The study was framed as exploratory. A qualitative approach was chosen as it is particularly suitable for the research objectives. Focus Groups (FG) were used as a data collection technique. Open questions were used to explore the meanings that the participants attributed (in their own words) to certain concepts, identifying in the process recurrent themes among the definitions and arguments that they used to explain their ideas (Morgan, 1996).

Based on a review of the literature, the research team developed a standardized protocol for FGs, which was subsequently approved by the board of directors of the Activa’t per la Salut Mental project. The protocol included information about the purpose of the study, the rights of the participants, and the benefits associated with participation.

The participants were people of legal age, not legally incapacitated (to sign their consent), users or ex-users of mental health services, and residents in Catalonia. In the recruitment, no distinction was made from the type of services. These could be both hospital and community settings, including: Psychiatric Hospitalization Units, Outpatient Mental Health Services, Community Rehabilitation Services, Residences, Social Clubs, etc.

For recruitment, 7 mental health services, collaborators of the Activa’t project, were contacted, as well as the people associated with the entities Federació Veus, Salut Mental Catalunya and ActivaMent Catalunya Associació. About 300 people were invited to participate in the FGs, 50 of whom responded. Finally, 36 of them attended the sessions that were recorded.

The participation in the study was voluntary. Those who were invited received a document explaining the objectives of the research, the data collection method and what would be required from them if they agreed to participate. Informed consent was obtained from all participants. A coding system was used to identify individuals and maintain anonymity.

The participants were selected using an intentional sampling strategy that took into account three demographic variables —age, gender and territory—, and the activism variable. The demographic variables were used for trying to have a certain equality of profiles among the people participating in all the FGs, but they have not been considered as categories of analysis in the configuration of the groups, nor in the elaboration of the protocol script.

Taking into account that this research was part of a larger study, the demographic variables of gender and age were considered in the sampling due to their relevance in relation to various aspects that affect mental health and recovery, for example: self-stigma and willingness to seek help (Eisenberg et al., 2009; Mackenzie et al., 2006); linkage to mental health services (Temkin-Greenyr & Clark, 1988); attitudes towards treatment (González et al., 2005); etc. The territory variable was considered because the differences between Barcelona and other Catalan towns in terms of mental health services may be relevant, which affects the resources that people use. Finally, the activism variable was considered due to its importance for empowerment and recovery processes (Adame & Knudson, 2007; Mancini, 2007). In this research, an activist has been considered as any person who defines himself/herself in this way and participates in the activities of an entity, social movement or collective that carries out activism in the first person in the field of mental health (Cazorla Palomo, 2018; Sampietro, 2016).

When creating the FGs, an equitable distribution in terms of gender, age and geographic location was sought, but the participants were divided based on the activism variable. Thus, FG1, FG2 and FG3 were made up of activists, while FG4 and FG5 were convened by service users without such experience. However, in practice, FG4 was finally considered a mixed group because two of its participants began attending the activities of a mental health user association in the period between recruitment and the day of the FG.

Likewise, in practice, parity of gender was not achieved in terms of the people who attended the FGs, resulting in twice as many men as women in the sample. In the FGs of activists, this difference was not as remarkable (10 women and 14 men), but only 2 women attended the non-activist FGs (see Table 1).

The data was collected from February to April 2017. The FGs were carried out in Barcelona and Vilanova i la Geltrú. The sessions lasted approximately 90 minutes, they were audiotaped and transcribed verbatim.

Transcribed data were analysed using ATLAS.ti 6.2 (Muhr, 1991). Data-driven coding was followed to answer the research question (Gibbs, 2009). The identification of the meanings attributed to recovery was based on the principles of Grounded Theory (Glaser & Strauss, 1967/2006). The open and flexible character of the Grounded Theory method allowed us “to create theoretical categories from the data and then analyse relationships between key categories” (Charmaz, 1990, p. 1162). Grounded theory is useful when research begins without a priori hypotheses, as it helps to minimize presuppositions and to build empirical knowledge grounded in the meanings of research participants. We specifically use Anselm Strauss and Juliet Corbin’s model of grounded theory (Corbin & Strauss, 1990; Strauss & Corbin, 1998), which acknowledges that empirical observation can never be totally free from theoretical influence. We drew upon constructivist grounded theory, which assumes that data and theories are constructed by the researcher as a result of his or her interactions with the field and its participants (Bryant, 2003; Charmaz, 2000, 2008).

Data analysis was performed in three stages. First, the transcription of each FG was read by two different researchers in order to identify extracts in which participants referred to recovery and what helped or hindered it. A third researcher was then involved in selecting the final pool of quotations. Her task was to choose the most intelligible option between two quotations expressing a similar view or idea, as well as to decide whether a quotation identified by just one of the first two researchers merited inclusion or not. In the second stage, we created a series of categories that reflected the themes that the participants used in their arguments to account for their experience. The two principal researchers compared their proposed categories in order to arrive at a consensus. The criterion for deciding the utility of each category was the theoretical saturation of codes, which was evaluated independently by each researcher (Charmaz, 2006; Glaser, & Strauss, 1967). This process led to the identification of eleven categories or emergent themes. Finally, researchers looked for relationships among quotations within each category.

Although the word “empowerment” only appeared in the activist FGs, the dimensions of the personal level of empowerment (World Health Organization - Regional Office for Europe, 2010) are factors that were highlighted as basic in all the FGs. Autonomy, self-management skills, the ability to make decisions, and overcoming self-stigma emerged as key variables in the recovery process and as prerequisites for well-being.

People defined their autonomy and self-determination in relation to possible external limitations, both from within the family environment and from mental health services:

Likewise, self-determination and autonomy are also linked to the mental health problem itself. For some people, it means being able to function in daily life without the constraints and limitations derived from the symptoms.

The participants highlighted that without the sense of one’s own existence nor a motivation to live, there is no possible recovery nor subjective quality of life.

A personal life project, however, does not need to reflect the social expectations. Some participants highlighted that recovery is not necessarily linked to productivity nor insertion in the labour market:

Accordingly, a lack of meaning in life would have a negative impact on a person’s mental health. The absence of such meaning is a barrier to recovery, and it could lead to a crisis or be experienced as a relapse.

In all the FGs, the participants highlighted the social-relational dimension of mental health. In these cases, people give a definition of recovery that is based on the human being’s need for interaction and connection with others.

From this perspective, any impairment on the ability to establish or maintain healthy social bonds is experienced as a relapse or a mental health problem.

For many people, what defines recovery is reversing the process that leads to isolation, regaining the ability to interact and bond.

Participants defined recovery as a process of self-knowledge. Some people highlighted the need to learn from experience and acknowledge one’s own limits as a condition of possibility for personal well-being, as well as a requirement to achieve and maintain a positive sense of one’s own life.

Participants also saw their crises as a learning experience, as an experience that is useful for their process of recovery, in the sense that this would allow them to be aware of what affects their well-being and quality of life. It is the idea of crises as a source of insights.

Some people define recovery as having a positive self-concept. Having wellness would basically mean feeling good about who I am and with what I do in life. Without a positive appraisal of oneself, there would be no emotional wellbeing.

Complementarily, other people described mental health crises as a loss of identity. For them, relapse feels like they cease to be themselves and they struggle to recognize themselves in their actions, thoughts, and emotions.

For these participants, recovery would be being themselves again, recognizing themselves in their own behaviours, thoughts, and feelings. Therefore the recovery process would suppose a reconstruction of the self.

For many people, recovery relies on the hope of having a better quality of life, with an optimistic view of the future. For these participants, recovery is not defined as a state but as a horizon to walk towards.

For some people, thinking of recovery as possible is a basic motivational element and a key facilitator to face their own process.

For these people, hope is a personal quality that can be cultivated. Their own life experiences can serve as reference and model to face the current situation.

Likewise, hope can also be cultivated from other recovery experiences, by knowing the life stories of those who have gone through a similar experience and who currently enjoy a full life.

In contrast to the previous topics, the remaining five themes that were identified in the analysis generated conflicting ideas, and discussion among the participants.

The issue of symptoms as an indicator to assess the recovery process itself is a topic that generated conflicting ideas. Some people saw the presence or absence of symptoms as the benchmark in this respect. For them, recovery means the absence of illness, whereas relapse is equated with being symptomatic.

When people think about mental health in absolute terms, that is to say, as the absence of symptoms, the fact that symptoms can reappear leads them to conclude that they can never recover.

Thus, some people who live with persistent symptoms think recovery as impossible and they see themselves as chronic patients.

It could be highlighted that the argument of recovery as impossible, with the "illness" defined as irreversible even though there are asymptomatic periods, only appeared in FG4 and FG5, the groups with participants without experience in contexts of mutual support and activism in mental health.

On the contrary, other people think on health as a continuum in which there can be more or less quantity, and not in absolute terms. When this idea is applied to the notion of recovery, participants argue that recovery is not a onetime action or an end state, but rather a permanent “journey”, a gradual process.

For these people, the key is not the presence of symptoms, but the extent of how these symptoms affect their quality of life and their ability to exercise autonomy and self-determination.

From this perspective, the appearance of symptoms or a relapse does not have the same dramatic meaning, as it is not seen as a failure nor the loss of everything that has been achieved.

By the same token, those who think about health as a process understand that, on the road to recovery, there may be stages or partial objectives that allow us to think of different levels or stages.

These stages would depend on the point or moment of the process in which the person is, a moment defined more from discomfort and suffering or more from well-being and quality of life.

The idea of normality generated a lot of conflicting opinions among the participants, and these can be grouped according to two definitions of what is normal. The first definition is normative and refers to what most people do. Recovering would mean being normal (being like most people in the own environment):

Accordingly, a relapse would mean going beyond the boundaries of what is considered normal in society, engaging in a behaviour that is not expected within a given social context.

The second definition of normality is built around what is usual or habitual for a particular individual. A relapse would be an interruption in that person’s ability to carry out daily activities.

From this conception, recovery means regaining the capacity to go on with their own daily life, regardless of whether they continue to have symptoms.

Finally, some people problematized the very idea of normality. In this case, there is no doubt that the behaviour itself is outside the norms or conventions, but it is questioned if these conventions are valid as a health criterion when evaluating behaviour.

How and who can measure recovery was another issue that generated conflicting opinions. Some people saw recovery as a subjective matter. For them, mental health is personal, and it would depend on the meaning that each person ascribes to it. Thus, each person must find out what recovery means for herself/himself, because there is not a single definition that fits everyone.

In these cases, recovery is synonymous with well-being. Some participants stated that only themselves can define when they feel good enough to consider themselves recovered or bad enough to realize they have relapsed. For these people, the experience of psychological well-being is the measure of recovery.

This argument has two sides. On the one hand, people may consider that their quality of life is good enough to see themselves as recovered and as having sufficient well-being, regardless of what other people think. Conversely, someone who still experiences high levels of suffering may refuse to accept that he or she has recovered, despite professional opinion of the contrary. The same occurs with the experience of relapse. Some participants considered that there are no external benchmarks of psychological suffering and only their own subjective experience can tell them when they have relapsed.

In contrast to the above, some participants used the notions of "lack of illness awareness" and "dangerousness" to argue against the possibility that people could adequately assess their own situation and take responsibility for their own recovery process. From this perspective, recovery should not be seen as a purely subjective matter, and the judgment of professionals should always be included in any assessment of a person’s mental health.

This topic is an intersection between the idea of normality, the value that people give to symptoms as a yardstick, and the debate on recovery understood as an idiosyncratic issue. Participants used the metaphor of reality as a place (where someone can be inside or outside) to explain their unusual perceptions and thoughts.

Accordingly, relapse is defined as a perception that goes beyond the margins of what is intersubjectively valid, a definition that is linked to a normative notion of reality.

Sensations and perceptions that do not fit within the framework of social normality would be located outside of reality. Recovery would be equated with the cessation of these experiences and the return to socially shared cognitive frameworks.

In contrast, there are people who question the very notion of reality and, especially, the value of this parameter as a criterion of health and illness.

Some participants saw recovery as a return journey to a time before their mental health crises. The idea here is that something lost (life, job, friends, etc.) can be regained. This is usually felt, specially, when the present is characterized by suffering or distress.

For people who think in this way, recovery and wellness might seem like impossible goals because they can never regain what has been lost.

Conversely, if the present is experienced as a time of wellness and people feel recovered, then the return to a previous time has the opposite meaning: it is a return to a time of crisis or suffering. This is experienced as a relapse and a backward step.

Finally, some participants were critical to the idea that you can turn the clock back. For them, time is irreversible, and believing otherwise only produces suffering and is a barrier to recovery.

From this perspective, recovery is a process of personal growth. Overcoming a mental health problem means having more tools and resources than before, and thus being better prepared for dealing with distress or suffering.

Although previous studies have explored the meaning of the concept of recovery from the perspective of mental health service users (Kidd et al., 2015; Mancini et al., 2005; Noiseux & Ricard, 2008; Piat et al., 2009), the present study is, to our knowledge, the first to examine this issue in a Spanish-speaking cultural context. A context in which recovery-oriented public policies are beginning to be implemented and coexist with care practices from a biomedical perspective. The study was also a key step in developing a new guide for implementing personalized plans for recovery and self-management of well-being, aimed at users of mental health services in Catalonia (Sampietro & GavaldàCastet, 2018).

On the contrary, although we intended to have a relative equity in the FGs demographic profiles, twice as many men as women attended. A difference that does not correspond to the use of mental health services in Catalonia, in which, although 56% of psychiatric admissions are men, in adult outpatient mental health services 6 out of 10 people served are women (Agència de Qualitat i Avaluació Sanitària de Catalunya, 2017). This is a limitation of the sample that could have influenced the collected meanings. Therefore, larger comparative studies of mental health service users are needed to examine the possible influence of demographic variables on how they understand and define recovery.