Issues and developments on the consumer recovery construct

Abstract

The consumer recovery model has had increasing influence on mental health practices in the United States, Western Europe, and several other countries. However, adoption of the model has reflected political decisions rather than empirical evidence of the validity of the model or its value for treatment services. The recovery construct is poorly defined, and until recently there has been no reliable and valid measure with which to base a research program. We have developed an empirical measure that is well-suited for both research and clinical applications: the Maryland Assessment of Recovery in Serious Mental Ill-ness (MARS). We briefly describe the MARS and present preliminary data demonstrating that recovery is not a simple by-product of traditional outcome do-mains, but seems to be a distinct construct that may have important implications for understanding consumers with serious mental illness and for evaluating the outcome of treatment programs.

Schizophrenia and most other forms of serious mental illness have traditionally been viewed as chronic conditions with poor outcomes. This pessimistic view has begun to change, as a series of long-term outcome studies have demonstrated that the course is more variable both across and within individuals, and that many people meeting strict diagnostic criteria have very good outcomes, often without maintenance medication (e.g., 1,2, 3, 4). There are now upwards of 20 contemporary studies of the long-term outcome of schizophrenia. These studies vary in specific criteria, measures, samples, and time frame, but overall some 50% of people with careful research diagnoses appear to have a good outcome, with substantial reduction of symptoms, and good quality of life and role function over extended periods of time.

At the same time as these new outcome data have been collected, there is growing recognition that traditional paternalistic mental health services have generated feelings of hopelessness and helplessness among many consumers, promoting dependence, and fostering stigma. In response to the failure of traditional services, consumers and many professionals have promoted a recovery movement, based on a model of recovery and health care that emphasizes hope, respect, and consumer control of their lives and mental health services 5.

Two important reports from the US federal government provided momentum to the recovery movement. First, the Surgeon General’s Report on Mental Health 6 concluded that all mental health care should be consumer and family oriented and have the promotion of recovery as its primary aim. This position was echoed more forcefully in the President’s New Freedom Commission report Achieving the promise: Transforming mental health care in America 7. Among other things, the report stated: “…care must focus on increasing consumers’ ability to successfully cope with life’s challenges, on facilitating recovery, and on building resilience, not just on managing symptoms”. The principles enunciated in these reports have been adopted by several state mental health systems in the United States, along with Canada, the United Kingdom, Italy, Australia, and New Zealand.

CONSUMER-ORIENTED DEFINITIONS OF RECOVERY

The consumer movement and the associated policy changes are based on the contention that recovery is a process that occurs over time in a non-linear fashion 8. Anthony 9 described it as “a deeply personal, unique process of changing one’s attitudes, values, feelings, goals, skills, and/or roles. It is a way of living a satisfying, hopeful, and contributing life even with limitations caused by illness. Recovery involves the development of new meaning and purpose in one’s life as one grows beyond the catastrophic effects of mental illness”. More recently, in a report on the New Freedom Commission, Hogan 10 described recovery as “a process of positive adaptation to illness and disability, linked strongly to self-awareness and a sense of empowerment”. The key elements of these definitions (recovery as a process in which the individual strives to overcome the fact of mental illness and its impact on one’s sense of self) have been echoed in many other definitions 11.

In 2004, the US Substance Abuse, Mental Health Services Administration (SAMHSA) held a 2-day consensus conference with over 100 consumers, mental health professionals and scientists, and developed the following definition of recovery: “Mental health recovery is a journey of healing and transformation for a person with a mental health disability to be able to live a meaningful life in communities of his or her choice while striving to achieve full human potential or personhood”. Ten characteristics of recovery and recovery-oriented services were also identified by SAMHSA: self-direction, individualized and person-centered, empowerment, holistic, non-linear, strengths-based, peer support, respect, responsibility, and hope. The SAMHSA definition is widely accepted by the field. It has been adopted by the Veterans Health Administration and several state mental health systems, and will guide future SAMHSA funding programs.

The SAMHSA definition and dimensions are each elaborated in an accompanying document, but they do not provide an operational definition of recovery. Rather, they comprise diverse dimensions of the recovery model, in- cluding: person characteristics (e.g., self- direction, empowerment, respect (for self), responsibility, hope), systems characteristics (e.g., individualized and person-centered, strengths-based, peer support, respect (from others)), and descriptors or parameters of recovery (e.g., holistic, non-linear).

As SAMHSA is the federal agency charged with developing and implementing national health policies, this conceptualization will likely have important implications for clinical practice and reimbursement in the United States. However, the recovery components specified by SAMSHA are not well-defined and there are marked redundancies across the items (e.g., empowerment and self-determination). Some of the elements refer to individual characteristics (e.g., hope, respect), while others refer to characteristics of the person’s environment or clinical service the person receives (e.g., the value of peer support). Overall, the elements are not adequate criteria for research, or for evaluating the effectiveness of clinical programs. They also do not provide adequate guidance about how to evaluate a person’s recovery status or changes over time, or to determine what other environmental or clinical factors are associated with recovery.

LIMITATIONS OF THE CONSUMER MODEL OF RECOVERY

No systematic data are available on rates of recovery as defined from the consumer perspective. Anecdotal data and commentaries by the many impressive consumer spokespersons for the recovery model are informative, but it is difficult to extrapolate from these sources of information. It is clear that the professional and scientific communities have not sufficiently appreciated the subjective experiences of people with severe mental illness, and their ability to recover from the debilitating effects of their illness. Conversely, it is not clear if the experiences of consumer-professionals are characteristic of the broader population of people with severe mental illness or if they represent a distinct good outcome subgroup.

The consumer recovery model has been referred to by some mental health professionals as “old wine in new bottles” and a “feel good” conceptualization that does not have true practical implications 12,13. If the concept is to have lasting impact, it is essential that it be tied to more objective measures of course of illness and community functioning that are viewed as relevant by scientists, clinicians, family members, and legislators. Studies are required to understand factors that contribute to consumer-defined recovery and determine its course. For example, consumer definitions generally suggest that recovery is independent of symptoms, but the few studies that have examined this issue report that recovery and symptoms are negatively correlated 14,15.

A social cognitive model of recovery

A major limitation of the consumer model of recovery is that is it not grounded in established psychological principles, and refers to vague constructs that have not been objectively defined 16. We conceptualize recovery in the context of Bandura’s social cognitive theory of human agency 17,18. Bandura postulated that people are agents of their experience and not simply passive respondents to a deterministic environment, or automatons who are driven by neurocognitive processes. The primary engine through which agency operates is self-efficacy. This is a set of beliefs about one’s capacity to manage internal and external experiences. It includes both generalized confidence in one’s abilities, and situationally specific efficacy beliefs. It also involves both personal agency (what people can do on their own) and interpersonal agency (ability to marshal help from others) 19.

Self-efficacy is determined by success/failure experiences, modeling (i.e., social learning), and the reactions of others over time. It has a powerful influence on motivation and goal setting, life choices, and action. The more people are confident in their ability to succeed or cope effectively, the more willing they are to set ambitious goals and take action. Self-efficacy also influences affect states. High self-efficacy can lead to mastery experiences, self-esteem, and life satisfaction, while low self-efficacy can lead to anticipatory anxiety, a sense of failure (regardless of actual performance), helplessness, and depression.

Figure 1 provides a graphic representation of the model. Negative experiences and attitudes (on the left) diminish self-efficacy, which leads to decrements in parameters of recovery, while positive experiences and attitudes (on the right) enhance self-efficacy and foster recovery. People with severe mental illness often have histories of personal failure in a host of social role experiences, experience harmful stigma (from the public, media, mental health professionals, and sometimes significant others), and often develop self-stigma 20. These experiences can erode self-efficacy for coping with their illness or dealing with a broad range of life demands. Diminished self-efficacy can lead to hopelessness, lack of self-respect, and lack of feelings of control (self-direction) or empowerment, which has been shown to happen in people with severe mental illness 21. Conversely, vocational success, effective shared decision making in health care, and other mastery experiences can produce increased self-efficacy and enhance feelings of empowerment, hope, self-respect, and capacity for self-direction.

Figure 1 A model of the relationship between self-efficacy and recovery.

Recovery from severe mental illness entails developing enhanced efficacy beliefs for key social roles (e.g., student/worker, spouse/parent), and developing a sense of control over one’s illness through both personal agency (e.g., the ability to manage one’s illness and its treatment) and interpersonal agency (e.g., ability to work collaboratively with health care providers and access support from peers and family members when needed). The contention that recovery entails adaptation to illness and disability 10 is consistent with the social learning view that efficacy beliefs are specific to situations and that a person can feel efficacious in some domains despite hav- ing difficulty in others.

The empirical literature on self-efficacy and agency in severe mental illness is limited, although the concepts have been widely linked 22,23,24 and several studies support the relationship between efficacy and outcomes in severe mental illness samples. Personal efficacy has been shown to be related to quality of life and community functioning, including employment, in several studies 23,25,26. Efficacy was found to have a strong negative relationship with perceived discrimination and self-stigma, and a strong positive relationship with empowerment in an outpatient schizophrenia cohort 27, and it was inversely related to depression and perceived loss of independence in a schizophrenia spectrum sample 28. While not directly measuring self-efficacy, Grant and Beck 29 examined the related construct of defeatist beliefs: overgeneralized conclusions about one’s ability to perform tasks (e.g., “If you cannot do something well, there is little point in doing it at all”). They found that these negative attitudes mediated the relationship between cognitive impairment and both negative symptoms and social and vocational functioning in a sample of people with schizophrenia and schizoaffective disorder.

Mediators and moderators of recovery

Another limitation of the current literature on the consumer model is that it is not clear to what extent recovery is mediated or moderated by functional outcome domains, such as work and social relationships: i.e., is productive activity like work or school a mediator of recovery, a consequence of recovery, or orthogonal to it? Does progress along the path toward recovery enable improved social relationships, do improved relationships contribute to recovery, or is movement along both dimensions somehow intertwined? We believe it is essential to develop a scientific base for the consumer model and to document that recovery has important practical and conceptual implications that extend beyond the subjective well-being of consumers.

Consistent with Bandura’s contention that people both influence and are influenced by their experience, we hypothesize that agency and self-efficacy are mediators between life experiences and recovery. Adverse experiences may diminish efficacy, and prevent or retard recovery, while positive experiences would have the opposite effect. There is also a feedback loop in which enhanced efficacy and progress along the path of recovery motivates and empowers the person to make positive life changes. For example, improved housing may increase hope and self-respect 30, which may increase the person’s inclination to seek employment. Conversely, an increased sense of hope and empowerment may enable the person to seek better housing.

Recovery may also be influenced by moderators. Recovery-oriented treatment can exert a positive influence, and paternalistic care can have a negative effect. Some domains, such as substance use, psychiatric symptoms and cognitive impairment, may function as (negative) moderators only when they are at significant levels.

Assessing recovery

Yet another limitation of the recovery construct is that there are no measures of recovery as defined by SAMHSA and only a handful that are based on other definitions 16. Andersen et al 31 identified only one recovery measure in a search of the published literature. Campbell-Orde et al 32 surveyed consumer and government organizations as well as the literature and identified eight measures, of which only six actually focus on recovery per se.

Results of these surveys reflect the fact that, for the most part, extant recovery instruments have been developed ad hoc by consumer groups and have not been published. They have evolved from small work group or consensus conferences with primary attention to face and consensual validity rather than a systematic psychometric program of scale development. Most are based on unsupported models or definitions of recovery. Most instruments have problems in scaling, and/or have inadequate floor or ceiling. Some are too long to be practical, and others are too heterogeneous to be useful as overall outcome variables.

In response to this situation we have developed a new measure based on our operationalized version of the SAMHSA recovery domains: the Maryland Assessment of Recovery in People with Serious Mental Illness (MARS) 33. The MARS is a 25-item self-report instrument designed to assess recovery status in people with serious mental illness. It was developed using an iterative process by a team of six doctoral level clinical scientists with expertise in serious mental illness and recovery in a series of ten face-to-face meetings, supplemented by structured interviews with six independent experts and a panel of consumers.

The MARS takes less than 10 minutes to complete and its items are written at a 4th grade reading level, making it practical for use in both research and for service delivery agencies. It should also be easy to translate to multiple languages. Notably, despite being developed to reflect the diverse SAMHSA recovery domains, a single primary factor accounts for the majority of variance.

We are currently conducting a longitudinal study to evaluate our social learning model and examine mediators and moderators of recovery. With a sample of more than 100 outpatients recruited to date, the data provide considerable support for our model. Table 1 provides a summary of a step-wise regression analysis of a range of domains on the MARS. Self-efficacy and human agency account for the largest proportion of variance in MARS scores: 59%. Other recovery constructs, including hope and empowerment, also account for a significant proportion of variance, but do not add appreciably to self-efficacy and agency. Positive and negative symptoms, neurocognitive functioning, social support, subjective quality of life, and health status, or receipt of recovery oriented treatment also add modest amounts of variance.

These data suggest that recovery is not a simple by-product of traditional outcome domains, such as symptoms, and is not a proxy for quality of life. Rather, it seems to be a distinct construct that may have important implications for understanding consumers with serious mental illness and for evaluating the outcomes of treatment programs. However, it should also be noted that the MARS was not highly correlated with either the receipt of recovery oriented treatment or with satisfaction with treatment. Thus, we have much to learn about what types of treatment services will enhance recovery and how to assess the extent to which services meet consumer’s recovery needs.

As indicated above, these data are preliminary and should be interpreted with caution. In addition, subjects were all receiving services at Veterans Administration hospitals in the United States. We are currently recruiting a larger and more diverse sample and will be assessing consumers 1 year after the baseline assessment. This will give us a picture of the influence of the diverse outcome and environmental factors on recovery over time, as well as how recovery status influences psychosocial functioning.

Table 1.

Table 1 Hierarchical regression of outcome domains on the Maryland Assessment of Recovery in People with Serious Mental Illness (MARS)

	Overall ANOVA				Change statistics
Step	F (df)	p<	R2		Change in R2	F (df)	p<
1	48.92 (3,102)	0.000	0.590
2	44.36 (6,96)	0.000	0.735		0.145	8.474 (6,93)	0.0001
3	24.24 (11,90)	0.000	0.748		0.013	0.385 (11,84)	0.0958
4	14.60 (16,80)	0.000	0.745		-0.003	0.047 (16,69)	0.999
5	12.76 (18,78)	0.000	0.747		0.002	0.022 (18,62)	0.999
1 - self-efficacy and human agency; 2 - hope, empowerment, and self-stigma; 3 - positive and negative symptoms and neurocognition; 4 - social support, subjective quality of life, and self-rated physical and mental health; 5 - receipt of recovery oriented treatment and satisfaction with services