As a result of government policies and the National Services Framework ¹ the concept of ‘recovery’ has become an increasingly discussed topic within the mental health field and is an important concept for mental health service providers and policy makers. However, its definition has never been explicitly formulated. This is of concern because of its current influence on research and clinical practice. ^{Reference Silverstein and Bellack2}

The definition and measurement of recovery in mental health has been approached in differing ways. Some people have simply outlined recovery by singular factors such as symptoms, relapse rates and functioning. ^{Reference Pilling, Bebbington, Kuipers, Garety, Geddes and Martindale3,Reference Addington, Penn, Woods, Addington and Perkins4} Others have considered it to be a long-term process incorporating hope for the future, rebuilding self and rebuilding life. ^{Reference Pitt, Kilbride, Nothard, Welford and Morrison5} Davidson et al outlined recovery as two superseding models: one that incorporates recovery as symptomatic and the other that adopts a sense of well-being regardless of symptoms. ^{Reference Davidson, O'Connell, Tondora, Styron and Kangas6} Others have described recovery as a set of internal and external conditions. ^{Reference Jacobson and Greenley7} In light of this diversity, it is important to explore what the term recovery means in order to prevent misunderstanding and misapplication. Within the domain of psychosis, recovery has traditionally been regarded as predominantly about symptom alleviation. ^{Reference Silverstein and Bellack2} This definition has been widely adopted in clinical and randomised controlled trials (RCTs) of psychological therapies (e.g. cognitive-behavioural therapy) ^{Reference Kuipers, Garety, Fowler, Dunn, Bebbington and Freeman8-Reference Sensky, Turkington, Kingdon, Scott, Scott and Siddle10} and medication. ^{Reference Jackson, Trower, Reid, Smith, Hall and Townend11,Reference Gitlin, Nuechterlein, Kenneth, Subotnik, Ventura and Mintz12} As RCTs are used as the benchmark for effective treatments for psychosis in National Health Service (NHS) and Department of Health policies and guidelines, symptom alleviation is clearly an important factor in relation to defining recovery in current services. In addition, symptom alleviation is important to recovery as psychotic symptoms have been shown to cause much distress, debilitation and reduction in social functioning. ^{Reference Addington, Penn, Woods, Addington and Perkins4,Reference McGlashan, Zipursky, Perkins, Addington, Miller and Woods13}

Further support for symptom change as an important indicator of recovery is illustrated by the number of psychometrically reliable symptom outcome measures that have been developed. Outcome measures such as the Positive and Negative Syndrome Scale (PANSS), ^{Reference Peters, Myin-Germeys, Williams, Greenwood, Kuipers and Scott14} the Brief Psychiatric Rating Scale (BPRS) ^{Reference Kay, Fiszbein and Opler15} and the Psychotic Symptom Rating Scales (PSYRATS) ^{Reference Overall and Gorham16} are commonly utilised illustrating the dominance of symptom alleviation as an indicator of recovery. Other outcomes have also been frequently used. For example, relapse reduction has been used as a measure with regard to recovery. ^{Reference Haddock, McCarron, Tarrier and Faragher17} Relapse often includes assessment of reoccurring symptoms, rehospitalisation rates and remission times. ^{Reference Garety, Fowler, Freeman, Bebbington, Dunn and Kuipers18} Assessing relapse allows clinicians to identify periods of symptom stabilisation and thus periods of recovery. There has also been a longstanding interest in assessing quality of life as an indicator of recovery in psychosis. ^{Reference Harrow, Grossman, Jobe and Herbener19} Quality of life integrates objective and subjective indicators, a broad range of life domains and individual values. ^{Reference Malla, Norman, McLean and McIntosh20} Studies have predominantly examined aspects such as independent living and employment, ^{Reference Felce and Perry21} e.g. using the Quality of Life Scale (QLS). ^{Reference Drake, McHugo, Xie, Fox, Packard and Helmstetter22} Similarly, functioning has been used as an assessor of recovery, e.g. Global Assessment of Functioning (GAF). ^{Reference Heinrich, Hanlon and Carpenter23} Quality of life and functioning are often assessed along with symptom outcome measures in RCTs and treatment studies. ²⁴

Collectively, these approaches to recovery have defined the term as a multidimensional outcome, although its dimensions are usually quantified and assessed individually. Although this approach has given great insight into recovery and impact of treatment, it does not represent the multidimensional and complex picture that has been highlighted by some service users. For example, some research has highlighted that recovery has also been considered as a process, which incorporates a range of personal and social factors. ^{Reference Lewis, Tarrier, Haddock, Bentall, Kinderman and Kingdon25}

A meta-analytic study of recovery research from a service-user perspective identified five clusters of important factors different to those already outlined. ^{Reference Allot, Loganathan and Fulford26} First, personal and self-empowering processes were highlighted as significant. This included taking control of one's life and developing a positive self-identity. ^{Reference Wilken27} Second, recovery was identified as a motivational process, incorporating items such as generating hope and being active in one's own recovery. ^{Reference Chadwick28} Third, developing one's own competences, including making sense of mental distress and seeking knowledge was highlighted. ^{Reference Pitt, Kilbride, Nothard, Welford and Morrison5} Fourth, making changes in the direction of social and community participation was highlighted as important, such as accessing social support, including support from other service users. ^{Reference Dinniss, Roberts, Hubbard, Hounsell and Webb29} Finally, incorporating resources from the environment was deemed imperative, including accessing mental health services and voluntary support services. Further factors have also been identified from service-user research. One dominating factor is the need for individuals to overcome their experiences of psychiatric treatment and medication. This may involve aspects such as overcoming social isolation, stigma and discrimination. ^{Reference Pitt, Kilbride, Nothard, Welford and Morrison5} Furthermore, the effectiveness of medication, appropriateness to the individual and side-effects have been shown to be important. ^{Reference Silverstein and Bellack2,Reference Lewis, Tarrier, Haddock, Bentall, Kinderman and Kingdon25,Reference Liberman and Kopelowickz30} Spirituality and religion have also been recognised as relevant, as a coping mechanism or an explanation for an individual's experience. ^{Reference Liberman and Kopelowickz30}

Collectively, the term recovery has been defined in a diverse manner dependant on an individual's perspective. A vast number of approaches have been outlined, but there is still uncertainty about what factors contribute to the construct of recovery, and whether recovery is related to symptom change or not. Studies to date have not taken an all encompassing approach to recovery, and have constrained their focus to either symptom alleviation or the idiosyncratic recovery process and its impact on life. This current study aims to alleviate this uncertainty by adopting an inclusive approach in further scrutinising what factors are important to this multidimensional concept. This study explores people's subjective experiences of recovery and, in particular, the relationship between recovery and symptoms. Interpretative phenomenological analysis was used to elicit data from participants because of its person-focused nature. Interpretative phenomenological analysis is concerned with the individual's understanding and interpretation of their own personal experiences. ^{Reference Forchuck, Jewell, Tweedell and Steinnagel31} It assumes that people are self-interpreting beings; therefore the researcher attempts to interpret the participant's experiences from the participant's perspective. A semi-structured interview schedule was utilised in order to elicit individual's personal views on recovery from psychosis. This facilitated the discussion of specific recovery topics such as symptom change and issues that affect recovery, but also allowed for flexibility in people's personal experiences; a method highly suited to interpretative phenomenological analysis.

Method

Results

Eight people were interviewed (six males and two females), with an age range of 24–35, and all had experiences of delusions and/or hallucinations within the past 12 months. Six were recruited from early intervention services and two were from community mental health teams. Overall, 132 themes were generated from the interviews by the authors. Overlapping and repetitious themes were identified and, where it was agreed by consensus, these were removed. The remaining themes were then further discussed, resulting in some being identified as reflecting the same concepts as others. This allowed a further fine-tuning, resulting in 50 clear themes that were representative of the expansive concourse. The final 50 themes broadly covered 8 areas of recovery: symptoms, emotional aspects, the self, behaviour, services and support, coping mechanisms, social functioning and occupational aspects. From these broad eight themes a logical grouping of four superordinate themes emerged.

The four superordinate themes were described as: ‘impacts on mental health’, ‘self-change and adaptation’, ‘social redefinition’ and ‘individualised coping mechanisms’. These themes were underpinned by change, highlighting that recovery is a process, not an end-point. The first theme was defined as ‘impacts on mental health’ because of the importance placed on symptom change by interviewees. The second theme, ‘self-change and adaptation’ was defined by the negative changes that participants felt had happened following experiences of psychosis. The third theme, ‘social redefinition’ represents the social changes often associated with psychosis. The final theme ‘individualised coping mechanisms’ is representative of the way that people chose to cope and overcome their experiences. These themes each had two further subthemes that consisted of smaller themes (Box 1).

Discussion

The findings from this study highlight the multidimensionality of the recovery process in psychosis, and that this does include a role for symptom change. There appears to be four main aspects that are important to consider: impacts on mental health; self-change and adaptation; social redefinition; and adapting an individual coping style. These factors were shown to be important to all those interviewed. The varied emphasis that the interviewee's placed on change within these four areas indicates that recovery may not be considered merely an outcome with clear cut differences between being recovered and not being recovered. This supports the notion that recovery is an ongoing process, consistent with previous literature. ^{Reference Lewis, Tarrier, Haddock, Bentall, Kinderman and Kingdon25}

Nevertheless, participants found symptom alleviation to have a major bearing on their recovery and this highlights the importance of considering symptoms within the recovery process, and as an important indicator of outcome from treatment. However, these findings also highlight that recovery is much broader than symptom alleviation alone. Improvements in psychotic symptoms may be important to recovery, but only in conjunction with a range of other factors. Furthermore, the findings in relation to the need for change within symptoms may indicate that although full symptom alleviation or removal may be important for some service users, for others, changes in the nature of the symptoms may be just as important. For example, recovery may mean the continued presence of symptoms but without their negative impacts. Clinicians should consider this when working with service users by working more holistically and being mindful of the importance of other social and psychological factors.

The emotional impacts of having psychosis were also identified. The effects on depression, anxiety, anger and frustration illustrate that emotional change is also important to consider. Psychotic symptoms are often concomitant with these aspects and measured accordingly (e.g. PANSS) ^{Reference Peters, Myin-Germeys, Williams, Greenwood, Kuipers and Scott14} so more importance may need to be placed on these areas. Anger and frustration was highlighted. This may be of particular interest as it is often not considered in typical outcome measures within trials or services for psychosis.

Another aspect that was the highlighted as important to recovery was self-change and adaptation. All people interviewed identified a change to their character, personality and identity. For example, they were less confident, less energetic and less motivated. Most people found that they no longer were able to do the things they used to do, i.e. hobbies and activities. Rebuilding identity and character understandably plays a key part in recovery, as outlined in previous service-user-led research, ^{Reference Pitt, Kilbride, Nothard, Welford and Morrison5} and should be considered as main factors with regard to therapy and research.

A major area of recovery that was also common throughout the themes was the social impact of psychosis. Most participants found that their financial stability, living arrangements and employment status were affected by having experienced psychosis. This highlights the continued need for social relationships and issues that affect social behaviour to be considered when developing services with regard to recovery. All interviewees spoke about a decrease in social activities, an affect on their relationships with friends and loved ones and feeling isolated. Social activities are measured briefly in such measures as the PANSS, ^{Reference Peters, Myin-Germeys, Williams, Greenwood, Kuipers and Scott14} as social relationships are in the GAF, ^{Reference Heinrich, Hanlon and Carpenter23} but are often otherwise ignored in terms of published research. This current study and other previous service-user research has shown how important developing social networks and activities are in regard to recovery (e.g. Pitt et al, ^{Reference Pitt, Kilbride, Nothard, Welford and Morrison5} Chadwick) ^{Reference Wilken27} but they are often not being considered in enough detail by large quantitative studies (e.g. Kuipers et al). ^{Reference Smith, Flowers and Larkin32}