful support or even in the face of straight-out rejection or sabotage of their efforts. In these episodes, agency had been achieved or maintained through lone efforts and self-exertion, resulting in the valued action toward recovery being accompanied by emotional tones of, e.g., struggle, depletion, resentment, disappointment and forced self-reliance. These were labeled as cases of “strained agency”, as the accumulative effect of such agentic efforts on personal recovery-related resources was implied to be negative.

Strained agency efforts were often cases in which participants had been rejected when reaching for help with a recovery task but managed nevertheless to hold on to their intention and continue pursuing the desired outcome, sometimes in explicit defiance of the setback. A young participant described their experience of help-seeking during conscript military service:

I explained [to the army doctor] that I hadn’t been doing so well [since elementary school] and I was depressed … he was really accusatory or- he let me talk, and he asked me some questions, and then I said ‘hey, I want concrete help, I don’t have to be discharged, but I would like some help with this’, so he thought for a moment and was just like ‘this should be all cleared up with this talk now’, and I told him ‘no, it’s not okay’, that I would really like to get help, or be told that I will get help, and then he looked really angry and started to tap on the computer saying that I would be discharged, and then he commented like ‘is it fair to the other guys’, that I’m just weaker than the others, how will the others react, if I leave, everyone else would start leaving too, stuff like that, just like you’d expect from an army doctor, and then he- well, it was pretty unpleasant to hear, but I thought I’d turn it into my strength, just to show him I’d go and seek help after I was discharged, so that was useful about it at least.

In another illustrative case of strained agency, a participant described solving their loss of autonomy during an inpatient stay by leaving, contrary to the professionals’ recommendation, in order to solve a recovery task they found urgent but which the professionals were not responsive to:

One of my problems was that I was losing my apartment and had to find a new one. I told [the hospital staff] about it but they were like ‘but you can’t go out’, so I couldn’t get a new apartment or do anything to take care of those things, and then when I asked for help they didn’t help me with it either, so it was really frustrating. I had like a week to find a new apartment, so it was a terrible stress … they wanted me to stay [in hospital] longer but I didn’t see the point because I couldn’t get anything done there.

The same participant described a struggle to be heard with their recovery needs, expressing powerlessness in relation to certain recovery tasks and a wish for support in working towards them:

I feel like something could have been done about [my problems with school and family], but I needed help with it … the doctor in the ward, they just told me that stuff is easy to fix, like it’s not a good enough reason [to feel suicidal] … and just last week I saw my [outpatient] psychologist, and they were like ‘you can just pick up the phone and call and that will solve it’, but they didn’t understand how difficult it is for me to call … [I would need] someone to do these things with me because I can’t do them on my own.

This participant described persistent strained agency efforts to acquire the needed support:

I’ve tried to make [the psychologist] understand that a phone call to the school, for example, it’s such a small thing to them, but to me it’s a really big thing. They haven’t gotten it yet, but maybe someday.

While not yet giving up on the intention and effort to form a collaborative alliance with the professional, this participant also described a sense of hopelessness (“I feel like my treatment is kind of a dead-end right now.”). Similar undertones of disillusionment and emerging hopelessness were typical in cases of strained agency.

Contained non-agency

In some episodes, participants described remaining incapable of recovery-related action even in a context experienced as helpful. In these cases, participants reported their experience of confusion, helplessness, or hesitation being accompanied by a sense of being supported, resulting in a more tolerable emotional state. These were labeled cases of “contained non-agency”, because the helpful context was perceived as providing protection from the most harmful effects of the non-agentic state.

One participant described their experience of being supported in a state of confusion and powerlessness after a devastating loss, and the feelings of safety this brought:

It was such a relief when I came [to the outpatient clinic], I was not very fit for work and then the doctor was like ‘okay, let’s take proper sick leave and defuse this situation’ … when you’re in these healthcare situations or talk about these difficult things you’re in a vulnerable position, so the fact that someone takes the initiative like ‘okay, let’s do this’, it’s so valuable, like you get to experience that you get a little control over your life when you may not really be in control of yourself … they have handled it really well because my anxiety is specifically related to worries about the future, about whether I’ll be left with nothing to support me … it has been really effective how they’ve engaged with me and assured me that help is available.

In cases of contained non-agency, participants often communicated a sense of relief, rest and/or hopefulness despite experiencing themselves as powerless and/or confused. One participant described the meaning of being provided with an emergency team during a vulnerable period:

The positive thing about it was the idea, that okay, if this person can’t get a permanent healthcare contact right now because it’s full everywhere, then that’s a really good idea, to find an unstable person a place they can visit and where someone checks on them that they’re still alive, that’s really great.

Another participant fondly remembered the nurses who had expressed worry and compassion during a vulnerable time:

[The nurses at the health center] were surprisingly supportive, they asked questions, and when I told them about the time I tried to get help but nothing came of it, they were like ‘oh, you got no help’, and then they asked me if I have any kind of plan for when I get back home, and when I talked to them about it they seemed worried and were like, ‘hey, can you manage these two nights, you’ll get a call then and an appointment will be booked’, and it was just like, for once they took it seriously.

Both participants reported this as an episode in which they had felt unsafe and lacking both the stable intentions and power needed to guide themselves towards recovery. While neither had received the immediate intensive help they wished for, both described the support they had received as making them feel better despite remaining unable to trust themselves to act in their own best interest, i.e., remaining non-agentic in relation to maintaining their safety. By alleviating some of the current emotional burden (driving suicidal behavior), the support they received had made it easier for them to stay safe even if it did not instill any immediate sense of their being able to control or regulate their behavior per se.

Uncontained non-agency

In cases of uncontained non-agency participants found themselves both unable to identify and/or act toward recovery tasks and lacking any meaningful support for so doing. These experiences were accompanied by emotional tones of, e.g., desperation, numbness, resignation, and anger. These were labeled as cases of “uncontained non-agency”, because the unhelpful context was described as leaving the participants without any protection from the most harmful effects of their non-agentic state.

The role of a supportive context in making disclosure possible was discussed by many participants. One participant, who cited “opening up” as a critical recovery task, described struggling with this task in meetings with psychiatric services:

Especially since you were not used to any kind of treatment, you were pretty closed off and feeling a bit of pressure and couldn’t really say anything about yourself, but then [the professionals] also didn’t know how to ask, so then many times the hour went on so that we were mostly just silent, and it was quite stressful. Somehow [the professionals] seemed to assume that I would be able to open up right from the start, even though the people were strangers and the whole context was completely unfamiliar to me, so that didn’t, that didn’t do any good at all for [my situation].

This participant eventually dropped out of these sessions. They reported that support provided by subsequent contacts with other services had empowered them to share personal experiences, resulting in both emotional relief and meaningful insights, i.e., experiences of sustained agency.

A high barrier to contacting services and asking for help was commonly associated with cases of uncontained non-agency. One participant described their dilemma when contacting outpatient services to make their next appointment had been left up to their own initiative:

That worries me a bit, because I would need some continuous support, but I don’t have it, not even a scheduled appointment for my outpatient clinic … I have the doctor’s number, so I can of course send them an SMS, but I don’t know if I dare to do that … I’m not very proactive about these things, so now I’m just waiting for them to maybe call me at some point … I’ve always felt like I’m a burden to others, that’s one of the biggest- that I don’t want to be a burden, that they have better things to do, that I wouldn’t dare [to bother them].

This participant also described a previous request for help (a strained agency effort toward recovery) being met by an unhelpful context (failure to make a promised referral), resulting in giving up (i.e., a shift to uncontained non-agency):

Either I wasn’t taken seriously or then [the referral] just wasn’t processed for some reason, because I was promised that they would make a referral to a psychiatric clinic, but I heard nothing, and then when I asked about it, they were just like ‘yeah, we’ll take care of this for you at some point’, but the referral never came, so I gave up on it.

In cases of uncontained non-agency, participants were sometimes able to hold onto recovery-related intentions despite experiencing a lack of power and support in acting towards realizing them. In these cases, they reported being frustrated with their sense of drifting toward recovery-hampering actions, such as being stuck at home, failing to maintain a reasonable daily rhythm, or failing to show up for appointments, and feeling their recovery goals slipping further away from them.

In some cases, experiencing the lack of both power and support was accompanied by a loss of recovery-related intentionality, with participants describing an emerging sense of indifference toward any recovery goals. These episodes could result in dangerous situations, including suicidal behavior. Some participants reported episodes in which their ambivalent intentionality had prohibited their spontaneous disclosure of suicidal thoughts or intentions. For two participants, a suicide attempt followed such an incident. When asked why they had not disclosed their suicidal intent (after voluntarily seeking help), one participant explained the effect that a direct question might have had on their ability to make the disclosure: “Honestly, at that point I thought I’d do it. But I wouldn’t have lied if the psychiatrist had asked. I was actually a bit surprised that they didn’t ask.”

Discussion

This abductive study explored, described and interpreted expressions of recovery-related agency as presented in suicide attempt survivors’ accounts of their interactions with services. Its findings illustrate how each response (or non-response) of services to help-seeking behaviors may significantly affect service users’ ability to achieve and sustain recovery-related agency, both directly (through alleviating or exacerbating emotional pain) and indirectly (by affecting willingness and capability to further engage with available support).

The current findings are congruent with previous studies that have found, e.g., that perceived staff attitudes critically affect help-seeking behaviors and that suicidal service users readily retreat from services that are not perceived as helpful [3, 4]. The participants’ accounts also illustrated the highly idiosynchratic, i.e., person-specific, nature of recovery goals and processes that is described by models of personal recovery from suicidality [18, 19]. However, the current study adds a valuable perspective by illuminating the agentic role of suicidal service users in using services to pursue recovery, previously scarcely described in the suicide literature.

“Coaching” behaviors as expressions of recovery-related agency

The participants’ agency was expressed both in setting and working toward goals for recovery and in “coaching” services to be more helpful in achieving these goals, i.e., in the various ways that the participants described having exerted their agency to adjust the help available to them, cf [15, 17]. We find Bandura’s [21] constructs of individual, proxy, and collective agency useful for understanding these coaching behaviors. Individual agency refers to the (limited) control individuals can directly exert on their circumstances. When goals are beyond individual agency (as the participants typically found their recovery goals to be) proxy agency and/or collective agency are needed to attain them [21].

Proxy agency is agentic effort directed at influencing others who may have the necessary resources, knowledge, or other means to act on one’s behalf [21]. The participants often described regulating their interactions with professionals in complex ways to secure the help they felt they needed to reach their recovery goals. In fact, some of their most tenacious strained agency efforts could be viewed as forms of proxy agency, as they were directed at influencing a gate-keeping professional in such a way that critical resources would become available. On the other hand, the participants’ (sometimes similarly strained) efforts at forming a collaborative alliance with professionals reflected an understanding that achieving recovery goals was a matter of interdependent effort rather than something another person could do for them, i.e., in these instances they seemed to pursue collective rather than proxy agency [21]. These findings illustrate how the participants’ recovery-related efforts took a variety of forms, and how services and professionals played a variety of roles in such efforts.

A self-determination theory perspective on recovery-related agency

The self-determination theory (SDT) [10] proposes that the satisfaction of three basic psychological needs (autonomy, relatedness, and competence) in service interactions predicts both service user engagement and outcome. Britton et al. [33] discuss the relationship between autonomy and treatment engagement as well as the role of relatedness and competence in the care of suicidal individuals, proposing the SDT as a framework for engagement-promoting care throughout services. Some recent theoretical works have also proposed SDT’s tenets as a foundation for effective practices in suicide prevention [34, 35], motivating a closer look at the SDT in relation to the current empirical findings.

In the SDT, autonomy is defined as voluntariness, self-endorsement and congruence with one’s authentic interests and values [10]. In the current study, autonomy support could be conceptualized as recognition of and respect for the participants’ intentionality. Such recognition and respect was reported as facilitating engagement in treatment processes, which then helped to further clarify recovery-related intentions and gain power. It also engendered feelings of being heard and thus alleviated emotional pain. When participants’ intentionality was not recognized or respected, they reported feeling, e.g., objectified, abandoned, and coerced. Perceived threats to participants’ autonomy often led to a shift from pursuing recovery-related goals to protective action, e.g., disengagement by dropping out or withholding information.

Relatedness is defined in the SDT as a sense of social connectedness, i.e., feeling cared for by and significant to others [10]. In the current data, support for relatedness was presented as relevant for recovery-related agency in at least two ways. First, participants expressed appreciation for service interactions in which their need for relatedness was met. When an empathetic other was available, participants described empowerment (sustained agency) or an experience of being safer even when they continued to feel powerless (contained non-agency). Second, participants often critiqued services for not providing enough support for recovery tasks pertaining to needs of relatedness [29, 30]. Such recovery tasks included forming a safe therapeutic alliance, resolving conflicts in significant relationships and (re-)connecting with loved ones or peers. Perceived lack of recognition and support for these tasks was reported as leaving participants powerless to progress towards many of their relationship-focused intentions.

The third basic psychological need posited in the SDT is competence, i.e., the experience of effectance and mastery. Appropriate task difficulty, positive feedback and the provision of structure are proposed as facilitating feelings of competence [10]. In the current study, support for competence can be conceptualized as support for gaining or maintaining agentic power. Interventions such as dividing goal work into small rehearsable tasks, noticing achievements and providing information were cited as empowering. Perceived lack of structure in interactions with services often left participants confused and powerless. In turn, structure-providing interventions (e.g., ASSIP’s program and tasks; professionals’ direct questions) were often explicitly cited as making recovery-related action possible [30].

Recovery-related agency as co-created

From the participants’ point of view, recovery-related intentionality and power were co-created moment-to-moment by the individual and their specific context. An understanding of recovery as co-created is in line with both the SDT [10, 12] and recovery models of suicidality [18, 19]. It is also reflected in calls to acknowledge that a collaborative alliance is critical for any helping efforts [2, 36]. However, the process of this co-creation has rarely been focused on in-depth in suicide research.

We argue that the current study contributes both conceptual tools and empirical findings useful for understanding the co-creation of recovery-related agency in suicidal individuals’ interactions with services. Importantly, the conceptualization of recovery-related intentionality and power as distinct but necessary components of recovery-related agency provides a useful perspective on assessing and dealing with obstacles to recovery-related action. It facilitates trouble-shooting when recovery-related action does not seem possible (distinguishing lack of recovery-related intentions from lack of power to act upon intentions). It also paves the way for identifying service users’ expressions of both recovery-related intentions and power as crucial resources for collaborative helping efforts. The concept of recovery tasks helps identify clients’ successes in acting toward their recovery-related intentions even during on-going crises, thereby facilitating encouragement of these efforts and feelings of competence.

Further, the empirical results demonstrate how offering interactions and resources that facilitate suicidal individuals’ recovery-related intentionality and power requires taking an interest in what recovery goals and tasks the person finds relevant, cf [19]. Giving primacy to the service user’s frame of reference does not exclude the possibility that services may also contribute in ways as yet unimagined by the service user. On the contrary, the provision of new perspectives and unexpected resources was perceived by service users as a valued aspect of helping efforts [30]. However, when professionals take the initiative in providing responses (e.g., suggestions, interventions, or formulations) that do not mesh with the service user’s current understanding of relevant recovery tasks, these should be made in a spirit of dialogue instead of being prescriptive, lest they be perceived as coercive or objectifying and thus motivating protective action rather than engagement, cf [10, 33].

Finally, acknowledging service user agency also entails appreciating how any offer of help is inevitably interpreted by and mediated through the subjectivity of the service user. Professionals’ well-intended actions do not automatically bring about the intended outcome. Thus, professionals and services need to accept the boundedness of their own helping-related power and remain curious about the real-time effects of their interactions with suicidal individuals. Consistent tracking of these effects should also inform corrective responses, i.e., attempts to repair ruptures in the collaborative alliance, cf [37].

Strengths and limitations

The main strength of this study is its focus on the contextuality of suicide attempt survivors’ agency in navigating recovery. Such explorations are rare in the suicide research literature, yet they serve to inform professionals in valuable ways. We explored participants’ expressions of recovery-related agency in relation to the range of services they perceived as (potentially) available to aid their recovery, broadening our scope beyond service-user agency in, e.g., psychotherapy alone. However, the existence or expression of service users’ recovery-related agency is not limited to the context of services, but is also in operation in their choosing to use or not to use any potentially available resources (e.g., the help of loved ones and peers), as well as in their independent recovery-related efforts [15, 22]. Further, the current analysis focused specifically on participants’ agency in relation to their self-identified recovery tasks, excluding expressions of agency or non-agency in other areas (such as agency related to suicidal intentions). Thus, a variety of relevant perspectives on suicidal individuals’ agency remain outside the scope of this article and await further research. Further, the location of this study in Finland means that its findings likely have most relevance in similar contexts (e.g., other Nordic countries). Exploring how recovery-related agency is expressed in very different service systems and/or cultural contexts would be of interest.

The current analysis is based on retrospective accounts, i.e., participants’ understandings of their service experiences as re-constructed at the time of the interview. While these narrative accounts provided a valuable window into the participants’ sense-making of their experiences, they were also inevitably affected by the interview context and the many heuristics known to affect human memory recall [38]. As all the participants had taken part in ASSIP, it is likely that their sense-making at the time of the interview was, in part, affected by the tenets of this specific intervention. Real-time data collection methods such as ecological momentary assessment [39] could help diminish some of these issues in future research by providing a longitudinal perspective on the evolving process of service users’ help-seeking behaviors and the interpretations of received and anticipated help associated with these behaviors.

Conclusions

The current findings illustrate both the fragility and the renewability of service users’ recovery-related agency during suicidal crises. It also highlights the possibilities that each service encounter presents to facilitate or hinder this agency. Specifically, service interactions that recognise service users’ recovery-related intentionality and power (or lack thereof) were described to facilitate service engagement, empowerment, and safety in suicidal crises. These findings underscore the need to implement suicide prevention practices that recognise and make use of service users’ agency rather than ignore or diminish it.

Abbreviations

SDT

Self-Determination Theory

ASSIP

Attempted Suicide Short Intervention Program

NGO

Non-governmental organisation

MIELI

MIELI Mental Health Finland

Author contributions

SGL and AL were responsible for the study design. SGL collected the data. The analytical and writing processes were led by SGL and reviewed and refined by AL, JV and JH.

Funding

This study was funded by grants from the Emil Aaltonen Foundation and the Alfred Kordelin Foundation. The funding bodies had had no role in the execution of the study.

Data availability

The original qualitative data are not available due to participants’ right to privacy.

Declarations

Ethics approval and consent to participate

This study received ethical approval from the Helsinki University Hospital Ethics Committee. All participants gave their written consent to use of their data for the purposes of this study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.