Treatment of Cognition in the Schizophrenia Spectrum: The Context of Psychiatric Rehabilitation

Abstract

Evidence-based approaches and modalities for targeting and treating the cognitive impairments of schizophrenia have proliferated over the past 15 years. The impairments targeted are distributed across the cognitive spectrum, from elemental perception, attention, and memory, to complex executive and social-cognitive functioning. Cognitive treatment is most beneficial when embedded in comprehensive programs of psychiatric rehabilitation. To personalize comprehensive treatment and rehabilitation of schizophrenia spectrum disorders, practitioners and participants must select from a rapidly expanding array of particular modalities and apply them in the broad context of the participant’s overall recovery. At present, no particular treatment, cognitive or otherwise, can be considered more important or primary than the context in which it is applied. Persistent difficulty in dissemination of new technology for severe and disabling mental illness compounds the significance of the context created by a full treatment array. In this article, a case-study of a mental health service system is described, showing the broad-ranging effects of degrading the rehabilitative context of treatments, obviating the benefits of cognitive treatment and other modalities. To realize the promise of cognitive treatment, the problems that prevent dissemination and maintenance of complete psychiatric rehabilitation programs have to be addressed.

Introduction

It has now been 15 years since publication of the first fully controlled study showing that the explicitly cognitive components of treatment for schizophrenia spectrum disorders contribute uniquely to overall outcome.¹ In the ensuing years we have seen a proliferation of new cognitive treatment approaches, so much so that we must now think systematically and holistically about the process of treatment development itself.² Computer technology has given rise to treatment techniques that were impractical or even impossible 10 years ago.³ We now differentiate neurocognitive from social-cognitive impairments,⁴ and we can fine-tune individual treatment accordingly. We have learned to integrate cognitive and neurocognitive constructs with complementary rehabilitation methods, ranging from social skills training⁵ to supported employment.⁶ We have treatment packages specialized for the schizophrenia spectrum population that systematically address cognitive, neurocognitive, behavioral, and social recovery goals.^7–9 All this progress is celebrated in the symposium in this special issue, although we still have fundamental scientific work to do, eg, to more fully understand cognitive impairments and their treatment in the context of physical health and biomedical dimensions of recovery.^10,11

Meta-analyses indicate that cognitive treatment has the most benefit when embedded in a comprehensive program of psychiatric rehabilitation.¹² Although there are direct benefits of improved cognitive functioning, the role of cognition in the skill acquisition processes that pervade psychiatric rehabilitation is also important. This was evident even in the earliest multivariate analyses of cognitive treatment effects.¹³ In addition, the first complete experimental segregation of treatment effects showed that even behavioral treatment modalities that do not explicitly target cognition do nevertheless benefit neurocognitive and social cognitive functioning, if delivered in an integrated treatment and rehabilitation program.^1,14 Since the early years, with the differentiation of neurocognitive and social cognitive approaches, and the proliferation of other psychiatric rehabilitation modalities that target various domains of personal and social functioning, we have come to see them all as embedded in an array of reciprocally interacting treatments.

This actually reflects an idea that has become central to psychiatric rehabilitation, that severe mental illness is a failure of multiple levels and components of the person’s biosystem, and these failures must be attacked at whatever points of access the biosystem provides.¹⁵ We have treatments that target the biosystem’s neurophysiological, neurocognitive, social cognitive, behavioral and social-environmental levels, addressing recovery goals ranging from stable brain chemistry to effective problem-solving to performance of desirable social roles. The pathways of recovery between targeted domains are complex and multidirectional (eg, as when improved illness management skills lead to better medication adherence, leading in turn to neurophysiological stabilization).¹⁰ Key pathways are probably different across individuals. Psychiatric rehabilitation will ultimately require a full understanding of the many ways in which cognitive treatment affects failures at all levels, directly and indirectly. Accordingly, further development of cognitive modalities must take into consideration the larger context of biosystemic psychiatric rehabilitation.

We have come full circle. For scientific reasons, it was initially critical to segregate the unique effects of the cognitive components from the other effects in a comprehensive treatment and rehabilitation regimen. Once this is accomplished, the higher priority for application is to optimally integrate the array, so that the interactions produce the most beneficial outcomes. Ultimately, this must be done on a case-by-case basis, applying the popular concept of personalization to individualized regimens of psychological and social, as well as pharmacological, treatments.¹⁶ However, to effectively personalize treatment for this notoriously heterogeneous patient population, practitioners and individual patients will still have to be able to select specific modalities from a complete and diversified formulary. For the time being, for the purposes of program development, it is difficult to argue that any one component of psychiatric rehabilitation is more important than any other, especially for the most severe, disabling and drug-refractory cases.

The Elephant in the Temple

The need for a complete and comprehensive psychosocial treatment array is a poignant reminder of the “elephant in the temple” of psychiatric rehabilitation. Despite strong empirical validation of the various treatments that make up our modern array, dissemination does not reach beyond a few academic research centers, and treatment for severe mental illness remains abysmal, at least in the United States.^17,18 Although dissemination barriers affect all new treatments including pharmacological ones, it appears that new psychosocial treatments are especially problematic.^19,20

Explanations of the dissemination problem usually focus on the broad economic consequences of historical stigmatization of the schizophrenia-spectrum population. However, recent social reforms, such as insurance parity legislation in the United States, indicate the effects of stigmatization are dissipating. Because the dissemination problem nevertheless continues, there are reasons to suspect that other factors are involved.²¹ One such factor is the conflict between professional and organizational conventions associated with traditional healthcare vs psychiatric rehabilitation. This was evident even before the deinstitutionalization movement was fully underway, when research was indicating that learning theory-based approaches produce outcomes superior to those observed in traditional “medical model” psychiatric institutions.²² Such approaches require an integrated but decentralized organizational structure and truly interdisciplinary treatment teams, not particularly attractive to providers irreversibly invested in a traditional medical-hierarchical organizational scheme.²³ Those most directly affected by such new organizational structures in public psychiatric institutions include the medical staff and nursing administrators, who have primary control over how treatment is provided. Adopting best practices in the form of psychiatric rehabilitation programs requires ceding a degree of institutional control to the treatment program and the professionals who operate it. Even the simple concept of organizing treatment into “programs” not accountable to department directors may generate vigorous resistance.

Closely associated with the organizational structure is the need to recognize direct care staff (the “psychiatric technicians”) as the most important psychosocial treatment agents.²⁴ This leads to a need for staff training, in extraordinary amounts compared with the amount training traditionally provided in psychiatric institutions. The trained staff then engages in therapeutic interactions with the patients that are directed by the treatment model and individualized treatment plans, not by nursing department policies or a physician writing orders. This shift of control is strongly resisted by the traditional institutional leadership.

As services evolve away from traditional government-operated public psychiatric institutions to smaller private providers in community settings, additional factors inhibit provision of evidence-based treatment for the historical institutional population. Ironically, one such factor is the historical segregation of people with serious mental illness (SMI), and the workforce that serves them, in institutions. Even after deinstitutionalization, people with SMI were often sequestered in “mental health ghettos” and there was little or no overlap between their services and the rest of community mental health service system.²⁵ The stigma of SMI may have applied to those who serve them as well as the patients themselves. Professionals and administrators in the community systems, insulated from the SMI population, do not necessarily have the background, experience or expertise to plan for the needed services.

If administrators and professionals do learn about serving the population, the risks, costs, and liabilities associated with serving people with refractory schizophrenia-spectrum disorders become powerful disincentives for providing any services at all. The result is “cherry picking,” the practice of accepting funding intended for serving a population or constituency, but systematically excluding individuals with more costly needs in order to preserve a corporate profit margin (or, in the case of “non-profit” organizations, overhead for executive salaries; The term “cherry picking” is also applied in other healthcare contexts, eg, to individual practitioners who select patients with positive prognoses. All forms of cherry picking in healthcare are arguably linked, but for present purposes the meaning is restricted to the level of corporate providers’ policies and business models.). Cherry picking is a well-recognized problem across the healthcare industry²⁶ and in the news media,²⁷ as well as the scholarly policy community.²⁸ Without vigorous regulation, public resources for the historical institutional population are especially vulnerable to cherry picking when the resources are made accessible to private providers. However, the public sector is also vulnerable. An example familiar to many in the mental health industry is policy that prohibits reimbursement for vocational rehabilitation services until the recipient has sustained competitive employment for 90 days. The well-known failure rate of vocational rehabilitation for people with SMI, even when evidence-based practices are applied, dissuades providers from investing in SMI recipients.

Cognitive treatment for severe mental illness may be even more susceptible to cherry picking than other rehabilitation modalities, due to its superficial resemblance to more familiar, traditional modalities administered to less disabled patients in a conventional office setting by an individual provider. Cognitive-behavioral therapy specialized for psychosis (CBTp)^29–31 has similar features, and might be similarly susceptible. It would arguably be more difficult for a provider to falsely claim, for regulatory purposes, that other rehabilitation modalities like community support, in vivo skill training and supported employment, are being delivered with high fidelity in an outpatient clinic office. However, even if those modalities were provided with high fidelity in the office, if they were not integrated with a comprehensive rehabilitation array they would not generate the expected benefit. Additionally, without the various modalities that stimulate engagement in treatment and rehabilitation across the array, such as collaborative pharmacotherapy, illness management skills training, contingency contracting, behavioral family therapy, and community support, there is selective attrition of people who most need them, ie, people with the most severe and disabling schizophrenia-spectrum disorders. The high cost, high-risk patients thus cherry pick themselves, complementing the providers’ business model, and removing the demand for a complete, integrated and accountable psychiatric treatment and rehabilitation array.

The logical antidote for cherry-picking is well-informed administrative regulation of funding and service provision. The risks and costs must be recognized and managed. This in turn requires regulatory recognition of the need for individualized treatment plans and fully integrated rehabilitation. Early in the post-deinstitutionalization era, assertive community treatment³² (ACT) emerged as an approach to organizing services for people with SMI. ACT was especially amenable to appropriate administrative regulation because it is a coherent and operationally defined package of services, with clearly identified provider responsibilities, specifically designed for the more severely disabled part of the post-institutional population. ACT emulated key features of the traditional institutional treatment model in community settings, most importantly a multidisciplinary treatment team, operating systematically according to set principles and procedures, and comprehensively responsible for patients’ treatment and social support needs. This led to characterization of early ACT as a “hospital without walls.” The original version of ACT proved effective at preventing recidivism among deinstitutionalized people, and was widely disseminated as one of the first “evidenced-based practices” of the post-institutional era. However, without inclusion of additional psychiatric rehabilitation modalities outcomes did not include progression to more independent personal and social functioning. In our contemporary era of the recovery movement,³³ simply preventing re-hospitalization is not a sufficient outcome, and even the differential recidivism achieved by early ACT is not good enough. Best practices for the SMI population require both administrative/organizational mechanisms to ensure accessibility and integration, as in the original ACT model, plus a full array of evidence-based treatment modalities, including cognitive treatments, and the capacity to thoroughly personalize rehabilitation according to recovery goals. Nevertheless, simple organizational models that neglect the modern treatment array may still appear to naïve administrators as convenient and economical service packages whose implementation takes care of the need for SMI services.

The factors that inhibit dissemination of best practices are potentially present in every mental health bureaucracy, service system, and institution. To realize the promise of cognitive treatment for schizophrenia, confronting and overcoming these factors will be at least as important as scientific and technological breakthroughs. The remainder of this discussion is a case example of how these factors can play out to prevent implementation of cognitive treatment in particular, and best practices in general, for people with SMI. It is a case of failure to maintain the administrative regulatory mechanisms necessary to ensure accessibility, personalization, integration, and comprehensiveness of psychiatric rehabilitation for the post-institutional SMI population. Understanding and foreseeing how these factors play out in ordinary real-world settings is prerequisite to effectively managing them.

The case is the regional mental health service system that serves Lincoln, NE, a metropolitan state capital city, plus surrounding rural counties and villages, with a total population of about 440000. In a reversal of the usual situation of best practices being introduced into an obsolete service system, this example shows the results of dismantling a relatively advanced service system spanning institutional and community settings. Due to a series of state- and regional-level policy decisions, psychiatric rehabilitation services were gradually degraded over a period of more than 10 years, first as a reduction of the individualization and integration components of the system, and later as removal of evidence-based modalities from the remaining service array.

A Case Analysis

The case analysis begins with a data set that was compiled almost a decade after the process under study began. In November 2013 the Lincoln Journal Star published a series of related news stories about the role of the city police department in community mental health. Since about 2000, the police had been collecting data on the number of police calls that were for mental health issues, eg, suicide threats, conflicts and confrontations between service consumers and providers, public bizarre behavior and so on. One of the newspaper stories³⁴ included a spreadsheet showing the actual number of police mental health calls, by year, from 2001 to 2013. That data are graphically represented in figure 1.

Fig. 1.

Mental health calls per 100000 (corrected for population growth) recorded by the Lincoln Police Department, 2001–2012.

The Lincoln area had unusually well-developed services for the historical institutional population, dating back to the initial wave of deinstitutionalization, which hit Nebraska in the mid-1970s. The county mental health center was specifically created to serve the people leaving Lincoln State Hospital, whose census shrunk from 2500 patients to 250 virtually overnight. Community-based services were initially organized around a generic case management/social support approach, but as psychiatric rehabilitation technology evolved, its specific evidence-based practices were incorporated, both in the county mental health center and a remaining long-term unit of the state hospital. Community as well as institutional services benefited from collaborations with the clinical psychology training program of the state university in Lincoln. In the 1990s the hospital rehabilitation program was the site of the aforementioned first RCT of cognitive treatment¹ and later a related attention-training modality.³⁵ A spin-off residential program was started in the community, a collaboration between the state hospital, the county mental health center and the university, using the intensive rehabilitation model developed at the state hospital. Community-based rehabilitation combined with new subsidized housing programs to yield unprecedented recovery outcomes for the historical institutional population. Evidence-based services for people with schizophrenia spectrum disorders were arguably as good as anywhere in the United States. Then, as figure 1 shows, in 2004 something happened.

In 2004 the state implemented a “second wave” deinstitutionalization reform intended to further reduce the remaining state hospital population. Although the planning for this change was initially systematic, it became unplanned as a new gubernatorial administration took office. The final reduction of state hospital beds, by 60%, reflected a political agenda to close 2 of the 3 facilities, and was not based on any assessment of the population’s needs or the future role of the remaining facility. As the local economic implications of closing large state facilities became apparent, the political reaction further fragmented the reform process. A detailed and comprehensive best practices document had been developed before the gubernatorial change, describing the projected needs of the population, namely for comprehensive psychiatric rehabilitation in community settings, but was subsequently ignored. One provision of the reform act, shifting responsibility for services from state-operated facilities to community providers, became an abdication of state-level responsibility for planning and funding services for the historical institutional population. A decade later, a state legislative investigation of the ensuing debacle determined that although some $30 million had indeed been liberated from the closed hospitals, nobody knew where those resources had gone. The money had been distributed without condition to administrators of multi-county regions, supervised by boards of county commissioners, with no provisions for subsequent tracking.

Although the funds released by institution downsizing were ultimately impossible to track, planning documents indicated that the regional administration’s plan for accommodating the discharged patients was mainly to provide new short-term case management and triage services. However, there were no new funds for existing treatment, support and rehabilitation services to accommodate the new population. The implicit expectation, reflecting administrators’ beliefs that the existing rehabilitation services were unnecessary, was that the new patients would need only to be triaged to subsidized housing, medication prescribers and a brief period of social support. The new case managers were generally nonprofessionals operating with little or no supervision. The service provision model, if there was one, was characteristic of the first round of deinstitutionalization, before development of even the earliest forms of ACT. Not surprisingly, the “short-term” case management services quickly became “chronic” community support and crisis intervention, as people with severe schizophrenia-spectrum disorders and years of institutionalization entered the community service system.

The effects are visible almost immediately in the Lincoln police calls data. As institution downsizing progressed, mental health calls increased. The infusion of new patients was complete within 4 years, and by then the increase in mental health calls appeared to be leveling off. However, they were rising again by 2010. In 2009, the state had abruptly closed the psychiatric rehabilitation program at the state hospital. The program had been closely integrated with the community programs of the county mental health center, and had been a key conduit back to the community for the most high-risk and “difficult-to-discharge” patients.³⁶ The rehabilitation program was converted to a generic inpatient unit with the same number of beds. Although administrators expected the result of this closing would be to accelerate discharges, the discharge rate from non-forensic hospital beds actually slowed by some 40% over the next 1.5 years (figure 2). The antecedents and consequences of that closing are quantitatively analyzed in detail in a previous case study.³⁷ For present purposes, the analysis showed an immediate and explosive increase in restraint and seclusion within the hospital, an increase in the size of the long-term inpatient population, increase in average duration of hospitalization, degraded outcome of community-based services for patients recently discharged, and increased stress on county emergency and crisis services. The police mental health calls continued to increase, and were still increasing 8 years after the initial institutional downsizing.

Fig. 2.

Change in state hospital discharge rate from nonforensic beds after closure of psychiatric rehabilitation program.

Although the closing of the state hospital rehabilitation program eliminated psychiatric rehabilitation from the state’s secure settings, a robust program continued in the county mental health center. There was a complete array of modern modalities, including cognitive treatment and CBTp, provided in outpatient, community outreach, day treatment, and residential settings, to formerly institutional patients with schizophrenia spectrum disorders. Then, in 2014, the county withdrew fiscal support, the mental health center was closed, and the remainder of its funding was distributed by the regional administrator to two private nonprofit behavioral health service providers. The transition process and consequent impact on services for people with schizophrenia-spectrum disorders is described in detail in the doctoral dissertation of K.L. Golba.³⁸

The regional administrator and the successor providers could have preserved the county mental health center’s service array, but they did not. None of the professionals and very few of the direct care staff from county programs were hired by the successor providers. Neither of the successor providers had staff psychiatrists or psychologists, which rendered them incapable of addressing forensic mental health needs. None of the treatment array for schizophrenia spectrum patients was retained, including cognitive and neurocognitive modalities. The successor provider funded to take over the county’s outpatient and community support services even declined to provide clozapine, with no other way for people dependent on this last-recourse antipsychotic medication to get it. As of this writing, there are still no cognitive, neurocognitive, or behavioral treatments evidence-based for schizophrenia spectrum disorders available in the region. A third provider that had been excluded from the funding redistribution managed to hire some of the experienced county staff for case management and community support, start a clozapine clinic, and take responsibility for patients with a forensic status, by using meager and unreliable funding channels not controlled by the regional administrator.

By the time the county services were dismantled, the patient turnover in the remaining state hospital had ground to a halt. The hospital’s psychiatric rehabilitation program and its steady rate of discharging the most difficult patients were long gone. The community-based residential program that provided the key step-down for the most disabled institutional patients was a county program that was not continued by the successor provider. Although the state hospital still has the statutory responsibility for civilly committed as well as forensic patients, only the latter are admitted, in response to the threat of courts holding administrators in contempt. At the end of the case-study period, the state hospital had ceased to be a factor in the state and regional mental health system, and there were no community-based programmatic psychiatric rehabilitation services in the region.

The one appropriate service that remained relatively available throughout the reforms was routine antipsychotic pharmacotherapy, because it could reliably be provided by independent practitioners and reliably reimbursed through the state’s indigent healthcare system. There may have been a shift from psychiatrist prescribers with extensive experience with schizophrenia spectrum to nurse practitioners with less experience, and medication adherence may have decreased without adjunctive supports, but otherwise all the reform outcomes are attributable to removal of the psychosocial components of the service array.

Although some features of this case example reflect local politics and economics, the sequence of events reveal the separate effects of factors inherent in service systems that inhibit implementation of best practices for SMI services. For 5 years after the initial reforms, the community service system struggled to accommodate a new post-institutional population with an obsolete service model, insensitive to long-term rehabilitation and recovery needs. Existing rehabilitation programs had not yet been dismantled, but the numbers of new patients in the system, without integrated and coordinated services, had a pervasively destabilizing effect, evident in the police data. This phase of the debacle was driven primarily by administrative failure to recognize the rehabilitation and recovery needs of the SMI population, and the consequent need for integrated, coordinated, programmatic services. However, the initial infusion of new patients would not be driving the continuing increase 6 years later. That was the result of dismantling the service array at a pivotal point in the system, the remaining secure institutional facility. After the institution-based rehabilitation program was closed, patients discharged from the institution were less psychiatrically stable and less progressed in their functional recovery. They compounded the pressure on a community system whose rehabilitation capacity had already been exceeded by the initial reform, so conditions continued to deteriorate. The rehabilitation program was closed because the persistent resistance of institutional interests, held in check for decades by state-level administration, was critically enhanced by the momentum of downsizing and by administrators’ beliefs that the need for best practices in the remaining institution would be obviated by community-based services. This is further evidenced by the fact that no institutional beds were eliminated when the rehabilitation program was closed—they were simply converted to a traditional, quasi-acute inpatient treatment model.

Closing the county mental health center and thereby eliminating the entire psychiatric rehabilitation array is a case of collaborative cherry-picking by administrators and providers who have no intention or ability to provide the services or serve the population. The cherries were picked because resources previously allocated for the SMI population were made available without administrative regulatory provisions to ensure that they stayed with the population. The administrative failure in turn reflects a failure to recognize that best practices for the historical institutional population, people with schizophrenia-spectrum disorders, are different from those for substance abuse and other behavioral health problems.

In the perspective of a treatment researcher and developer, the most visible and salient events in this case study may be the removal of specific evidence-based modalities associated with a modern psychiatric rehabilitation array, including cognitive, neurocognitive, and cognitive-behavioral treatment. Perhaps less visible and salient is the shift to a case management and triage model without interdisciplinary teams, functional assessments, integrated treatment and rehabilitation plans, or accountability for functional recovery. These dimensions of a service array are the most difficult to achieve, especially in traditional metal health service models, so it is no surprise they were not sustained by the successor providers. It is more surprising that there was no effort to adopt specific modalities that can be provided in a traditional outpatient office setting, eg, cognitive remediation or CBTp. The professional resources were available. The demand for the specific services and the reliability of funding for them was beyond doubt. On the other hand, a provider whose business model involves insulation from the schizophrenia spectrum population has no need for services or providers for that population.

The continuing condemnations and calls for reform of American mental health services suggest that the undesirable processes observed in this case-study do operate in most systems. State institutions have not disappeared, and most of them still employ obsolete quasi-acute inpatient hospital models and operate under hospital regulatory regimens. This sustains a small but entrenched reactionary institutional industry. Professional and administrative experience and expertise with the schizophrenia spectrum population is still relatively restricted to these settings, preventing sensitization of community professionals, administrators and provider organizations to the needs of the population. Misconceptions about the nature of severe mental illness abound among administrators, eg, that schizophrenia is a strictly episodic condition, with periods of acuity interspersed by normal functioning, comparable to episodes of substance abuse. Cherry picking has become a business model. Strategic plans, administrative policies and regulations do not generally recognize, and certainly do not facilitate, the centrality of interdisciplinary treatment teams, treatment and rehabilitation planning, integrated and coordinated treatment, and accountability for outcomes in best practices for the schizophrenia spectrum population. These are challenges to dissemination of cognitive treatment and many other best practices, in most venues.

Conclusions and Projections

To realize the promise of cognitive and neurocognitive treatment for schizophrenia spectrum disorders we will have to find ways not only to build but to sustain the broader infrastructure of psychiatric rehabilitation. Beyond stigmatization of the schizophrenia population, the barriers include business models designed to secure revenue without actually serving the intended population, regulatory practices that allow this to happen, and institutional interests intolerant of accountability for outcomes.

In the United States, the regulatory failures may extend to the national policy level, where there has been systematic obfuscation of the difference between rehabilitation and recovery as it has evolved in the self-help and substance abuse industries, vs psychiatric rehabilitation as it evolved in rehabilitation medicine and psychology for physical disabilities, and recovery from schizophrenia spectrum disorders^39–41). Even the term Serious Mental Illness, originally coined to refer to the historical institutional population, has come to include virtually any psychiatric disorder.⁴² In the Nebraska case example, terminological obfuscation played a key role in disenfranchising the entire schizophrenia-spectrum population. The successor providers could credibly claim experience and expertise with “serious mental illness” while having none with the post-institutional schizophrenia-spectrum population.

At the level of individual practitioners, we need not only to train therapists to perform high-fidelity cognitive and neurocognitive treatment, but to educate mental health professionals across the disciplines about the unique contributions of such treatment and the need for integrated, personalized application in comprehensive psychiatric rehabilitation. It will take a coalition of practitioners, scientists, administrators, and consumer advocates to overcome the political and economic disincentives against providing psychiatric rehabilitation and its respective evidence-based modalities, including cognitive treatment.