Running head: INTERVENTION FOR INVOLUNTARY INPATIENTS
THE ROLE OF GUARDIANSHIP IN THE COURSE OF TREATMENT AND
TREATMENT OUTCOME FOR INDIVIDUALS RECOVERING FROM SEVERE MENTAL ILLNESS
by
Thea L. Rothmann, M.A.
University of Nebraska, 2006
Adviser: William D. Spaulding
The purpose of this study was to evaluate the role of guardianship in treatment and treatment outcome for people recovering from severe mental illnesses (SMI) in a psychiatric rehabilitation context. Research in the intersecting field of mental health care and the law is in its nascence. Using the unifying theory of therapeutic jurisprudence, this study investigates the clinical correlates of guardianship in a population of people with SMI. An archival database from an inpatient psychiatric rehabilitation program in a Nebraska state hospital was used in analysis. This was an ideal context and population for the study because this group of people is highly affected by the legal constructs evaluated in this investigation. The archival database contained comprehensive clinical, demographic, and outcome data for all participants. This included assessments of neurocognition, social cognition, symptomatology, behavioral functioning, and treatment compliance. In addition, outcome data regarding discharge location and rehospitalization were available. Two main hypotheses were put forth towards the overall purpose of this study. First, it was hypothesized that people with guardians would demonstrate lower overall functioning at the time of admission and throughout the course of treatment across multiple domains when compared to those without guardians. Second, it was hypothesized that people with guardians would be discharged to more restrictive community placements than those without guardians, but that they would have a lower rate of rehospitalization. Partial support for both hypotheses was obtained. Results suggest that, in Nebraska, people with guardians can be discriminated from those without guardians based on behavioral functioning. In addition, people with guardians were found to have a longer length of stay. Other areas of clinical functioning assessed – neurocognition, social cognition, symptomatology, and treatment compliance – were not found to differ between those with and without guardians. People with guardians were discharged to more restrictive settings and there is some evidence that they were rehospitalized sooner than those without guardians. This is the first known study to empirically investigate the role of guardianship in the recovery of people with SMI.
Table of Contents
List of Tables and Figures iv
Dedication v
Acknowledgements vi
Chapter 1 - Introduction 1
Chapter 2 – Literature Review 4
Severe Mental Illness and Psychiatric Rehabilitation. 4
Intersection between Mental Health Care and the Law. 6
Guardianship. 8
Related Studies. 10
Present study. 19
Chapter 3 - Method 23
Design Overview. 23
Setting. 23
Participants. 24
Measures. 26
Demographic and Clinical Characteristics. 26
Neurocognition Measures. 27
Social Cognition Measures 29
Behavioral Functioning Measures. 30
Symptomatology Measures. 31
Treatment Compliance Measure. 32
Outcome Measures. 32
Procedure. 35
Data Collection. 35
Data Cleaning. 38
Data Analysis. 38
Chapter 4 – Results 40
Demographic and Clinical Characteristics of People with Guardians at CTP. 40
Hypothesis 1 48
Functioning at admission. 49
Neurocognitive functioning. 49
Social cognitive functioning. 54
Behavioral functioning. 56
Treatment compliance. 59
Symptomatology. 60
Functioning over the course of treatment. 63
Neurocognitive functioning. 63
Social cognitive functioning. 65
Behavioral functioning. 66
Treatment compliance. 71
Symptomatology. 72
Hypothesis 2 75
Outcome. 75
Discharge location. 75
Rehospitalization rate. 76
Chapter 5 – Discussion 79
General Discussion. 79
Limitations of the Present Study. 88
Future Directions. 89
References 92
Appendix 102
List of Tables & Figures
Table 1. Demographic and Clinical Characteristics of Participants (N=162)
Table 2. Guardianship Status of Participants (N=162)
Table 3. Relationship Between Axis I Diagnosis and Guardianship Status (N=162)
Table 4. Relationship Between Race and Guardianship Status (N=162)
Table 5. Guardianship Status by Demographic and Clinical Characteristics Multivariate Analysis of Covariance (MANCOVA)
Table 6. Mean Scores and Standard Deviations for Demographic and Clinical Variables as a Function of Guardianship Status and Age
Table 7. Standardized Canonical Coefficients and Structure Weights from Discriminant Analyses of Demographic and Clinical Variables to Predict Guardianship Status
Table 8. Guardianship Status by Neurocognitive Variables (Group 1) at Admission Between Group Multivariate and Univariate Statistics
Table 9. Guardianship Status by Neurocognitive Variables (Group 2) at Admission Between Group Multivariate and Univariate Statistics
Table 10. Guardianship Status by Neurocognitive Variables at Admission One-Way Analyses of Variance (ANOVAs)
Table 11. Guardianship Status by Social Cognitive Variables at Admission Between Group Multivariate and Univariate Statistics
Table 12. Guardianship Status by NOSIE Subscales at Admission One-Way Analyses of Variance (ANOVAs)
Table 13. Standardized Canonical Coefficients and Structure Weights from Discriminant Analyses of NOSIE Subscales to Predict Guardianship Status
Table 14. Guardianship Status by BPRS Factor Scores at Admission Between Group MANOVA
Table 15. Guardianship Status by BPRS Items One-Way Analyses of Variance (ANOVAs)
Table 16. Guardianship Status by COGLAB Repeated Measures ANOVAs
Table 17. Mean Scores and Standard Deviations for NOSIE Total Assets at Admission, 6 Months, and 12 Months as a Function of Guardianship Status
Table 18. Guardianship Status by NOSIE Subscales Repeated Measures ANOVAs
Table 19. Means and Standard Deviations on NOSIE Subscales at Admission and 12 Months as a Function of Guardianship Status
Table 20. Relationship Between Level of Discharge Location Restrictiveness and Guardianship Status (N=147)
Figure 1. NOSIE Total Assets Scores Over Time as a Function of Guardianship Group
Dedication
For my Dad . . .
You are missed muchly.
Acknowledgements
I had intended for graduate school to take me out of the Midwest for a while, but will be forever grateful I decided to come to Nebraska. I wish to express my gratitude to my advisor, William Spaulding, for just the right amount of direction and more than enough inspiration throughout my graduate career. I could not have found a better supervisory committee for this dissertation and am thankful for the collective wisdom of Mario Scalora, Brian Wilcox, and Robert Schopp. I am also grateful for the encouragement and mentorship from Cal Garbin and support of Mary Sullivan and the entire staff and all the participants at CTP. Finally, Mark Krejci from Concordia College deserves my thanks for steering me into psychology in the first place.
To the UN-L Serious Mental Illness Research Group, members past and present, I am grateful to have been among you. I especially want to offer many thanks to Jason Peer for raising the bar in our research lab just beyond my reach, so I was always striving for more. I am glad that both Jason Peer and Srividya Iyer were able to take the walk through graduate school with me, sharing times of struggle and success. This project would not have been possible if not for the NRSA fellowship awarded to Myla Browne, a member of the UN-L Serious Mental Illness Research Group. To those who continue to fund research in this much-needed and oft-overlooked area, you have my gratitude.
To my fellow transient Lincolnites, I will be forever indebted to you for maintaining my sense of humor and sanity and for always keeping my social calendar on powerbook and my costume drawer overflowing. My far-away-friends - you know who you are - have always “been there” for me and opened their hearts and doors at a moment’s notice; I couldn’t be luckier. My Mom and Dad have supported me 110% since Day 1. All the good things in me I got from them. My Mom has put up a brave front this past year so that I could continue to pursue my dreams and I am grateful for her courage. Finally, I am overwhelmed and overjoyed that Guy has stuck with me through the end of this adventure and is joining me for all the rest of life’s adventures.
Chapter 1 - Introduction
The Role of Guardianship in the Course of Treatment and Treatment Outcome for Individuals Recovering from Severe Mental Illness
The relationship between the legal system and mental health system is a tenuous one. Research on the two separate but intersecting fields has received increasing interest over the last fifteen years (e.g., Wexler, 1988; Wexler, 1990; Wexler & Winick, 1991; Winick, 1995; Monahan, Hoge, Lidz, Roth, Bennett, Gardner, & Mulvey, 1995; Slobogin, 1995; Spaulding, Poland, Elbogin, & Ritchie, 2000; Schopp, 2001). However, much of this research remains theoretical in nature with few applications in clinical settings. There are exceptions (for examples regarding coercion, see Hoge, Lidz, Eisenberg, Gardner, Monahan, Mulvey, et al., 1997; McKenna, Simpson, & Coverdale, 2003; regarding involuntary commitment, see Strachan, 2004; Ridgely, Borum, & Petrila, 2001; regarding legal status and self-report of symptoms, see Hopko, Averill, Cowan, & Shah, 2002). If the legal system allows itself to be informed by the current science of psychopathology and treatment, it may more fully serve its purpose of protecting the individual or society. Likewise, if mental health providers become more aware of the legal system and its impact on mental health care, they may better equipped to pass the benefit on to consumers to help them navigate the system and optimize their recovery. Bridging the gap between mental health law theory and application is the intention of this study. In particular, the role of guardianship as it pertains to the recovery from severe mental illness for people in psychiatric rehabilitation is evaluated.
Two main hypotheses were put forth towards the overall purpose of this study. First, it was hypothesized that people with guardians would demonstrate lower overall functioning at the time of admission and throughout the course of treatment across multiple domains when compared to those without guardians. Second, it was hypothesized that people with guardians would be discharged to more restrictive community placements than those without guardians, but that they would have a lower rate of rehospitalization.
An archival database constructed with nearly ten years of clinical data from an inpatient psychiatric hospital was used in this study. Clinical data on multiple levels of functioning, consistent with the multidimensional nature of impairments in SMI, was available for analyses. Data on symptomatology, neurocognitive and social cognitive functioning, treatment compliance, and behavioral functioning were used in analyses. In addition, data regarding discharge location and rates of rehospitalization following discharge were available. There were two main groups being compared in this study. First, there are those with court-appointed guardians, referred to as the Guardian group. Second, there are those without court-appointed guardians, referred to as the No Guardian group.
While these two groups are the primary focus of analyses, several other groups were evaluated since these distinctions were not mutually exclusive. Specifically, some people with guardians were admitted with a Voluntary per Guardian (VpG) legal status while others were admitted following civil commitment (CC). These legal statuses will be discussed further below. Therefore, there are subgroups based on legal status and guardianship: the VpG Guardian group and the CC Guardian group. Another variation on these groupings which was explored in analyses is based on whether the guardian was acquired before or after admission to treatment. The resulting groups from this distinction are: the Guardian Admission group and the Guardian Acquired group. A final variation which was explored was whether or not the guardian was a family member. This dissertation, then, takes on the following structure. First, current conceptualizations of SMI and optimal treatment practices are addressed. Next, there is a brief discussion of the intersection between the legal and mental health systems, including the concept of therapeutic jurisprudence. Then, a review of guardianship, the area of focus for this proposal, is presented. Finally, research relevant to this area is discussed. The study addresses several specific questions. First, this study addresses whether or not people with guardians differ in any way from those who do not have guardians with regards to specific areas of functioning including neurocognition, social cognition, treatment compliance, behavioral functioning, and symptomatology. This study aims to identify ways in which people with guardians differ from those without guardians at the time of admission to the psychiatric rehabilitation program and over the course of treatment. Second, this study identifies ways in which people with guardians differ with regards to outcome from those without guardians. Specifically, the rate of rehospitalization and the discharge location of those with guardians are compared to those without guardians. Finally, exploratory analyses delve into within-group differences regarding guardianship. That is, the impact of whether or not the guardian is a family member, whether or not guardianship was acquired before or after admission to the psychiatric rehabilitation program, and whether or not the legal status of the person is Voluntary per Guardian (VpG) or Civil Commitment (CC) is investigated.
Chapter 2 – Literature Review
Severe Mental Illness and Psychiatric Rehabilitation.
A majority of individuals categorized under the rubric of SMI are those diagnosed with schizophrenia spectrum disorders. Increasingly, research efforts in schizophrenia seek to understand schizophrenia not as a disorder with a single causal deficit but as a biosystemic disorder in which component processes are in a state of dysregulation (Spaulding, 1997). These components include processes related to neurophysiology (i.e., neurotransmitter systems), neurocognition (i.e., basic cognitive functions such as attention and memory), social cognition (i.e., higher order cognitive functions such as the formation of beliefs and abstract reasoning), and sociobehavioral functioning (i.e., performing behavioral activities in a socially meaningful context) (Spaulding, Sullivan, & Poland, 2003). Therefore, we understand the course of the illness, its nature, the deficits present, and the recovery process as reflections of varying degrees of dysregulation within the person (Ciompi, 1989; Spaulding, 1997; Spaulding et al., 2003) and between the person and their environment (Strauss, 1989). Dysregulation in one domain may affect functioning in another domain, in a precarious balance of reciprocal interactions (Spaulding, 1997). This differs from the dominant medical model perspective of mental illness as proceeding in a linear cascade from molecular levels of functioning to molar levels of functioning (e.g., a bacterial infection). Understanding severe mental illness as a multidimensional model of reciprocal processes calls for interventions which target multiple domains in order to re-regulate the system.
Significant advances have been made in specific psychopharmacological and psychosocial treatments for severe mental illnesses, including schizophrenia (see Spaulding et al., 2003; Hofman & Tompson, 2002; American Psychological Association [APA], n.d.; McEvoy, Scheifler, & Frances, 1999; Lehman, Thompson, Dixon, & Scott, 1995; Kendall, 1998). The advent of more effective antipsychotic medications in particular shaped service delivery for individuals with SMI, prompting a shift toward deinstitutionalization in the 1950s and 1960s. Greater emphasis was then placed on community care, which was not well-established at the time, with only brief hospital stays being necessary to stabilize symptoms with medication (Cook & Wright, 1995). Well-known today are the failures of deinstitutionalization with the resulting increase in criminalization and homelessness of the mentally ill. We have learned that antipsychotic medications are rarely sufficient in and of themselves to treat all levels of functioning impaired by the disorder. Researchers estimate that twenty-five to fifty percent of patients with schizophrenia experience residual medication resistant symptoms, which highlights the importance of psychosocial interventions (Garety, Fowler, & Kuipers, 2000; Spaulding, Johnson, & Coursey, 2001).
Despite the advances made in treatment strategies, outcomes remain modest (Wallace, Liberman, Kopelowicz, & Yaeger, 2001). As a group, people with schizophrenia have much in common, but at the same time there is considerable heterogeneity, with each individual representing a unique set of impairments, requiring multiple approaches in varying combinations. Any specific treatment addresses only a subset of an individual’s problems. Heterogeneous patient characteristics and multiple treatment approaches create complexity in choosing the appropriate application of treatment and contribute to modest treatment outcomes. Psychiatric rehabilitation has emerged as an approach to organize a diversity of treatments and target multiple levels of functioning.
Psychiatric rehabilitation is a complex, integrated approach which optimizes treatment outcome for those with severe mental illnesses by targeting multiple levels of functioning, consistent with the conceptualization described above. It is hardly a new concept, beginning nearly four decades ago (e.g., Anthony, Buell, Sharratt, & Althoff, 1972; Paul & Lentz, 1977). Psychiatric rehabilitation can include psychopharmacological interventions, contingency management, social skills training, occupational skills training, individual and group therapy, family intervention strategies, independent living skills training, and more.
As mental illness conceptualizations and treatment strategies have evolved, so have the laws, legal procedures, and legal roles surrounding them. The legal context of psychiatric rehabilitation creates more complexity. Often, individuals are subject to involuntary treatment and some have others making surrogate decisions on their behalf which may diminish their autonomy[1] and involvement in treatment. Considerations such as these can have a profound impact on how treatment and rehabilitation is provided and perhaps on the outcomes of treatment.
Intersection between Mental Health Care and the Law.
Both the legal system and mental health system have always shared at least one commonality: addressing deviance in society. Both are designed to address those behaviors which fall outside of a socially constructed range of normal, acceptable, or condoned behaviors. Care for those with mental illnesses had previously been reserved as a responsibility of the communities or families of individuals with mental illness. The legal system exerted little authority over the mental health care system – for better or for worse. Then, asylums were created to care for these individuals. In the United States, these asylums were characterized by some as a zoo-like public spectacle in the late 1700s and early 1800s (Whitaker, 2002). The era of moral therapy in the mid-1800s was trumped by the early 1900s eugenic movement which characterized the mentally ill as “unfit to breed.” Since these early times in the development of mental health care systems, the legal system has been juxtaposed between what it sees as protecting the rights or interests of the individual (e.g., deinstitutionalization) and the interests of society (e.g., civil commitment). A far cry from early history on mental health care, today, the legal system has not left any corner of the mental health care system unturned. We now have all kinds of psycholegal concepts: competency to stand trail, the Not Guilty By Reason of Insanity defense, parens patriae and police power civil commitment, outpatient commitment, guardianship, limited (or specific) guardianship, durable power of attorney, advanced directives, rights to treatment and rights to refuse treatment, and many others.[2] What is more, the arm of the law finds itself bent over mental health care to a far greater extent than it does even in general physical health care (Winick, 1996a).
Out of this mire of legal concepts in mental health care has emerged a unifying theoretical approach called therapeutic jurisprudence (TJ). TJ recognizes that the legal system itself acts as a social force which can impact mental health outcomes in both positive and negative ways and asks how the legal system may promote well-being (Schopp, 1999; Slobogin, 1995; Wexler & Winick, 1991; Wexler, 1990; in severe mental illness, Spaulding et al., 2000). TJ, then, becomes a framework for research investigations such as the one proposed here to determine how the legal system can be therapeutic, or, on the other hand, anti-therapeutic.
Guardianship.
While the legal system has not always exerted authority over mental health care, guardianship has long been a traditional practice if not a codified one. Historically, common law in Rome and England allowed for the nation to exert authority over an incompetent person’s estate. Parens patriae, from which the guardianship process gains authority, was borne out of this tradition (Reisner, Slobogin, & Rai, 1999). Parens patriae is essentially the provision for the government to act as a parent to its citizens for the expressed purpose of preventing harm. Guardianship, then, allows the decision-making power for adults with mental disabilities who are deemed incompetent by the court (or children, due to their minority status alone) to be given to a designated person who is often a family member, but may also be a non-family member (e.g., lawyer, member of an advocacy group).
Guardianship varies by jurisdiction and many forms of guardianship have developed (in Nebraska, see Neb.Rev.St. §30-2619 & §30-2630). For example, the scope of guardianship may vary; it may be full (general) or limited (specific).[3] Full guardianship allows the appointed guardian the authority to make decisions in all areas conferred upon by law while limited guardianship refers to any guardianship which is limited to certain areas of decision making, presumably those in which the individual is deemed incompetent (e.g., estate management; Neb.Rev.St. §30-2601). For the purpose of the proposed study, guardianship with the afforded decision making authority over at least treatment-related issues is of focus. Limited guardianship is an appealing, but not often exercised, mechanism which reflects that being incompetent in one area of life does not mean that an individual is incompetent in all areas of life. It allows an individual to make self-regarding decisions in areas of life in which competence remains intact.
In addition, the duration of guardianship varies. Some states limit the time frame over which the period of guardianship lasts or requires that the finding of incompetence be reevaluated. However, in other states, the finding of incompetence is indefinite unless someone (the ward or interested party) files a petition for a hearing to determine that the individual is no longer incompetent (Neb.Rev.St. §30-2623).
The nature of the decision-making guardians use also varies. Sometimes, guardians as surrogate decision-makers may be instructed to make the decision in an individual’s best interests (which may or may not be the decision the individual would have made if he or she was competent at the time) and, sometimes, guardians are instructed to make a substitute decision (making the same decision the individual would have made were he or she competent to do so, which may or may not be in the individual’s best interest) (Reisner et al., 1999).
Finally, the basis for guardianship varies. Without a finding of incompetence, the previous points are moot since a court will not appoint a guardian without it. Originally, criteria primarily revolved around whether or not the person could care for himself or herself. Later, however, the Uniform Probate Code was developed and it remains in widespread use. The Uniform Probate Code was designed to focus on the quality of cognitive processes used in decision-making. That is, an incapacitated person is one who “by reason of mental illness, mental deficiency, physical illness or disability, or advanced age . . . or other cause . . . lacks sufficient understanding of capacity to make or communicate responsible decisions concerning his person” (Uniform Probate Code §5 as quoted in Reisner et al., 1999, p.869). The focus here is not on the decision that is made, but rather on the process of making it. As an alternative to the Uniform Probate Code, some states use a “functional” approach which harkens back to original criteria and involves the impairment of ability to perform “minimal” activities of daily living (Reisner et al., 1999).
Related Studies.
While few studies have examined guardianship, several have looked at the relationship between mental illness and competency to consent to treatment. In particular, the MacArthur Treatment Competence Study was designed specifically to address the concerns of policymakers and clinicians alike regarding the decision-making capacities of people with mental disorder. The study was conducted at three sites by Paul Appelbaum, Thomas Grisso and colleagues. The articles generated by this study examine the decision-making capacity of patients hospitalized with mental illness as it compares to patients hospitalized for medical illness and a matched sample of non-patients from the community (Appelbaum & Grisso, 1995; Grisso, Appelbaum, Mulvey, & Fletcher, 1995; Grisso & Appelbaum, 1995).
Appelbaum and Grisso began their study by identifying the legal standards associated with the determination of competence and developing measures to assess decision-making abilities in those areas. Roth and colleagues were among those to originally discuss the evolving legal standards for determining competence (Roth, Meisel, & Lidz, 1977). The standards were later refined by the work of Roth, Appelbaum, and Grisso, among others (Roth et al., 1982; Appelbaum & Roth, 1982; Drane, 1984; Tepper & Elwork, 1984; Grisso, 1986; Appelbaum & Grisso, 1988). The four areas of decision-making competence identified as relevant to the legal system are as follows: the ability to state a choice, to understand relevant information, to appreciate the nature of one’s own situation and the potential consequences, and to reason rationally with information (Appelbaum & Grisso, 1995). These four legal standards for assessing competence to consent to treatment were used to develop the three MacArthur Treatment Competence Research Instruments, Understanding Treatment Disclosures (UTD), Perceptions of Disorder (POD), and Thinking Rationally About Treatment (TRAT; Grisso, Appelbaum, Mulvey, & Fletcher, 1995). While the authors have deemed these instruments inappropriate for use in a clinical context due to the time and complexity associated with their administration, they have developed a tool for use by clinicians called the MacArthur Competence Assessment Tool-Treatment (MacCAT-T; Grisso, Appelbaum, & Hill-Fotouhi, 1997). It has since been used in studies regarding decision-making in psychiatric contexts (e.g., Palmer, Nayak, Dunn, Appelbaum, & Jeste, 2002).
The results of the initial studies conducted using the UTD, POD, and TRAT indicated that people diagnosed with schizophrenia and depression did indeed show deficits in their decision-making abilities (Grisso & Appelbaum, 1995) as compared to medical patient and community non-patient control groups. In particular, people diagnosed with schizophrenia scored significantly lower than their matched non-patient community counterparts, with 52% showing impairment in at least one area, in contrast to 12% of medical inpatients and 4% of community non-patients. People diagnosed with depression likewise showed impairment in decision-making, though to a lesser degree.[4] Despite the overall finding of impaired decision-making, there was considerable heterogeneity among the group hospitalized for mental illness leading to the conclusion that one cannot make a blanket statement regarding impairment in decision-making solely on the basis of having a mental illness.
In fact, Grisso and Appelbaum (1995) found that a majority of people with schizophrenia demonstrated adequate decision-making capabilities, as defined by the researchers, in at least one of the three areas assessed. For any one measure, only one out of four people with schizophrenia scored in the impaired range. This finding is important since many jurisdictions use only one or a combination of the four legal standards discussed above as opposed to all four. In addition, even when considering all measures, nearly half of those diagnosed with schizophrenia performed adequately on the four measures combined.
A subgroup emerged to explain, in part, the lower scores on the decision-making measures by the group of people diagnosed with schizophrenia. Grisso and Appelbaum (1995) identified that those who demonstrated deficits were by and large those with more severe symptomatology (e.g., delusions, disorganized thinking) and when assessed after a two-week period of treatment (i.e., hospitalization with antipsychotic medication), improvement in all decision-making areas was evident for those whose symptomatology had decreased.
The current study takes place in a context in which the vast majority of participants are diagnosed with schizophrenia and have arrived in treatment because they have been civilly committed or because of a decision made by a guardian to hospitalize. In either case, the competence of the person to consent to treatment has been questioned. The MacArthur studies are relevant to the current study in that they identified impairment in decision-making for people with mental illness, especially those with schizophrenia. However, they also identified considerable heterogeneity in decision-making abilities among those with mental illness, leaving important questions about whose competence regarding treatment decisions should be questioned. There are several key differences between the research of Appelbaum and Grisso and the current study. In particular, the current study takes place in the context of a psychiatric rehabilitation program lasting between nine and 18 months for people with a chronic mental illness rather than an acute setting. In addition, participants in this program are admitted only after a period of stabilization in an acute context such that the decrease in symptomatology with medication as seen in the Grisso and Appelbaum (1995) sample is unlikely to occur since acute symptoms have already diminished prior to admission. Furthermore, in the current study, treatment is considered involuntary in nature whereas in the Grisso and Appelbaum (1995) study, 55.9% of those diagnosed with schizophrenia were considered voluntary admissions while only 27.5% were considered involuntary and 16.7% were considered emergency admissions. These key contextual differences offer the current study an important new set of data from which to consider the questions of competency and decision-making.
Grisso and Appelbaum (1995) accurately point out that the competence to consent to treatment is almost exclusively only questioned when an individual refuses treatment. However, their study suggests that a far greater number of people consent to treatment without adequate decisional capabilities than refuse treatment (Appelbaum & Grisso, 1995). Aside from the contextual differences described above, the current study is also unique in that it employs a measure of treatment compliance which will help to delineate how treatment compliance is associated with different legal statuses of individuals, including those who are civilly committed and/or those who have guardians. While this study does not include an explicit measure of decision-making such as the measures developed and used in the MacArthur studies, a measure of insight which includes subscales regarding need for treatment and awareness of illness as well as extensive assessments of cognitive functioning is included in the database.
Clinical research on guardianship for individuals with SMI is limited. Some work has been done in the UK, though these limited studies remain mostly descriptive in nature, regarding patterns of use of guardianship (e.g., Shaw, Hatfield & Evans, 2000, Hatfield, Bindman, & Pinfold, 2004). Most research in the area of mental health law as it pertains to individuals with SMI has been in the related area of civil commitment. There is overlap in theory and in practice between the processes of guardianship and civil commitment, but they remain distinct legal processes.
One main distinction between guardianship and civil commitment statutes, as far as they pertain to treatment, is regarding competency. While a finding of incompetence is a necessary component for the appointment of guardianship, this is no longer the case for civil commitment. Prior to 1970, parens patriae civil commitment and guardianship were indistinguishable with regards to treatment (Reisner et al., 1999). In fact, civil commitment carried with it the presumption of incompetence, “de facto.” Today, however, in most jurisdictions, including Nebraska, civil commitment does not require a finding of incompetence. In fact, there is an explicit rejection of incompetence in many civil commitment statutes. For example, the Nebraska civil commitment statute (§83-1068) specifically states that, “subjects in custody receiving treatment shall have the right to be considered legally competent for all purposes unless they have been declared legally incompetent.” Furthermore, the mental health board (or other decision-making body) does not have the authority to declare incompetence; a separate hearing regarding competency must be held to do so. However, the statute also goes on to declare that an individual who is civilly committed retains the right “to refuse treatment, except such treatment as is essential in the judgment of the medical health professional in charge…to…substantially improve [the patient’s] mental illness” (Neb.Rev.St. §83-1068). This raises the question, what does it mean to be afforded the right to refuse treatment if you cannot refuse treatment that somebody else thinks will help? It appears that the true underpinnings of civil commitment remain an apparent assumption that people with mental illness lack adequate decisional capacity – at the very least, in areas regarding treatment – consistent with the de facto determination of incompetence civil commitment previously held. Herein lies the problem. Current statutes in the area of parens patriae civil commitment are incoherent. While explicitly denying the presumption of incompetence for people with mental illnesses, the implicit premise in the statutes is the assumption of incompetence. If it is, in fact, an empty promise, Schopp (2001) questions how civil commitment under parens patriae authority is warranted.
After all, recall that parens patriae authority is a provision for the government to intervene to prevent its citizens from coming to harm. If someone poses a danger to only his or her self, the state’s only justification for intervention is that the person is incompetent to make that decision. Whereas, if someone poses a danger only to his or her self and is competent, the state has no justification for intervention under parens patriae authority.
Essentially, then, guardianship and civil commitment differ, at least semantically, in the inclusion (guardianship) or exclusion (civil commitment) of a legal finding of incompetence. The distinction between the two is not to say that an individual with a guardian cannot be civilly committed, or, vice versa, that someone who is civilly committed can not have a guardian, but the means by which they enter and exit treatment settings may differ. Civil commitment is sought for the sole purpose of involuntary treatment for an individual whereas guardianship may invade the individual’s autonomy on multiple levels, including treatment-related decision making. Through the lens of therapeutic jurisprudence, we ask the questions: Is the distinction between the two a helpful or therapeutic one? Does it serve a functional purpose? However, it is no longer helpful to ask these questions outside of a clinical context. We must begin to ask questions such as: What are the correlates of legal status to clinical functioning? How do we identify which legal mechanism is the most beneficial for a given individual? We cannot begin to answer questions about the utility or efficacy of these legal mechanisms without a greater understanding of the clinical characteristics of individuals for whom these legal mechanisms are designed.
Guardianship has been posed as an alternative to civil commitment and this raises some important questions about the clinical characteristics of those with guardians and those without guardians (Spaulding et al., 2000). The proposed study does not intend to advocate for one approach or the other. Rather, it intends to provide a scientific basis for answering the relevant questions. For example, do those with guardians have demonstrably different functional capacities than those who do not have guardians, which would suggest a need for different legal mechanisms? There are reasons to think that people with SMI in a psychiatric rehabilitation program with guardians differ from those without guardians, and there are reasons to think that they don’t. One might presume that since a person with a guardian has been found incompetent to make self-governing decisions in some area of life, this person has a lower level of functioning compared to someone who has not been found incompetent. One might see this difference become apparent in the course of illness. That is, the long-term nature of a full (as opposed to limited) guardianship implies more pervasive and chronic impairments in that person such that a surrogate decision-maker is needed to help the individual navigate through life. On the other hand, the fact that civil commitment typically ends when the period of hospitalization ends implies that these individuals have a more cyclic course involving periods of recovery in which they may effectively govern all aspects of their own lives independently. On the continuum of autonomy afforded them, then, one would expect these two populations to differ functionally with those with guardianship having relatively lower overall levels of functioning. However, it may also be anticipated that at the beginning of treatment differences in functioning are imperceptible since all people have experience an exacerbation of their illness at that time; only as treatment progresses will differences emerge with those with guardians demonstrating less recovery in areas of impairment indicating more constant or chronic functional impairments than those without guardians.
However, there are also reasons to believe these populations do not differ. The reasons for which an individual interacts with the legal system are complicated and there is no one straight road to guardianship, civil commitment or treatment, in general. For example, family involvement for people with SMI varies greatly. If one person has an active family and another does not, the former may be more likely to have a guardian than the latter, independent of their functional abilities, simply because a family member sought it. Or, an individual may repeatedly receive involuntary inpatient treatment based on a series of civil commitments, typifying a more chronic course that may be better served through guardianship, but no interested party seeks guardianship on behalf of the individual. Convolutions such as these suggest that we need to know more about the legal process and its correlates to the clinical picture for people with SMI.
To date, there are no empirical studies conducted in a clinical setting that compare the characteristics of people with court-appointed guardians to those who do not have guardians. A major factor contributing to the paucity of research in this area is the substantial methodological problems associated with this kind of research. The impact of legal mechanisms on SMI likely takes years to demonstrate full, measurable effects. Likewise, then, it takes years to accumulate the needed data in order to be able to understand this process.
Another important consideration of this type of research is the clinical context. The psychiatric rehabilitation setting is an appropriate starting place for a study of this nature. Individuals in this setting have almost exclusively arrived there through civil commitment or guardianship proceedings, making this the population most affected by these processes. Also, because the rehabilitation setting targets multiple levels of functioning, it is most likely to bring about change in the things associated with the need for guardianship or civil commitment. For example, it targets aspects of neurocognitive functioning associated with decision making, such as attention and executive functioning.
The archival database to be utilized in the proposed study is ideal for this type of research. It affords the opportunity to study two populations under a similar set of clinical circumstances. For the vast majority of people in the psychiatric rehabilitation setting from whose data this database is constructed, treatment is typically viewed as “involuntary.” That is, they have either been civilly committed (CC) by the state or a guardian has made the decision regarding admission for treatment, likely without the full support or consent of the person, resulting in a “Voluntary” per Guardian (VpG) admission. Significantly, this data has been accumulated over the course of seven years, allowing for a preliminary look at the therapeutic consequences of these legal mechanisms and the longitudinal effects of psychiatric rehabilitation treatment for these individuals. It must be reiterated that the proposed study is highly exploratory in nature since there is no previous research on the clinical differences between those with and without guardians. This is a critical first step in the analysis of the role of guardianship in the treatment of SMI.
Present study.
The present study has two primary objectives:
1. To determine if there is a pattern of differential functioning for individuals with or without guardians across various levels of functioning. In general, it is anticipated that individuals with a guardian will demonstrate lower overall functioning than those without guardians. For the most part, those in the “other” group consist of individuals who have been civilly committed, but do not have a guardian. As described above, these are individuals for whom the court maintains, at least in theory, that they are presumed competent. For those with guardians, however, there has been a legal finding of incompetence. Because of this, it is predicted that those individuals who have a guardian will show demonstrably lower functioning in the clinical setting. Therefore, it is hypothesized that the legal distinction will correlate with clinical functioning. Specifically, the groups will differ in statistically significant ways across all domains of functioning measured (neurocognitive, social cognitive, symptomatology, behavioral functioning, and treatment compliance).
A. Upon Admission. It is predicted that at the time of admission, individuals with guardians will demonstrate lower overall functioning than those without guardians on neurocognitive, social cognitive, and behavioral functioning measures. In addition, it is anticipated that those with guardians will have lower levels of treatment compliance and higher levels of symptomatology.
B. Over the course of treatment. It is predicted that overall differences in functioning, symptomatology, and treatment compliance will remain over the course of treatment. It is hypothesized that as a result of rehabilitation there will be an increase in neurocognitive functioning, insight into disorder, internal locus of control, and behavioral functioning and a decrease in external locus of control, symptomatology, and treatment compliance for those with and without guardians. That is, it is anticipated that both groups will show improvement in functioning over the course of rehabilitation. However, while the overall pattern of improvement will not differ between groups, it is predicted that significant differences will remain between those with guardians those without guardians in all areas with those individuals with guardians demonstrating lower functioning, higher symptomatology, and lower treatment compliance.
2. To determine if there is a pattern of differential outcome following discharge for individuals with guardians from those without. While previous research does little to inform hypotheses in this area, it is generally hypothesized that there will be differences between groups in terms of treatment outcome. Specifically:
A. Discharge disposition. It is predicted that discharge disposition will differ between the two groups in that individuals with guardians will be discharged to a more restrictive setting. This hypothesis is based on the idea that the legal status of the individual may influence treatment providers’ notions such that there is an assumption that individuals with guardians require higher levels of care and supervision.
B. Rehospitalization rate. It is predicted that there will be a greater rate of rehospitalization for those without guardians than those with guardians. This hypothesis is founded on the functional basis for guardianship. That is, if an individual has a guardian, the state has seen to it that this person is “protected” whereas an individual without a guardian is not afforded this same protection.[5]
Based on the results of these analyses, exploratory analyses will be undertaken to glean additional information about the role of guardianship in the psychiatric rehabilitation of individuals with SMI.
Chapter 3 - Method
Design Overview.
The primary purpose of this study was to assess the relationship of guardianship to functioning before, during, and after treatment. The participants with guardians were compared to those without guardians to determine if any differences existed. Univariate and multivariate analyses were conducted within and between the two groups across with respect to overall functioning using multiple measures described below. In addition, within-group analyses of the participants with guardians were conducted.
Setting.
The Community Transition Program (CTP) is an inpatient unit at the Lincoln Regional Center (LRC) a public state psychiatric hospital in Lincoln, Nebraska. This 40-bed inpatient unit hosts a comprehensive psychiatric rehabilitation program for those most disabled by mental illness in the State of Nebraska. Individuals are typically discharged to a less restrictive setting after a 9 to 18 month period of intensive treatment. Treatment engagement is fostered through the use of contingency management with a backbone in social learning theory. The regimen includes pharmacotherapy, psychoeducational groups and classes to target improved management of symptoms and disorder, and training aimed at improving occupational, leisure, and social skills. The treatment is designed to target multiple levels of functioning for individuals with SMI, rather than only targeting an isolated area of dysregulation (e.g., symptoms). As such, treatment does not focus on clinical diagnosis. Rather, treatment plans are individualized and based on making incremental gains in compromised areas of functioning and capitalizing on areas of relative strength of the individual. Treatment plans are designed by a multidisciplinary treatment team including nurses, social workers, psychiatrists, and psychologists. In addition, program participants are encouraged to be active members of their own treatment team for the purpose of increased engagement in treatment. Hence, the CTP refers to patients as “participants” instead of patients. Treatment and Rehabilitation of Severe Mental Illness (Spaulding et al., 2003) outlines the theoretical underpinnings and practical applications of this innovative, state-of-the-art treatment technology.
The Lancaster County Community Mental Health Center (LCCMHC) also contributed data to this study. Since a majority of participants from CTP are served by LCCMHC upon discharge, data collected as part of ongoing program evaluation at LCCMHC pertaining to outcome such as rehospitalization rate, discharge location, and community functioning were obtained from LCCMHC records.
Participants.
Participants discharged from the CTP program from 1996 through December 2004 contribute data to the archival database utilized in this study. The CTP program participants were not directly involved in any specific research procedure. Rather, the database includes the ongoing clinical data collected as part of the routine assessment process described further below at the CTP and additional data gathered thro ugh chart review.
As a criterion for admission, participants of the CTP have a chronic, but stable Axis I major mental disorder. This sample represents a severe and treatment refractory subpopulation. These are individuals for whom other short-term treatments have proved unsuccessful and who are now involved in an intensive long-term rehabilitation program.
Table 1
Demographic and Clinical Characteristics of Participants (N=162)
|Demographic and Clinical Variables |N |Mean (SD) or Percentage|
|Age (range: 19 to 66) |162 |40.45 (11.83) |
|Education, # of years in school (range: 7 to 19 years) |158 |12.47 (2.14) |
|Length of stay at CTP in days (range:145-2745 days) |162 |659.54 (373.99) |
|Age at first hospitalization (range: 4[6] to 50) |158 |21.04 (8.28) |
|Number of hospitalizations in lifetime (range: 1 to 15) |156 |9.09 (4.44) |
|Gender (n=162) | | |
| Male |87 |53.7% |
| Female |75 |46.3% |
|Race/Ethnicity (n=162) | | |
| Caucasian |144 |88.9% |
| African American |11 |6.8% |
| Hispanic |2 |1.2% |
| Native American |- |- |
| Asian American |2 |1.2% |
| Other |3 |1.9% |
|Marital Status (n=162) | | |
| Single |104 |64.2% |
| Married |8 |4.9% |
| Divorced |43 |26.5% |
| Widowed |3 |1.9% |
| Separated |4 |2.5% |
|Legal Status (n=162) | | |
| Civil commitment (CC) |118 |72.8% |
| Voluntary per guardian (VpG) |35 |21.6% |
| Not responsible by reason of insanity (NRRI) |5 |3.1% |
| Voluntary (V) |4 |2.5% |
|Primary Axis I Diagnosis (n=162) | | |
| Schizophrenia, Paranoid Type |55 |34.0% |
| Schizophrenia, Chronic/Undifferentiated Type |49 |30.2% |
| Schizoaffective |37 |22.8% |
| Bipolar |8 |4.9% |
| Psychotic Disorder NOS |1 |.6% |
| Pervasive Developmental Disorder |1 |.6% |
| Impulse Control Disorder |1 |.6% |
| Other |10 |6.2% |
|Axis II Diagnoses (n=156) | | |
| None |79 |50.6% |
| Borderline |10 |6.4% |
| Paranoid |12 |7.7% |
| Antisocial |6 |3.8% |
| NOS |31 |19.9% |
| Other |18 |11.5% |
Diagnoses largely consist of schizophrenia spectrum disorders (87%). In addition, 25.3% have comorbid substance abuse (22.8%) or dependence (2.5%) diagnoses. Table 1 summarizes the major demographic and clinical characteristics of the participants.
Initially, the archival database included 177 participants. However, 12 of these admissions were “return participants,” meaning they had previously completed treatment at CTP. Because available outcome data in the database pertains to the most recent admission to CTP, all previous admissions before the most recent one were excluded from analyses to maximize the number of valid cases available for analyses. After excluding participants with multiple admissions, 165 participants remained. Three of the 165 participants stayed less than two months in the CTP program and were therefore excluded from analyses since treatment effects could not be evaluated as there was not adequate time to participate in the CTP rehabilitation program. The resulting sample size for use in this study was 162.
CTP participants are ideally suited for this study because a vast majority are either civilly committed or admitted by a legal guardian. In either case, treatment is typically viewed as involuntary in that most do not decide for themselves to enter the program. For every individual who is admitted to the CTP program with a VpG legal status, three are admitted via CC. In addition, as described further below, many of the people who are civilly committed also have a guardian. This study examines guardianship in particular. It is estimated that one third to one half of CTP participants are affected by guardianship.
Measures.
Demographic and Clinical Characteristics.
Through chart review at CTP, information regarding the demographic and clinical characteristics of participants was included in the archival database. These variables included gender, age, number of years in school, race/ethnicity, marital status, legal status, number of previous hospitalizations in the participant’s lifetime, age of first hospitalization, Axis I and Axis II psychiatric diagnoses, length of stay at CTP, and other relevant variables.
Neurocognition Measures.
1) Repeatable Battery for the Assessment of Neuropsychological Status (RBANS; Randolph, 1997). The RBANS is a brief neurocognitive screening test (approximately 25 minutes) ideal for individuals who may lack attentional capacity for longer assessments. There are two forms of the RBANS (Form A and Form B), which intend to minimize the practice effects of repeated administrations. There are twelve subtests on the RBANS which are grouped into five neurocognitive domains. The RBANS yields five domain scores including immediate memory, delayed memory, attention, language, and visuospatial/constructional functioning. In addition, a total index score represents overall cognitive functioning. The convergent validity of the RBANS with other neuropsychological constructs, like memory and intelligence, has been established in people diagnosed with schizophrenia (Gold, Queern, Iannone, & Buchanan, 1999; Hobart, Goldberg, Bartko, & Gold, 1999). In addition, sensitivity to patterns of cognitive impairment in SMI and general reliability and validity have been evaluated (Gold et al., 1999; Hobart et al., 1999). The RBANS total score is the primary measure of interest in this study.
2) COGLAB (Spaulding, Garbin, & Dras, 1989). COGLAB was created as a computerized test battery comprised of tests common in the psychopathology literature. Currently, CTP standard assessment includes two of the tests from the battery, focusing on attention and executive functioning - the Span of Apprehension (SPAN) task and the Card Sorting Task (CST; an adaptation of the Wisconsin Card Sorting Task). SPAN hits and false alarms provide a measure of attention and CST perseverative and random errors serve as a measure of executive functioning. Good discriminant validity was demonstrated between individuals diagnosed with schizophrenia and controls in an early study of the COGLAB (Spaulding et al., 1989). The results from large multivariate studies conducted with normal and patient populations indicate overall acceptable psychometric properties of the COGLAB (Spaulding, Hargrove, Crinean, & Martin, 1981; Spaulding et al., 1989).
3) Rey Auditory Verbal Learning Test (RAVLT; Schmidt, 1996). The RAVLT , is a seven trial list-learning task with alternative forms consisting of 15 words presented in an auditory format. Participants are instructed to recall as many words as they can from the list immediately following each of five trials. A distractor trial is then presented consisting of a different list of 15 words and participants are required to recall as many words from this distractor list. Finally, participants are required to recall as many words as possible from the original list without it being presented again. In general, the RAVLT provides a measure of verbal memory. The number of words remembered after the fifth trial is the most commonly used RAVLT score. Good discriminant validity between memory impaired vs. memory intact patients and normal vs. neurological patients has been demonstrated, as well as adequate test-retest validity (Schmidt, 1996).
4) Rey-Osterrieth Complex Figure Test (RCFT; Rey, 1941): This is a test of visuoconstructional ability and nonverbal memory. It is comprised of four tasks: a copy trial, immediate recall trial, delayed recall trial, and a recognition task. Figures are scored using the 18-point scoring system, originally developed by Osterrieth (1944), and outlined in Meyers & Meyers (1995).
5) The Trailmaking Test (A&B) (Army Individual Battery, 1944). This two part test assesses attention, visual scanning, and information processing. In Part A, individuals connect circles numbered 1 through 25 by drawing a line sequentially from beginning to end as quickly as they can. In Part B, individuals complete a similar task alternating between numbers and letters. Performance is measured by subtracting the number of errors from the total possible score, resulting in two scores, one for each trial.
Social Cognition Measures
1) Inventory for the Measurement of Self-Efficacy and Externality (I-SEE; Krampen, 1991). The I-SEE provides a measure of global attributional style, or locus of control. It is comprised of 32 items which are each rated on a six point Likert scale ranging from strongly disagree to strongly agree. The measure consists of four primary scales: “internality” (i.e., “whether I have an accident is based on my own behavior”); “self-concept of one’s own competence” (i.e. “I can do many things to protect my self-interest”); “powerful others’ control beliefs” (i.e., “other people often prevent my plans from becoming reality”); and “chance control beliefs” (i.e., “whether I fall ill is a matter of chance”). These scales are combined to yield two composite scales a general external scale (“externality”) and a general internal scale (“self-concept of one’s own efficacy”) which represents participant’s beliefs about their self-efficacy. Krampen (1991) established reliability for the two composite scales.
2) Insight Scale (IS; Birchwood, Smith, Drury, Healy, Macmillan, & Slade, 1994). This brief self-report measure (8 items) allows participants to choose one of three responses: agree, disagree or unsure, for each item. It yields a total score and three subscale scores representing David’s (1990) three domains of insight. The three subscales, therefore, are: “need for treatment” (i.e., “I do not need medication”), “ability to relabel psychotic experiences” (“some of my symptoms were made by my mind”), and “awareness of illness” (“I am mentally well”). This measure of insight focuses on insight into functional impairment rather than specific illness categorizations. Internal consistency and test-retest reliability were demonstrated by Birchwood et al. (1994).
Behavioral Functioning Measures.
1) Nurse Observation Scale for Inpatient Evaluation (NOSIE-30; Honigfeld, Roderic, & Klett, 1966). This measure is a 30-item behavioral checklist format completed by nursing staff at CTP has responses from zero (never) to four (always). Two psychiatric technicians complete the checklist weekly for each participant. Items cover six areas of ward functioning: social competence (“refuses to do ordinary things expected of him or her”), social interest (“tries to be friendly with others”), neatness (“keeps clothes neat”), irritability (“gets angry or easily annoyed”), psychoticism (“talks, mutters, or mumbles to self”), and motor retardation (“is slow-moving or sluggish”). When summed, these six areas represent a total assets score. Adaptive functioning scales (i.e., social interest) are positively weighted and maladaptive scales (i.e., irritability) are negatively weighted when determining the total assets score. This measure has been widely used as part of the psychiatric rehabilitation treatment and is a routine assessment in the treatment program. Analyses within the CTP population have yielded Pearson correlations between 0.68 and 0.72 for all scales (Penn, Mueser, Spaulding, Hope, & Reed, 1995; Spaulding et al., 1999b). Also, a recent reliability update of this measure confirms it remains reliable in modern treatment settings, with inter-rater reliability on the total assets score at 0.76, on maladaptive scales at 0.68, and adaptive scales at 0.75 (Lyall, Hawley, & Scott, 2004).
Symptomatology Measures.
1) The Brief Psychiatric Rating Scale – Extended Versions (BPRS-E; Lukoff, Nuechterlein, & Ventura, 1986; Van der Does, Linszen, Dingemans, Nugter, & Scholte, 1993). To evaluate symptoms, the BPRS-E is used routinely at CTP. In general, the BPRS-E is widely used to assess changes in psychiatric symptoms. Using a Likert scale from 1 (symptom is not present) to 7 (symptom is very severe), clinicians rate individuals based on interview content and general behavior on 24 items. The BPRS-E is a widely used instrument and reliability and validity have been demonstrated across several studies (Bailley, Lachar, Rhoades, Diefenbach, Espadas, & Varner, 2004). Reliability analyses in one study yielded Cronbach’s alpha of 0.76 (Perlick, Rosenheck, Clarkin, Sirey, & Raue, 1999). Factor analyses on the former 18-item version and the newer 24-item version have yielded four, five, and six factor solutions of symptom items (e.g., Guy, 1976, Spaulding, Reed, Sullivan, Richardson, & Weiler, 1999a; Panos, 2004; Perlick, et al., 1999; Burger et al., 1997). The six-factor solution validated by Spaulding, Fleming, Reed, Sullivan, Storzbach, & Lam (1999a) was used in this study because the original validation took place with the same population. A standard principal component analysis of the BPRS (Spaulding et al., 1999a) yielded six factors: Psychotic Disorganization, Hallucinations/Delusions, Paranoia, Emotional Blunting, Agitation/Elation and Anxiety/Depression.
Treatment Compliance Measure.
1) Rehabilitation Noncompliance (RNC) As part of the rehabilitation program for all CTP participants, data are kept on the extent to which they are participating actively in the rehabilitation program. Instances of noncompliance may include not attending to activities of daily living, refusing to attend groups and classes, not following program rules, and the like. These data are often used in the context of the contingency management programs which are incorporated into most CTP treatment plans. The CTP program infrastructure includes a computerized system for collection of this behavioral data based on the contingency management interventions. The data collection is subject to monthly fidelity checks by trained graduate students to ensure behaviors are being recorded accurately and contingencies are being implemented correctly. Rehabilitation Noncompliance (RNC) data is collected on a weekly basis and, for the purposes of this study, are measured as an average number of weekly instances of noncompliance during the last three months of each six month period. The last three months were selected because behavior management programs are established during the first month of treatment and further refined during the first few months of treatment making initial RNC data less reliable due to frequent changes in programs. Choosing six month period is primarily due to the fact that it coincides with the assessments completed at CTP at six month intervals.
Outcome Measures.
1) Rehospitalization rate. One of the most primary goals of psychiatric rehabilitation is the prevention of future hospitalizations and the decrease of inpatient hospital days and use of emergency services (Iyer, Rothmann, Vogler, & Spaulding, 2005; Cook, Pickett, Razzano, Fitzgibbon, Jonikas, & Cohler, 1996; Anthony, Cohen, & Vitalo, 1978). The inclusion of data from LCCMHC is a critical part of analyzing outcome from the CTP program since a majority of participants are served through LCCMHC upon discharge from CTP. Recent program evaluation activity at the LCCMHC has established a rehospitalization data tracking program and the data available here was cross-checked with chart reviews at CTP and chart reviews and interviews with staff at LCCMHC. While outcome is often calculated on the basis of number of rehospitalizations or the amount of days in the hospital following discharge, this method can potentially be confounding. Since the archival database in this study includes people discharged from CTP back in 1996 all the way up to people discharged from CTP in December 2004, people may range in the amount of time since discharge from 150 days to 3,285 days. Therefore, the number of hospitalizations or hospital days as a measure of outcome may be misleading because the amount of possible time in the community varies. To control for this variability, the rehospitalization rate may be operationalized in several different ways. For example, Paul and Menditto (1992) recommend that rehospitalization rate will be derived based on the amount of continuous community tenure during the first six months following discharge in order to control for different amount of times since discharge. Setting an arbitrary timeframe such as Paul and Menditto suggest of six months, nine months, or twelve months following discharge is one way to address the potential time confound. This study also explores other ways of addressing this such as the percentage of hospital days out of all hospital and non-hospital days since discharge and the survival rate (or how long before the first rehospitalization).
2) Discharge disposition. The restrictiveness and nature of the setting to which individuals are discharged from the CTP will also serve as a measure of outcome. There are at least 35 different discharge locations to which people go following CTP. Less restrictiveness (e.g., independent living) is characteristic of better outcome. For the purposes of this study, there were essentially four categories of discharge location from most to least restrictive: 1) Regional Center transfer, 2) Psychiatric Residential Rehabilitation, 3) Assisted Living, and 4) Independent living. These categories were conceptualized based on interviews with the CTP program director and CTP social workers who are primarily responsible for discharge planning and most familiar with community services. The broadest of all these categories is Assisted Living because the levels of care within the Assisted Living category in the state of Nebraska vary considerably. These four categories do not encompass all discharges from CTP. For example, the categories do not include nursing homes or developmentally disabled (DD) group homes. Very few individuals are discharged to nursing homes, which would be considered more restrictive than even psychiatric residential rehabilitation, but discharges such as this from CTP are rare and typically due to the person’s medical rather than psychiatric condition, so are not included here. Likewise, very few people are discharged to DD settings because one of the exclusion criteria from admission to CTP is developmental disability. However, following admission, a few people have qualified for DD services. These discharges are not included in analyses since their discharge services are conceptually different than a majority of participants at CTP.
Procedure.
Data Collection.
An archival database was used in this study. Most of the archival data comes from the CTP clinical archives, but additional data pertaining to outcome and community functioning was collected from LCCMHC. Clinical data are routinely collected as part of the CTP program and contribute significantly to the database. In addition, extensive chart review at both CTP and LCCMHC added additional data regarding hospitalization history and general clinical and demographic characteristics. Data from 162 participants discharged from the CTP program between 1996 and December 2004 were used in analyses.
CTP. All participants complete a comprehensive clinical assessment upon admission to the CTP program and most of these assessments are repeated at six month intervals throughout a person’s hospitalization in order to monitor treatment response and inform future treatment planning decisions. Currently these assessments primarily include measures of neurocognitive and social cognitive functioning. Clinical psychology graduate students or trained clinical assistants administer and score all measures according to standardized instructions. Scoring is assisted by several computerized scoring programs. One significant change occurred in the routine assessment battery during the nine year period from which the archival data was extracted. The RBANS and social cognitive measures were added to the assessment battery between 2000 and 2001 when new admissions arrived during that time. The RAVLT was phased out at that time in order to maintain that the assessment battery could be completed in a manageable amount of time. Therefore, individuals discharged before 2000 do not have RBANS and social cognitive data. Likewise, people who entered the program at the time of the change do not have RAVLT data. Nevertheless, assessments continued to be collected at six-month intervals and therefore biannual neurocognitive and/or social cognitive data is available for most participants.
In addition, assessments of behavioral, social, and overall functioning are regularly completed in the context of the general milieu. For example, psychiatric technicians complete NOSIE assessments on a weekly basis. The monthly average of the weekly ratings were included in the present database, making monthly NOSIE data pertaining to each person’s ward functioning available. Most participants also have behavioral management programs in place which monitor specific behaviors identified on individualized treatment plans such as rehabilitation noncompliance (RNC), as described above. These behaviors are monitored and tracked continuously in individualized tracking forms. The RNC data and other behavioral management data are entered into Excel spreadsheets as the number of instances of the target behavior each week. Data entry and management is completed by a trained clinical psychology graduate student on a monthly basis to be used in treatment progress meetings. At the time of extraction, the data is subjected to fidelity checks to monitor if the behaviors are being correctly recorded and contingencies implemented as intended.
For the present study, the first 18 months of rehabilitation were followed resulting in the maximum number of possible data points for each person as four neurocognitive and social cognitive measure time points, 18 monthly NOSIE total assets scores, and 72 weeks of RNC data. The 18 month time period was selected because the average length of stay at CTP is around 18 months which maximizes the amount of data available at any given time point. A majority of treatment in terms of psychoeducational skills training is delivered within this time frame, as well. A qualitative analysis completed as part of program evaluation of the CTP program in February 2005 indicated that people with the most protracted lengths of stay were associated with legal status (i.e., court approval and availability of appropriate community placement for NRRI participants), severe treatment-resistant aggression, and long waiting lists for specific community placements (e.g., residential rehabilitation in Omaha, developmental disability services).
The current format of the archival database does not lend itself to pre-post analyses. Future studies may consider reformatting the archival database so as to have an admission data point and a discharge data point (or the assessment closest to discharge) as an approximation of pre- and post- rehabilitation functioning.
LCCMHC. Extensive chart review and interviews with LCCMHC staff were completed by a dedicated clinical psychology graduate student involved in program evaluation activity as part of an assistantship at LCCMHC. In addition, after collecting hospitalization data, the data was cross-checked with LRC records through chart review to ensure its accuracy. Not all CTP participants were served by LCCMHC upon discharge and hospitalization data for those individuals was obtained through LRC records whenever possible.
After the archival database was completed with data from both settings, two graduate students completed additional quality assurance checks. Data was subjected to cross-checking with original and computerized archival data to ensure its reliability and accuracy.
Data Cleaning.
Before analyses, data were examined for skewness and potential outliers. It was necessary to ensure normal distribution of the dependent variables because most of the analyses used in this study assume normality. Distributional skewing and asymmetrical outliers can both produce skew and therefore transformation and/or outlier windsorizing were applied only after examining the nature of the skew. Any dependent variables that demonstrated a skewed distribution (skewness > +/- 1.00) without outliers were normalized using conservative transformation procedures. All variables were able to be brought within acceptable skewness range ( 0.14, (See Table 10). The one exception was on COGLAB Asarnow false alarms. Those with guardians have a significantly higher number of false alarms than those without guardians, F (1, 92) = 3.82, p = 0.05, η2 = .040. However, this isolated finding does little to support the hypothesis, especially considering the number of analyses conducted.
Table 9
Guardianship Status by Neurocognitive Variables (Group 2) at Admission Between Group Multivariate and Univariate Statistics
|Multivariate F (6, 12) = .403, p=.863 |Univariate |
|Neurocognitive Variables |M (SD) |F |df |p |
|RBANS Total | |.16 |1, 9 |.70 |
|Guardian |78.00 (23.07) | | | |
|No Guardian |73.25 (15.59) | | | |
|COGLAB Asarnow Hits | |.01 |1, 9 |.94 |
|Guardian |26.00 (2.65) | | | |
|No Guardian |25.88 (2.47) | | | |
|COGLAB Asarnow False Alarms | |2.12 |1, 9 |.18 |
|Guardian |4.49 (3.36) | | | |
|No Guardian |2.04 (.67) | | | |
|COGLAB Card Sort Random Errors | |.01 |1, 9 |.92 |
|Guardian |18.00 (14.42) | | | |
|No Guardian |17.00 (13.56) | | | |
|COGLAB Card Sort Perseverative Errors |14.33 (9.50) |.23 |1, 9 |.65 |
|Guardian |18.75 (14.69) | | | |
|No Guardian | | | | |
|RCFT Copy |29.33 (6.43) |1.13 |1, 9 |.32 |
|Guardian |32.75 (4.13) | | | |
|No Guardian | | | | |
|RCFT Immediate Memory |11.55 (1.41) |.22 |1, 9 |.65 |
|Guardian |14.12 (1.21) | | | |
|No Guardian | | | | |
|RCFT Delayed Memory |9.56 (1.41) |.60 |1, 9 |.46 |
|Guardian |13.71 (1.34) | | | |
|No Guardian | | | | |
|RCFT Recognition |17.33 (4.16) |2.28 |1, 9 |.17 |
|Guardian |19.88 (1.73) | | | |
|No Guardian | | | | |
|Trails A |28.58 (16.17) |.89 |1, 9 |.37 |
|Guardian |40.76 (19.83) | | | |
|No Guardian | | | | |
|Trails B |80.58 (54.96) |.03 |1, 9 |.88 |
|Guardian |84.50 (29.65) | | | |
|No Guardian | | | | |
|WRAT |67.67 (32.65) |.219 |1, 9 |.65 |
|Guardian |61.25 (14.92) | | | |
|No Guardian | | | | |
Table 10
Guardianship Status by Neurocognitive Variables at Admission One-Way Analyses of Variance (ANOVAs)
|Neurocognitive Measures |M(SD) |F |df |p |
|RBANS Total | |.41 |1, 56 |.53 |
|Guardian |72.81 (16.95) | | | |
|No Guardian |75.38 (13.66) | | | |
|COGLAB Asarnow Hits |24.39 (3.85) |2.19 |1, 92 |.14 |
|Guardian |25.50 (3.46) | | | |
|No Guardian | | | | |
|COGLAB Asarnow False Alarms |4.26 (1.52) |3.82 |1, 92 |.05 |
|Guardian |2.63 (.93) | | | |
|No Guardian | | | | |
|COGLAB Card Sort Random Errors |21.85 (14.77) |.71 |1, 96 |.40 |
|Guardian |24.69 (18.21) | | | |
|No Guardian | | | | |
|COGLAB Card Sort Perseverative Errors |21.61 (13.30) |.31 |1, 96 |.58 |
|Guardian |23.12 (14.32) | | | |
|No Guardian | | | | |
|RCFT Copy |28.68 (5.58) |1.71 |1, 33 |.20 |
|Guardian |31.26 (5.81) | | | |
|No Guardian | | | | |
|RCFT Immediate Memory | |.69 |1, 33 |.41 |
|Guardian |15.23 (1.79) | | | |
|No Guardian |12.49 (1.56) | | | |
|RCFT Delayed Memory |14.02 (1.82) |.05 |1, 33 |.83 |
|Guardian |13.29 (1.87) | | | |
|No Guardian | | | | |
|RCFT Recognition |21.07 (6.13) |1.76 |1, 33 |.19 |
|Guardian |18.50 (5.25) | | | |
|No Guardian | | | | |
|Trails A |43.38 (16.50) |1.75 |1, 63 |.19 |
|Guardian |38.05 (15.86) | | | |
|No Guardian | | | | |
|Trails B |98.18 (28.45) |1.84 |1, 60 |.18 |
|Guardian |87.45 (33.20) | | | |
|No Guardian | | | | |
|WRAT |61.21 (14.10) |2.23 |1, 83 |.14 |
|Guardian |65.38 (11.52) | | | |
|No Guardian | | | | |
|RAVLT |8.80 (3.52) |.81 |1,74 |.37 |
|Guardian |9.53 (3.52) | | | |
|No Guardian | | | | |
In addition, examination of the means between these two groups reveals that some differences on neurocognitive variables, though not statistically significant, are counterintuitive. That is, contrary to hypothesis, on certain measures those without guardians demonstrated poorer neurocognitive functioning than those with guardians. For example, as expected, people with guardians have fewer hits and more false alarms on the COGLAB Asarnow task, take longer on Trails A and B, and have lower RBANS total, RCFT copy, RAVLT, and WRAT scores than those without guardians. On the other hand, contrary to hypothesis, those with guardians have fewer perseverative and random errors on the COGLAB and higher scores on the RCFT immediate memory, delayed memory, and recognition tests. Therefore, there is no clear pattern indicating lower neurocognitive functioning for those with guardians.
Exploratory analyses within the guardianship group based on whether the guardian was acquired before or after admission, whether the guardian was a family member, and whether the person with a guardian was admitted CC or VpG using one-way ANOVAs revealed only two marginally significant differences. On the WRAT, those who acquired guardians (n = 9) had a lower score than those who had guardians at admission (n = 34), Ms = 53.44, 63.27, respectively, F (1, 41) = 3.67, p = .06, η2 = .082. Also on the WRAT, those whose guardians were family members (n = 29) had lower score than those whose guardians were not family members (n = 14), Ms = 58.72, 66.36, respectively, F (1, 41) = 2.89, p = .097 η2 = .066. There were no differences on any of the other neurocognitive variables on any of the three within-group dimensions, all Fs < 2.38, all ps >.14.
Overall, contrary to hypothesis, these results suggest that participants with guardians do not significantly differ from those without guardians at the time of admission, with respect to neurocognitive functioning. However, given the one significant finding supporting the hypothesis, differences in neurocognitive functioning will be assessed to determine if the difference on COGLAB false alarms remains stable over time. In addition, there do not appear to be substantial within-group differences in neurocognitive functioning for the Guardian group.
Social cognitive functioning.
The I-SEE, a measure of attributional style, and the Insight Scale, a measure of insight into functional impairment and need for treatment, were used to examine social cognitive differences between the two groups. It was hypothesized that people with guardians would demonstrate lower insight than those without guardians. In addition, a lower assessment of self-efficacy and greater external locus of control as measured by the attributional measure was expected for those with guardians than those without guardians. The bivariate correlation matrix for the social cognitive measures and their subscales can be found in the Appendix, see Table A3. One relationship to note was a significant negative correlation between total insight and self-efficacy. This is somewhat counterintuitive in that the relationship suggests that participants with lower insight have a greater assessment of their self-efficacy. Because this study predicted higher levels of insight and higher self-efficacy for those without guardians, the negative correlation between these two scales makes it difficult to support that hypothesis.
To assess the relationship between social cognitive functioning and guardianship status, the composite scales of the I-SEE, self-efficacy and externality, and the total insight score were entered as within subjects factors into a MANOVA with guardianship status as the between groups factor. The 2 (guardianship status) x 3 (measure) MANOVA revealed no significant differences for any of the social cognitive variables between guardianship groups (See Table 11). This suggests that, contrary to hypothesis, participants in this sample with guardians do not differ with regard to social cognitive functioning than those without guardians.
Table 11
Guardianship Status by Social Cognitive Variables at Admission Between Group Multivariate and Univariate Statistics
|Multivariate F (3,42) = .145, p=.93 |Univariate |
|Social Cognitive Measures |M(SD) |F |df |p |
|I-SEE Self-Efficacy | |.23 |1, 43 |.75 |
|Guardian |69.06 (8.52) | | | |
|No Guardian |67.88 (8.03) | | | |
|I-SEE Externality | |.13 |1, 43 |.21 |
|Guardian |48.00 (10.66) | | | |
|No Guardian |49.32 (13.69) | | | |
|Insight Total Score | |.06 |1, 43 |.59 |
|Guardian |7.93 (3.95) | | | |
|No Guardian |7.64 (3.89) | | | |
Exploratory analyses using one-way ANOVAS revealed no significant differences within the guardianship group based on whether the guardian was acquired before or after admission or whether the guardian was a family member, all Fs < 3.17, all ps > .09. On the I-SEE externality composite scale, those admitted VpG (n = 13) had a higher score (M = 56.46) than those admitted via CC (n = 9, M = 44.44), F (1, 20) = 5.14, p = .04, η2 = .204. There were no other significant differences on social cognitive variables for those admitted VpG versus those admitted via CC, all Fs < 1.80, all ps > .20.
Behavioral functioning.
The NOSIE total assets score was used as a measure of general behavioral functioning. It was expected that those in the Guardian group would demonstrate poorer behavioral functioning than those in the No Guardian group. As hypothesized, a one-way ANOVA revealed a significant difference, F (1, 148) = 7.92, p = .006, η2 = .051, on total assets between those with guardians (M = 149.94) and those without guardians (M = 160.18) indicating lower behavioral functioning at admission for those participants with guardians.
Because the NOSIE is comprised of three adaptive functioning scales and three maladaptive functioning scales, further exploratory analyses were conducted to determine if differences were apparent across all areas of functioning assessed by the NOSIE or whether the differences were specific to particular subscales. Pearson product moment correlations revealed significant intercorrelations among all subscales and the total assets score, except for the relationship between the irritability and social interest subscales. The bivariate correlation matrix can be found in the Appendix, see Table A4.
Follow-up analyses revealed significant differences between those with guardians and those without on the social competence, neatness, and motor retardation subscales (See Table 12). The difference between the two groups on psychoticism approached significance. All differences were in the expected direction with lower levels of social competence and neatness and higher levels of psychoticism and motor retardation for those with guardians than without guardians.
Table 12
Guardianship Status by NOSIE Subscales at Admission One-Way Analyses of Variance (ANOVAs)
|NOSIE Subscales |M |F |Df |p |η2 |
|Social Competence | | | | | |
| Guardian |31.71 |14.20 |1, 148 |.0002 |.088 |
| No Guardian |34.88 | | | | |
|Social Interest | | | | | |
| Guardian |15.21 |.46 |1, 148 |.50 |.003 |
| No Guardian |15.92 | | | | |
|Neatness | | | | | |
| Guardian |22.09 |10.06 |1, 148 |.002 |.064 |
| No Guardian |24.75 | | | | |
|Irritability | | | | | |
| Guardian |5.19 |1.68 |1, 148 |.20 |.011 |
| No Guardian |4.73 | | | | |
|Psychoticism | | | | | |
| Guardian |1.51 |3.46 |1, 148 |.07 |.023 |
| No Guardian |1.05 | | | | |
|Motor Retardation | | | | | |
| Guardian |7.25 |14.02 |1, 148 |.0003 |.087 |
| No Guardian |5.11 | | | | |
Discriminant analyses were used to determine if guardianship status (Guardian v. No Guardian) could be predicted with the NOSIE social competence, neatness, motor retardation, and psychoticism subscales. The discriminant function is significant with a Wilk’s lambda of 0.90 (p = .004), an R2-canonical of 0.099, and 61.3% correct re-classification (92 out of 150). The model correctly classifies 78.8% of those without guardians (63 out of 80), but less that half, or 41.4%, of those with guardians (29 out of 70). Table 13 shows the standardized canonical coefficients and the structure weights. Although the structure matrix identifies all four variables as contributing to the multivariate effect, the standardized canonical coefficients reveal that neatness and psychoticism have little unique contribution to the model. This indicates that the most efficient model for discriminating between the Guardian and No Guardian groups include unique contributions from the NOSIE subscales of social competence and motor retardation. These results are consistent with the ANOVA analyses described above.
Table 13
Standardized Canonical Coefficients and Structure Weights from Discriminant Analyses of NOSIE Subscales to Predict Guardianship Status
|NOSIE Subscales |Standardized Coefficients |Structure Weights |
|Social Competence |-.353 |.926 |
|Neatness |-.138 |-.896 |
|Psychoticism |.090 |-.785 |
|Motor Retardation |.578 |.440 |
Results of NOSIE analyses support the hypothesis that those with guardians demonstrate poorer behavioral functioning upon admission than those without guardians. Specifically, those without guardians were better at adhering to a daily schedule (the social competence subscale) and keeping themselves and their living space neat (the neatness subscale) than those without guardians. In addition, people with guardians showed higher levels of lethargy and amotivational syndrome (the motor retardation subscale) than those without guardians.
When exploratory analyses were repeated to identify differences on the NOSIE total assets within the Guardian group, no significant differences emerged between the Guardian Admission and Guardian Acquired groups, F (1, 68) = 0.84, p = .36. There is also no difference on the NOSIE total assets score between people whose guardians are family members and those who are not, F (1, 67) = 0.67, p = 0.42. Likewise, comparing people with guardians whose legal status was CC to those with a VpG status, there is no significant difference, F (1, 67) = 0.03, p = .85.
Treatment compliance.
Only participants for whom compliance is thought to be problematic by the treatment team have a RNC behavior management program as a component of his or her treatment plan. Therefore, not all participants have RNC data available. Simply having RNC data may be considered an indicator of compliance in the CTP program. For this analysis, there were six people with missing data for RNC. In addition, there were 66 people discharged from CTP before computerized data management of RNC data began. Of the remaining 90 people in the sample, only four people did not have a RNC behavior management program at any time during rehabilitation. One of these four people was a Voluntary admission and did not have a guardian. The other three were all admitted via CC and did not have guardians. Of the 86 people who had RNC data available at some point during treatment, 46 had guardians and 40 did not, which indicates that compliance is identified as a problem by CTP treatment teams for people with and without guardians.
People with guardians were expected to demonstrate lower levels of compliance with treatment than those without guardians. The first time point of data available for this analysis was the average number of weekly instances of noncompliance during the second three months of the first six months in treatment. Contrary to hypothesis, there are no significant differences between the Guardian (M = 2.45) and No Guardian (M = 2.07) groups, F (1, 74) = 0.41, p = .53. This suggests that guardianship status is not related to treatment compliance.
Within-group differences in the Guardian group were explored using one-way ANOVAS. There is a nonsigificant trend between the Guardian Admission and Guardian Acquired groups, F (1, 39) = 3.17, p = 0.08, with those who acquired guardians having a higher number of instances of noncompliance (M = 4.36) than those who had guardians at admission (M = 2.12). However, for this analysis, there were only seven people in the Guardian Acquired group compared to 34 people in the Guardian Admission group. There is no significant mean difference on instances of weekly noncompliance between people whose guardians are family members and those who are not, F (1, 38) = 0.01, p = 0.94. Likewise, comparing people with guardians whose legal status was CC to those with a VpG status, there is no significant difference, F (1, 38) = 0.08, p =.78.
Symptomatology.
It was anticipated that people with guardians would have a higher level of symptomatology than those without guardians. Contrary to hypothesis, a one-way ANOVA revealed no significant mean difference, F (1, 27) = 0.94, p = .34, in overall symptomatology as measured by the BPRS total score between those with guardians (M = 45.58) and those without guardians (M = 42.71).
Six factor scores for the BPRS were computed in order to evaluate symptoms groupings as opposed to the gross overall measure of symptomatology provided by the total score. The six factors used were Psychotic Disorganization, Hallucinations/Delusions, Paranoia, Emotional Blunting, Agitation/Elation and Anxiety/Depression. The bivariate correlation matrix of the six factors and the total BPRS score can be found in the Appendix, see Table A5. A 2 (guardianship status) X 6 (BPRS factor scores) MANOVA was used to assess whether differences existed between groups in symptom areas. There are no significant differences between groups for any of the symptom factor scores, all Fs < 2.41, all ps >.13, see Table 14.
Table 14
Guardianship Status by BPRS Factor Scores at Admission Between Group
MANOVA
|Multivariate F (6, 22) = 1.23, p=.33 |Univariate |
|BPRS Factor Scores |M (SD) |F |df |p |
|Psychotic Disorganization | |.09 |1, 27 |.77 |
|Guardian |6.42 (1.30) | | | |
|No Guardian |6.64 (2.35) | | | |
|Emotional Blunting | |2.41 |1, 27 |.13 |
|Guardian |5.37 (2.20) | | | |
|No Guardian |4.24 (1.72) | | | |
|Paranoia | |.78 |1, 27 |.39 |
|Guardian |7.77 (1.56) | | | |
|No Guardian |8.50 (2.56) | | | |
|Anxiety/Depression | |1.70 |1, 27 |.20 |
|Guardian |9.29 (2.19) | | | |
|No Guardian |8.03 (2.80) | | | |
|Hallucinations/Delusions |6.00 (3.16) |.57 |1, 27 |.46 |
|Guardian |5.26 (2.11) | | | |
|No Guardian | | | | |
|Agitation/Elation |3.46 (1.42) |.04 |1, 27 |.84 |
|Guardian |3.35 (1.36) | | | |
|No Guardian | | | | |
Finally, one-way ANOVAs for each of the 24 BPRS items were conducted to determine if there were differences in any specific symptoms between the Guardian and No Guardian groups. Only one significant difference emerged. People with guardians were rated with a higher level of Motor Retardation, F (1,27) = 4.29, p = .05, η2 = .137, than those without guardians. A marginally significant difference was found between the groups on Uncooperativeness, F (1,27) = 3.10, p = .103, with those in the Guardian group being rated at a higher level of Uncooperativeness than those in the No Guardian group. Results of all the one-way ANOVAs can be found in Table 15.
Table 15
Guardianship Status by BPRS Items One-Way Analyses of Variance (ANOVAs)
|BPRS Items |M (SD) |F |df |p |
|Somatic Concern |Guardian |2.42 (1.83) |1.13 |1, 28 |.30 |
| |No Guardian |1.81 (1.32) | | | |
|Anxiety |Guardian |2.74 (1.41) |1.70 |1, 28 |.20 |
| |No Guardian |2.17 (1.04) | | | |
|Depression |Guardian |2.63 (1.15) |.79 |1, 28 |.38 |
| |No Guardian |1.22 (1.26) | | | |
|Suicidality |Guardian |1.33 (0.65) |.16 |1, 28 |.69 |
| |No Guardian |1.47 (1.06) | | | |
|Guilt |Guardian |1.67 (0.89) |1.24 |1, 28 |.28 |
| |No Guardian |2.08 (1.07) | | | |
|Hostility |Guardian |1.92 (1.17) |.83 |1, 28 |.37 |
| |No Guardian |2.33 (1.27) | | | |
|Elevated Mood |Guardian |1.50(1.24) |.52 |1, 28 |.48 |
| |No Guardian |1.25 (0.65) | | | |
|Grandiosity |Guardian |1.67 (1.37) |.09 |1, 28 |.77 |
| |No Guardian |1.53 (1.17) | | | |
|Suspiciousness |Guardian |2.83 (1.47) |.37 |1, 28 |.55 |
| |No Guardian |2.53 (1.28) | | | |
|Hallucinations |Guardian |1.58 (1.24) |1.78 |1, 28 |.20 |
| |No Guardian |1.17 (0.42) | | | |
|Unusual Thought Content |Guardian |2.25 (1.80) |.001 |1,27 |.98 |
| |No Guardian |2.26 (1.50) | | | |
|Bizarre Behavior |Guardian |1.42 (0.79) |.51 |1, 28 |.48 |
| |No Guardian |1.67 (1.03) | | | |
|Self-Neglect |Guardian |2.17 (0.91) |.27 |1, 28 |.61 |
| |No Guardian |2.00 (0.82) | | | |
|Disorientation |Guardian |1.58 (1.16) |.05 |1, 27 |.83 |
| |No Guardian |1.50 (0.85) | | | |
|Conceptual Disorganization |Guardian |1.83 (1.12) |1.05 |1, 27 |.32 |
| |No Guardian |1.47 (0.80) | | | |
|Blunted Affect |Guardian |2.96 (1.48) |.13 |1, 27 |.72 |
| |No Guardian |2.76 (1.35) | | | |
|Emotional Withdrawal |Guardian |2.38 (1.67) |.03 |1, 27 |.88 |
| |No Guardian |2.47 (1.55) | | | |
|Motor Retardation |Guardian |2.32 (1.20) |4.29 |1, 27 |.05* |
| |No Guardian |1.59 (0.71) | | | |
|Tension |Guardian |1.50 (0.91) |.55 |1, 27 |.46 |
| |No Guardian |1.29 (0.59) | | | |
|Uncooperativeness |Guardian |1.33 (0.49) |3.10 |1, 27 |.09 |
| |No Guardian |1.97 (1.18) | | | |
|Excitement |Guardian |1.58 (1.17) |.02 |1, 27 |.90 |
| |No Guardian |1.53 (1.13) | | | |
|Distractibility |Guardian |1.46 (0.89) |.02 |1, 27 |.89 |
| |No Guardian |1.41 (0.87) | | | |
|Motor hyperactivity |Guardian |1.25 (0.87) |.03 |1, 27 |.87 |
| |No Guardian |1.29 (0.59) | | | |
|Mannerisms and Posturing |Guardian |1.25 (0.62) |.003 |1, 27 |.96 |
| |No Guardian |1.24 (0.75 | | | |
*p .11. Recall that there was a significant difference noted between groups at admission on COGLAB false alarms; this isolated significant finding did not remain stable over time.
Table 16
Guardianship Status by COGLAB Repeated Measures ANOVAs
| |Measures |
| |COGLAB Asarnow Hits|COGLAB |WCST Perseverative |WCST Random Errors |
| | |Asarnow |Errors | |
| | |False Alarms | | |
|Variable |F (3,31) |F (3,31) |F (3.32) |F (3, 32) |
|Main Effect |376.60*** |4.27** |53.29*** |3.17* |
|Time | | | | |
|Main Effect |1.60 |.311 |.248 |.135 |
|Guardianship Status | | | | |
|Interaction |.74 |2.42 |1.19 |2.44 |
|Time * Guardianship Status | | | | |
*p .31.
In summary, results of analyses on neurocognitive functioning at admission and over the course of treatment suggest that improvements in neurocognitive functioning are evident over the course of treatment for the entire sample, but that, contrary to hypotheses, those with guardians do not significantly differ from those without guardians with respect to neurocognitive functioning.
Social cognitive functioning.
It was anticipated that those with guardians would demonstrate lower insight, lower assessment of self-efficacy, and higher external locus of control than those without guardians throughout the course of treatment. MANOVA could not be used to evaluate change over time in the social cognitive variables as they relate to guardianship status because fewer data points were available since social cognitive measures were not added to the routine assessment battery until 2000-01. Using paired sample t-tests with the overall combined means of the two guardianship groups, change in social cognition with treatment was tested. From admission to 6 months, there were no significant changes in insight or attributional style, all ts < 1.15, all ps >.26. However, a significant increase in the total insight score is evident from 6 months to 12, increasing from an average score of 7.07 to 8.43, t (35) = -2.81, p = .01. From admission to 12 months of treatment, the increase in insight is marginally significant, t (27) = -1.80, p = .08. There were no significant changes in attributional style as measured by the composite scales of externality and self-efficacy on the I-SEE from 6 months to 12 months. However, as expected, a significant decrease in externality from admit to 12 months was found, t (31) = 2.27, p = .03. There were no significant changes in self-efficacy as measured by the I-SEE found to occur with treatment.
One-way ANOVAs were used to determine if there are group differences in social cognition based on guardianship status at 6 months and 12 months. There are no significant differences for any of the social cognitive variables between guardianship groups, all Fs .30, congruent with results of analyses with social cognitive variables at the time of admission.
This suggests that, contrary to hypothesis, participants with guardians do not differ from those with guardians in social cognitive functioning. Specifically, there is no difference in attributional style or in the overall level of insight between the two groups.
Behavioral functioning.
Poorer behavioral functioning over the course of treatment was hypothesized for those with guardians as compared to those without guardians. A mixed model MANOVA was used to assess whether differences noted between groups on the NOSIE at admission persisted throughout treatment. The average monthly total assets scores during the first, sixth, and twelfth months of treatment were included in analyses. As hypothesized, the 2 (guardianship status) x 3 (time) repeated measures ANOVA revealed a significant main effect for time on the NOSIE, F (2, 117) = 20.38, p < .001, η2 =.258, indicating that, overall, improvements in NOSIE occur with treatment. As hypothesized, the between-subjects main effect for guardianship status is also significant, F (1, 118) = 4.11, p = .05, η2 =.034, indicating that differences in behavioral functioning found at admission as measured by the NOSIE between those with guardians and those without persist over the course of treatment. The interaction of NOSIE over time with guardianship status was not significant, F (2, 117) = .23, p < .78. The means and standard deviations for these analyses are found in Table 17.
Table 17
Mean Scores and Standard Deviations for NOSIE Total Assets at Admission, 6 Months, and 12 Months as a Function of Guardianship Status
| |NOSIE Total Assets Scores |
| |At Admission |Six Months |Twelve Months |
|Guardianship Status |M (SD) |M (SD) |M (SD) |
| Guardian |149.94 (23.16) |155.65 (23.51) |159.95 (20.21) |
| No Guardian |160.18 (21.39) |165.40 (23.55) |169.91 (18.45) |
Follow up analyses using one-way ANOVAs revealed that differences between the two groups were significant at six, F (1, 145) = 6.29, p = .01, η2 =.042, and 12 months, F (1, 123) = 8.18, p = .005, η2 =.062 and that they remained in the expected direction with lower NOSIE total assets scores for those with guardians. It should be noted that the 18th month time point was not included in the mixed model MANOVA above because a lower N at that time point decreases the power available for the overall analyses. In a follow-up one-way ANOVA, differences between the two groups were found to remain significant at 18 months, F (1, 84) = 5.60, p = .02, η2 =.062, in the expected direction, with those in the Guardian group having lower NOSIE total assets scores (M = 159.18) than those in the No Guardian group (M = 169.65). The NOSIE total assets scores from admission to 18 months as a function of guardianship status are displayed in Figure 1.
Figure 1
NOSIE Total Assets Scores Over Time as a Function of Guardianship Group
[pic]
Additional follow-up analyses using the NOSIE total assets score at the time of admission as a covariate revealed that the difference between the Guardian and No Guardian group at six months was not significant when controlling for functioning at admission, F (2,142) = 0.62, p = .43, and only marginally significant at 12 months, F (2,123) = 3.55, p = .06, η2 =.029.
Because of the pattern of results evident at admission on the six NOSIE subscales, follow-up analyses were conducted to determine if the pattern of results remained the same over the course of treatment. A 2 (guardianship status) x 2 (time: 6 months and 12 months) repeated measures ANOVA was completed for each of the 6 subscales. As can be seen in Table 18, there were significant main effects for time on the social competence (η2 =.218), social interest (η2 =.304), neatness (η2 =.200), psychoticism (η2 =.036), and motor retardation (η2 =.131), indicating significant change in all these areas over time. Improvements were seen in social competence, social interest, and neatness while a decrease in motor retardation is evident. The significant change over time on the psychoticism subscale represents an increase in psychoticism as measured by the NOSIE. There was not a significant main effect of time for irritability, indicating minimal change over the course of treatment in irritability. Means and standard deviations between groups can be seen in Table 19. Mirroring the pattern found at admission, significant main effects of guardianship status are present for social competence (η2 =.069), neatness (η2 =.076), and motor retardation (η2 =.063). Differences are in the same direction as evident at admission, with those with guardians having lower levels of social competence and neatness and higher levels of motor retardation than those without guardians. There are no significant interaction effects for any of the analyses.
Table 18
Guardianship Status by NOSIE Subscales Repeated Measures ANOVAs
| |NOSIE Subscales |
| |Social Competence |Social Interest |
|NOSIE Subscales |M |SD |M |SD |
|Social Competence |31.03 |6.98 |34.67 |5.35 |
|Guardian |34.17 |5.47 |36.79 |4.53 |
|No Guardian | | | | |
|Social Interest |15.27 |6.08 |19.29 |5.91 |
|Guardian |15.17 |6.13 |19.83 |6.38 |
|No Guardian | | | | |
|Neatness |22.03 |5.74 |23.97 |4.70 |
|Guardian |24.31 |4.62 |26.70 |4.32 |
|No Guardian | | | | |
|Irritabilitya |5.63 |6.14 |5.60 |5.13 |
|Guardian |5.13 |6.06 |4.36 |5.09 |
|No Guardian | | | | |
|Psychoticismb |1.98 |3.08 |2.73 |3.73 |
|Guardian |1.51 |2.65 |1.84 |2.59 |
|No Guardian | | | | |
|Motor Retardation |7.65 |4.52 |5.58 |3.97 |
|Guardian |5.51 |3.50 |4.34 |3.34 |
|No Guardian | | | | |
a, b Note: Standard deviations for these variables are large and results should be interpreted cautiously.
These results suggest that there are significant differences in behavioral functioning between those with guardians and those without guardians. These results support the hypothesis that those with guardians have poorer behavioral functioning as measured by the NOSIE than those without guardians. In addition, it appears that the differences in specific areas of functioning remain stable over time, even over the course of rehabilitation.
Finally, while exploratory analyses on the within-group characteristics were only conducted with the admission data for other domains of functioning, analyses were repeated for the NOSIE total assets score over time since consistent differences in the overall Guardian versus No Guardian analyses were found. Analyses indicated that the pattern of results described above do not change with respect to within-group characteristics of guardianship. When one-way ANOVAs were repeated to examine differences in NOSIE total assets at six months, twelve months, and eighteen months of treatment, no significant differences were found between the Guardian Acquired and Guardian Admission groups, all Fs < 0.84, all ps > .48. In addition, there are no significant differences between people whose guardians are family members and those who are not on NOSIE total assets, all Fs < 0.34, all ps > .57. Likewise, comparing people with guardians whose legal status was CC to those with a VpG status, no significant differences were found, all Fs < 1.96, all ps > .17.
Treatment compliance.
As described above, an average number of weekly instances of noncompliance was computed for the second three months of each six month time periods of treatment (i.e. months 4, 5, & 6; months 10, 11, & 12, and months 16, 17, & 18). It was predicted that those with guardians would demonstrate less compliance with treatment than those without guardians. Consistent with analyses at admission, a one-way ANOVA revealed no significant differences on RNC between those with guardians and those without during the second six months of treatment, F (1, 66) = 0.24, p = .63, or the third six months of treatment, F (1, 42) = 1.58, p = .22.
As expected, there was a significant decrease in weekly instances of noncompliance for all participants from the first six months (M = 2.55) to the second six months (M = 2.04) of treatment, t (61) = 2.15, p = 0.04, indicating an increase in compliance with the demands of the rehabilitation program. The difference from twelve months (M = 2.26) to eighteen months (M=1.87) is not significant, t (43) = 1.19, p = .24.
Follow-up exploratory analyses were conducted to determine if the pattern of change (i.e., slope) in weekly instances of noncompliance differs between the two groups. The change in weekly number of instances of noncompliance was calculated by subtracting the average from the first six month period from the average from the third six month period. The slope for both groups was negative indicating a decrease in the number of instances of noncompliance, but there was not a significant difference between the Guardian (M = -.17) and No Guardian (M = -.30) groups based on the slope, F (1, 38) = 0.42, p = .52.
These results suggest there is not a relationship between guardianship status and treatment compliance at admission or over the course of treatment.
Symptomatology.
There were fewer people who had repeated assessments of the BPRS because the BPRS was added to the routine assessment battery later in this archival database cohort. Therefore, low N prevented use of repeated measures ANOVA to assess the change in BPRS over time by guardianship status group. Instead, to maximize the number of valid cases available for analyses, one-way ANOVAs were completed on assessments administered at 6 months, 12 months, and 18 months of treatment between guardianship groups on the BPRS Total Score. Higher symptomatology was predicted in the Guardian group than the No Guardian group. At 6 months, there is no significant mean difference, F (1, 47) = 0.003, p = .95, between those with guardians (M = 44.34) and those without guardians (M = 44.57). Likewise, at 12 months, there is no significant difference, F (1, 37) = 0.26, p = .61, between the groups (M = 41.95, 43.63, respectively). Therefore, the lack of a statistically significant difference in symptomatology between groups found at admission persists at 6 months and again at 12 months.
At 18 months, however, there is a significant difference, in the expected direction, between groups, F (1, 25) = 7.41, p = .01, η2 =.229, with higher BPRS scores for those with guardians (M = 47.47) than whose without guardians (M = 37.67). At 18 months, there are only nine people in the No Guardian group as opposed to 18 in the Guardian group. The low N for this analysis calls its results into question. In addition, a protracted length of stay for those with guardians was identified in earlier analyses. Because of the identified relationship between length of stay and guardianship status, an ANCOVA analysis with length of stay as the covariate was done to determine whether differences between the two groups on BPRS total score are still evident when controlling for length of stay. With this analysis, the significant difference remains between the groups, F (2, 24) = 4.64, p = .04, η2 =.162.
To further evaluate symptomatology, the six BPRS factor scores were analyzed for group differences. A 2 (guardianship status) X 6 (BPRS factor scores) MANOVA for each time point, (i.e., 6, 12, and 18 months) was done to assess whether differences existed between groups in symptom groups. There are no between group effects at any of the three time points on any of the factor scores, all Fs < 1.67, all ps > .19.
Because isolated differences between groups on individual BPRS items were found at admission, those analyses were repeated here to determine whether there is a consistent pattern over time on any particular BPRS items. At six months, there are no significant differences on any items, all Fs < 3.75, all ps > .06. The two items which approached significance are Hostility, F (1, 50) = 3.75, p = .06, η2 =.070, and Suspiciousness, F (1,50) = 3.09, p = .09, η2 =.058, but both are in the unexpected direction with those without guardians having higher symptom ratings than those with guardians. At 12 months, only one significant difference emerged on all items. On Suspiciousness, people without guardians again have a higher rating on suspiciousness than those without guardians, F (1, 38) = 3.95, p = 0.05, η2 =.094. The difference between the groups on Disorientation at 12 months approaches significance, F (1, 37) = 3.24, p = .08, η2 =.080. This difference is in the expected direction with symptom ratings on disorientation being higher for those with guardians than those without guardians. No other differences were noted at 12 months, all Fs < 1.44, all ps > .24. Finally, at 18 months, significant differences between the two groups were found on the following BPRS items: Anxiety, F (1, 27) = 4.67, p = .04, η2 =.148, Elevated Mood, F (1, 27) = 4.28, p = .05, η2 =.137, Bizarre Behavior, F (1, 27) = 4.47, p = .04, η2 =.142, Conceptual Disorganization, F (1, 27) = 4.27, p = .05, η2 =.136, and Tension, F (1, 27) = 4.47, p = 0.04, η2 =.142. All of these differences were in the expected direction with those with guardians having higher symptom ratings than those without guardians. However, there are only 9 people in the No Guardian and 20 people in the Guardian group at the 18 months time point. Overall, analyses of the BPRS items indicate that there is no consistent pattern over the course of treatment regarding symptomatology between groups.
In summary, despite the difference noted at Time 4 on the BPRS total score and the various differences noted on individual BPRS items, results of analyses of the relationship between symptomatology as measured by the BPRS and guardianship status suggest that there is considerable variability in symptomatology and that this is not likely attributable to group membership.
Hypothesis 2
Outcome.
A pattern of differential outcome was hypothesized for the Guardian and No Guardian groups. Specifically, it was predicted that those with guardians would be discharged to more restrictive settings than those without guardians. Conversely, it was hypothesized that there would be a higher rate of rehospitalization for those without guardians than those with guardians.
Discharge location.
It was predicted that people with guardians would be discharged to more restrictive levels of care than those without guardians. Chi square analyses revealed a significant relationship between the level of restrictiveness of discharge setting and guardianship status, X2 (3) = 9.70, p =0.02 (See Table 20).
Table 20
Relationship Between Level of Discharge Location Restrictiveness
and Guardianship Status (N=147)
| Guardianship Status |
|Discharge Location Restrictiveness |Guardian |No Guardian |Total |
| |1 |5 |6 |
|1 – Same or Higher Restrictiveness (LRC Transfer) | | | |
| |27 |35 |62 |
|2 - Psychiatric Residential Rehabilitation | | | |
| |34 |26 |60 |
|3 - Assisted Living | | | |
| |4 |15 |19 |
|4 – Independent Living/Living with Family | | | |
| |66 |81 |147 |
|Total | | | |
There are no differences in the relative number of people from each group who went to discharge locations characterized as highest in restrictiveness (i.e., LRC transfers and Psychiatric Residential Rehabilitation). However, partial support for the hypothesis is generated by the distribution of discharges to the least restrictive setting, independent living. Not even 1% of people with guardians were discharged to independent living (n = 4 out of 66) whereas nearly 20% of those without guardians were discharged to independent living situations (n = 15 out of 81).
Rehospitalization rate.
Following discharge, CTP participants spent and average of 89.1% of days during the follow-up period in the community rather than in the hospital. Seventy-four of the 123 people with rehospitalization data available were not rehospitalized from the time of discharge through the time the outcome data was collected. The amount of time for follow-up period varies considerably because discharges from CTP happened from 1996 until 2004, therefore the number of days between CTP discharge and when the outcome data was collected was included as a covariate in analyses. There are no significant differences between those with and without guardians on the total number of days hospitalized after CTP discharge, the number of days hospitalized within the first 6 months, the number of days hospitalized within the first year, or the number of days hospitalized in the first year and a half, all Fs < 2.07, all ps > .13. There is, however, a significant difference in the number of days before the first hospitalization following discharge F (2, 69) = 5.82, p = .005, η2 =.144. Contrary to hypothesis, people with guardians were rehospitalized sooner than those without guardians (Ms = 213.46 days and 268.12 days, respectively). While the differences between groups on the total number of hospital days and the number of hospital days in six month intervals following discharge were not statistically significant, the means were all in the direction suggesting more hospital days for those with guardians than those who do not have guardians. These results suggest poorer outcome in terms of rate of rehospitalization for people with guardians.
As described in the methods section, the rehospitalization data was considerably skewed and could not be transformed or windsorized into an acceptable skewness range. This is largely a function of the large proportion of people who were never rehospitalized during the follow-up period. In order to corroborate the above results since the data used in the above analyses were skewed, categorical variables of “recidivists” and “non-recidivists” were created based on all or none cutoff levels. That is, people who were never rehospitalized were categorized as “non-recidivists” whereas if they were ever rehospitalized they are “recidivists.” Chi square analysis revealed no significant relationship between guardianship status and recidivist categorization, X2(1) = 2.22, p = .14. Likewise, recidivist categorizations were made based on rehospitalization in six month intervals following discharge (i.e., rehospitalized within that time frame = recidivist, not rehospitalized within that time frame = non-recidivist). At six months post-discharge, there is not a significant relationship between guardianship status and recidivist categorization, X2(1) = 1.20, p = .29. At 12 months post-discharge, again no significant relationships between guardianship status and recidivist categorizations are found, X2(1) = 3.29, p = .07. Finally, at 18 months post-discharge, there remains no relationship between the two variables, X2(1) = 2.22, p = .14.
In summary, results of analyses of outcome data with regard to guardianship status indicate that at the time of discharge and during the follow-up period after discharge, relatively few differences between those with guardians and those without guardians are evident. However, as hypothesized, people with guardians appear to move to more restrictive settings at the time of discharge. As for the second part of the hypothesis, there is no evidence to suggest that people without guardians recidivate more than those with guardians. To the contrary, there is some evidence that people with guardians return to the hospital sooner than those without guardians.
Chapter 5 – Discussion
General Discussion.
The purpose of this study was to provide an empirical investigation of guardianship as it relates to clinical functioning and outcome within an SMI population involved in psychiatric rehabilitation. To date, no such study is known to exist. In general, there are no known studies examining the relationship between guardianship and clinical functioning in any population. Because of the paucity of research in this area, this study was highly exploratory in nature, giving a first look at the clinical correlates of the legal construct of guardianship.
Overall, there was mixed support for the hypotheses of the study. In terms of the first hypothesis evaluating differences between those with guardians and those without guardians on several domains of functioning at admission and over the course of treatment, only one stable difference in functioning was identified between the two primary groups of interest. Behavioral functioning, as measured by the NOSIE, was poorer upon admission and consistently over the course of treatment for those with guardians than those without guardians. In general, other areas of functioning evaluated in this study – neurocognitive functioning, social cognitive functioning, treatment compliance, and symptomatology – did not differ with regard to guardianship status. Behavioral functioning may have emerged as the only significant difference between these two groups because it represents a comparatively molar level of functioning to the neurocognitive, social cognitive, and symptomatology measures used in this study. It may be the case that legal decisions regarding guardianship have more to do with molar levels of behavior rather than molecular levels of cognition and symptomatology. Intuitively, this makes sense in that guardianship proceedings, especially those in which a family member petitions for guardianship, are typically accompanied by testimony regarding the individual’s behavior in areas such as self-care, finances, work, and treatment-related decisions. The results of this study suggest that guardianship determinations for those with severe mental illness in Nebraska are, at least in part, related to behavioral levels of functioning rather than more molecular levels of functioning.
Most states, including Nebraska, use the Uniform Probate Code to make judicial decisions regarding competency and guardianship. This standard defines an incapacitated person as one who “lacks sufficient understanding or capacity to make or communicate responsible decisions” (Reisner et al., 1999, p. 869). However, some states use a more functional approach which considers a person’s ability to complete basic activities of daily living (Reisner, et al., 1999). While the line between the two approaches is not at all clear, the findings of this study appear consistent with the latter despite the fact that Nebraska operates under the Uniform Probate Code. The central premise of the Uniform Probate Code is the ability to “make or communicate responsible decisions,” rather than the ability to complete necessary tasks. This distinction means that people with serious physical impairments, like paralysis due to spinal cord injury, who are functionally unable to care for themselves would not meet the criteria for guardianship under the Uniform Probate Code. Therefore, guardianship determinations under the Uniform Probate Code often include both psychological evaluation of cognitive capacites and behavioral evidence of functioning level in order to demonstrate that a person is unable to make or communicate responsible decisions. This study, however, identified a relationship between guardianship status and behavioral levels of functioning and not other domains of functioning considered related to decision-making abilities.
With the exception of behavioral functioning, in general, few differences between those with guardians and those without with regards to the various domains of functioning were evident at admission or over the course of treatment. This finding muddies the waters in terms of bridging that gap between mental health law and theory. In other words, if the only clinical correlate related to legal findings of incompetence within this population is in behavioral functioning, the psychological construct of decision-making seems grossly out-of-line with the legal construct, which may in fact be the case. The relationship between clinical functioning and guardianship remains unclear since many of the more molecular measures, as opposed to the molar level of functioning captured by the NOSIE, are hypothesized to be correlates of decision-making abilities. Previous research has identified relationships between cognitive functioning, symptomatology, and decision-making (e.g., with psychiatric symptomatoloty, Grisso, Appelbaum, & Hill-Fotouhi, 1997; with neurocognition, Shurman, Horan, & Nuechterlein, 2005; Hutton et al., 2002). Since guardianship requires a legal finding of incompetence and incompetence implies impaired decision-making, it is surprising that no clear distinctions between those with guardians and those without guardians were evident on measures of neurocognition. Likewise, since higher psychiatric symptomatology has been linked to impaired decision-making (Grisso, Appelbaum, & Hill-Fotouhi, 1997), it is surprising that this study did not identify symptomatology as related to guardianship status. The lack of convergence between the current study and previous studies may be because of varied conceptualizations of the construct of decision-making. Take the research cited above, for example. While the work of Appelbaum, Grisso, and colleagues used a measure that is theoretically based on the legal standards associated with findings of incompetence, the measure of decision-making used in the work of Shurman, Horan, and Nuechterlein was the Iowa Gambling Task (Bechara, Damasio, Tranel, & Anderson, 1994) which involves selecting cards that have varying punishment and reward profiles. It is hard to imagine two measures that seem more dissimilar, at least in terms of face validity. However, they are discussed as both measuring “decision-making abilities” – and perhaps they do. While correlates of decision-making are often sought in terms of symptomatology, neurocognitive variables, brain injury, diagnosis, etc., the relationship between measures of decision-making and legal incompetence has not been studied. It probably comes as no surprise, then, that this study questions if psycholgical assessment of decision-making is at all similar to legal assessment. It is likely the case that the psychological construct of decision-making is not as closely related to the legal construct of decision-making as we might presume. It is also possible that while the constructs are fairly similar, the way they get operationalized in the treatment and legal contexts varies.
Another potential explanation for the hypotheses of this study going largely unsupported may be found in the nature of the population from which study sample was drawn. Because of the severity and chronicity of psychiatric disorder within the CTP population, the general lack of differences between groups is perhaps attributable to the overall high level of impairment present in this specific population. That is, the CTP participants represent a particularly treatment-refractory population and differences in functioning between groups those with guardians and those without guardians may be less apparent than they would be in a less severe or less chronic psychiatric population with more variability. Discriminating between any groups within the CTP population may require not only the existence of differences, but substantial differences. That is not to say that there is not considerable heterogeneity within this group, but it may mean that fine, subtle differences between groups may be hard to detect, and these differences may or may not be meaningful. This conclusion is congruent with studies which have had difficulty detecting treatment effects between groups in the CTP setting (Spaulding, Reed, Sullivan, Richarson, & Weiler, 1999b).
Since this study identified relatively few differences in functioning based on guardianship status, one wonders if there were a “screening procedure” of sorts, hypothetically speaking, as a standard component of admission if all people would be appointed guardians. As in, would every person in this population meet the legal criteria for a finding of incompetence? Now, this study is not intended to suggest that the entire population of the psychiatric rehabilitation program at the state hospital in Nebraska is in need of guardianship. Rather, it raises questions regarding the lack of specificity in the legal criteria for guardianship and the reasons for which guardianship is acquired. We may presume that most people with mental illnesses have guardians because their disorders are particularly severe and/or chronic. Since the population in this study is by nature both severe and chronic, there must be another factor involved in guardianship since only around half in this study were affected by guardianship. This third factor could be described as the resources the person has available in terms of family support and involvement, financial assets, and the like. This study did not evaluate the aspects of individual resources which may have led to guardianship, but did identify a greater likelihood of having a family member as a guardian at admission than if the guardian was acquired during hospitalization. Further study of “the road to guardianship” is needed.
In summary, results pertaining to the first hypothesis of this study regarding differences in functioning as a function of guardianship indicate that most differences are apparent in the area of behavioral functioning. Differences between groups in other areas of functioning seem negligible. Therefore, one clinical correlate to guardianship, as identified by this study, is behavioral functioning.
With regards to the second hypothesis of this study regarding outcome, again, there was only mixed support. Results suggest that people with guardians are discharged to more restrictive settings than those without guardians. People with guardians in this study were rarely discharged to an independent living situation whereas nearly one out of every five people who did not have guardians went on to independent living following discharge from CTP. There are at least two probable driving forces behind this pattern of discharge. First, treatment teams involved in planning discharges may not consider independent living an option for a person who has been found incompetent. Treatment teams are aware of the social history of each person and it is likely that any independent living situation for people with guardians prior to admission at CTP has failed. In fact, results from this study support this theory. Because behavioral functioning, including social competence and neatness, were related to guardianship status it could be inferred that people with guardians were unable to maintain adequate self-care or an appropriate level of care for an apartment or home. Based on past failures in independent living, treatment teams may be less inclined to consider it as a future discharge option. Second, and perhaps even more likely, discharges are largely dictated by the community providers themselves. That is, the less restrictive the setting is, the less likely it would be to accept someone who has a guardian. The availability of a guardian might make a client more attractive to some providers. Therefore, those with guardians are likely placed in more restrictive settings than those without guardians.
The second part of the outcome hypothesis pertained to rate of rehospitalization. Results of this study cannot be used to fully substantiate nor disprove the notion that people without guardians have higher rates of rehospitalization and that guardianship may serve as a protective mechanism from rehospitalization. However, results suggest that, on the contrary, those with guardians are rehospitalized sooner following discharge than those without guardians. In reality, access to inpatient hospitalization is facilitated by having a guardian because a VpG admission may be used instead of more time-consuming civil commitment proceedings. This may account for the relatively faster return to the hospital for those with guardians. However, as is the case in this study, exploratory analyses revealed that a fair amount of people with guardians do not return to the hospital until they are civilly committed. More qualitative analyses or case studies regarding rehospitalization rate for those with guardians and those without may be warranted to further evaluate this hypothesis.
Finally, exploratory analyses provided further insight into the nature of guardianship. Longer lengths of stay were found to be associated with those with guardians. The protracted length of stay among those with guardians is partially attributable to those who acquire guardians while at CTP. Often, guardians are acquired at CTP because it has been determined by the treatment team that it will be difficult to find a community placement for an individual unless a guardian is in place prior to discharge. As noted earlier, community providers sometimes make decisions about whether to accept a new admission based on whether or not a guardian is in place. Therefore, not only are lengths of stay protracted for this group because of identifying an appropriate placement, but often because community placements sometimes prefer or require guardianship to be established prior to placement at the facility. The court proceedings to establish guardianship alone extend the length of stay for those acquiring guardians at CTP.
Longer lengths of stay for those with guardians may also suggest that this subpopulation is a particularly treatment-refractory subpopulation within CTP. However, an alternative explanation to there being a subpopulation of non-responders is that instead, there is a subpopulation of people with high interepisodic symptomatology. The seemingly anomalous finding of a significant difference in total BPRS scores at 18 months, with those with guardians having a higher level of symptomatology than those without guardians, may be evidence of a group with higher symptomatology at admission that does not decrease over the course of treatment.[8] At admission and early on in treatment, higher symptom ratings for this group may have been masked by an overall higher level of symptoms early in treatment that, for most, decreased with treatment. Only later in treatment, then, when many have been discharged and most of those remaining have experienced symptom reduction, do those whose symptom picture has remained largely unchanged – and comparatively higher - become identifiable. If this is the case, coupled with the finding those with guardians have a longer length of stay, high interepisodic symptomatology may prove to be a vulnerability factor for having a guardian. Further investigation is needed.
The difference in level of education found between the two groups is probably due to an earlier age of onset of disorder[9] for those with guardians which would have truncated their education. Since those with guardians were not different from those without guardians with regard to neurocognitive functioning, it seems unlikely that actual cognitive ability contributed to the difference between the groups on years of education.
A few anomalous findings in the study deserve discussion. At 18 months of treatment, more differences in symptomatology on BPRS items were found between groups. However, while several items on the BPRS showed significant differences at different time points, there was no stability to these differences over time. As discussed earlier, those with guardians who are still in treatment at 18 months may have higher interepisodic symptomatology. Results may also simply be affected by the low number of people available at that time point for analyses.
Finally, none of the results of this study suggest differences in functioning or outcome for people whose guardians are family members or not or people who are admitted via civil commitment or Voluntary per Guardian.
Before moving on to discuss limitations of this study and areas for future research, a word of caution is warranted. Because the population in this study does represent a particularly severe and chronic subpopulation of those with mental illnesses, those who are civilly committed and those who have guardians in this population are likely quite different from those who are civilly committed or have guardians in the general population. Since this area of research has important policy implications and because policy surrounding guardianships is constantly evolving, it is important to acknowledge that any conclusions drawn from this study regarding guardianship may or may not generalize to the larger population of those with guardians. However, since the population in this study is among the most highly affected by guardianship, results represent a significant contribution both to the literature and to future policy decisions.
Limitations of the Present Study.
The goal of identifying correlates of clinical functioning to the legal construct of guardianship in this study potentially missed one key intermediate variable: decision-making. One weakness of this study is that it did not include an explicit measure of decision-making ability. Since this study utilized archival clinical data, only measures part of routine assessment at CTP were available. While poor cognitive functioning has been linked to impaired decision-making (e.g., Shurman, Horan, & Nuechterlein, 2005; Hutton et al., 2002) and this study included various measures of neurocognitive functioning as well as a measure of functional insight, no relationship was identified between guardianship and these molecular levels of functioning. It may be that a more global measurement of decision-making would better capture the relationship between impairments in decision-making and the legal finding of incompetence as opposed to the more molecular level of functioning that was captured by the measures in this study. Because the NOSIE captures a more molar level of functioning and identified differences between the groups, it seems logical that a more global measure of decision-making might better discriminate between these two groups. The MacCAT-T (Grisso, Appelbaum, & Hill-Fotouhi, 1997) is one example of the kind of measure that could be employed towards this goal.
Another limitation of this study was low power for several key analyses due to missing data and/or changes in the clinical assessment battery. The use of multivariate analyses in this study was undermined by a low number of valid cases on many of the variables over time. Therefore, conclusions about differences between groups, or the lack thereof, over the course of treatment are tentative due to insufficient power. However, consultation of a standard power table reveals that given the F-values obtained for most results in this study, an infinitely large sample size would likely have been necessary to detect any differences between groups (Friedman, 1982; Cohen, 1988). Therefore, it seems unlikely that any effects were “missed.” On the other hand, due to the exploratory nature of this study, many analyses were conducted, inflating the Type I error rate. Confident rejection of the null hypothesis is complicated because of the risk of experimentwide Type I error. At any rate, this study explored the characteristics of guardianship such that future studies in this area can make more informed research hypotheses using stricter constraints in research design to circumvent problems related to lack of power and missing data.
Future Directions.
Areas of needed research have already been alluded to in the above discussion. Specifically, a replication or study similar to the one undertaken here, with more defined hypotheses and greater power is needed to further clarify the characteristics of those with guardians. Likewise, a similar study in a broader population would allow for more generalizability of results. Finally, a study – or perhaps, initially, a series of case studies – examining the different events proceeding guardianship hearings and the resources available to each person is necessary to better understand why some people end up with guardians and others do not given the relative lack of differences in functioning that this study identified.
In addition, as implied above, research studies using a more molar or explicit measure of decision-making are needed. The purpose of such a line of research would be two-fold. First, whether or not current measures of decision-making (e.g., MacCAT-T, Grisso & Appelbaum) in fact measure what they purport to measure with regards to legal incompetence must be established. While the MacCAT-R is theoretically based on the legal standards related to decision-making (Appelbaum & Grisso, 1995; Roth, Meisel, & Lidz, 1977), no validation within a population of people determined to be legally incompetent has been conducted. Therefore, using measures of decision-making to identify correlates to clinical functioning is moot until a relationship between the psychological construct and the legal construct has been established. Studies identifying correlates of decision-making to various clinical variables have not been conducted within a population defined as legally incompetent. Or, if they have, legal status has not been considered in analyses. Such a study might simply include measures of decision-making, like the MacCAT-T, within a population such as the one in this study to determine if the measures make distinctions between those with guardians (who have been found legally incompetent) and those without guardians.
In general, more empirical, as opposed to theoretical investigations, of the concepts found within the intersecting field of mental health law are needed. The current study is a critical first step in identifying the relationship between the legal construct of guardianship and the clinical functioning of those defined by it.
References
American Psychological Association, Division 12 (n.d.). Empirically Supported Treatments. Retrieved June 6, 2004, from .
Anthony, W., Buell, G., Sharratt, S., & Althoff, M. (1972). Efficacy of psychiatric rehabilitation. Psychological Bulletin, 78, 447-456.
Anthony, W.A., Cohen, M.R., Vitalo, R. (1978). The measurement of rehabilitation outcome. Schizophrenia Bulletin, 4, 365-383.
Appelbaum, P.S. & Grisso, T. (1988). Assessing patients’ capacities to consent to treatment. New England Journal of Medicine, 319, 1635-1638.
Appelbaum, P.S. & Grisso, T. (1995). The MacArthur Treatment Competence Study. I: Mental illness and competence to consent to treatment. Law and Human Behavior, 19(2), 105-126.
Appelbaum, P.S. & Roth, L.H. (1982). Competency to consent to research: A psychiatric overview. Archives of General Psychiatry, 39, 951-958.
Army Individual Test Battery. (1944). Manual of directions and scoring. Washington D.C.: War Department, Adjutant General’s Office.
Bailley, S., Lachar, D., Rhoades, H.M., Diefenbach, G.J., Espadas, A., & Varner, R.V. (2004). Quantifying symptomatic change during acute psychiatric hospitalizations using new subscales for the anchored Brief Psychiatric Rating Scale. Psychological Services, 1, 68-82.
Bechara, A., Damasio, A., Damasio, H., & Anderson, S. (1994). Insensitivity to future consequences following damage to humanprefrontal cortex. Cognition, 50, 7–15
Birchwood, M., Smith, J., Drury, V., Healy, J., Macmillan, F., & Slade, M.A. (1994). A self-report insight scale for psychosis: reliability, validity, and sensitivity to change. Acta Psychiatrica Scandinavica, 89, 62-67.
Burger, G.K., et al. (1997). Factor structure of the expanded Brief Psychiatric Rating Scale. Journal of Clinical Psychology, 53, 451-454.
Ciompi, L. (1989). The dynamics of complex biological-psychosocial systems: four fundamental psycho-biological mediators in the long-term evolution of schizophrenia. British Journal of Psychiatry, 155(suppl. 5), 15-21.
Cohen, J. (1988). Set correlation and contingency tables. Applied Psychological Measurement, 12(4), 425-434.
Cook, J.A., Pickett, S.A., Razzano, L., Fitzgibbon, G., Jonikas, J.A., & Cohler, J.J. (1996). Rehabilitation services for people with schizophrenia. Psychiatric Annals, 26, 97-104.
Cook, J.A. & Wright, E.R. (1995). Medical sociology and the study of severe mental illness: Reflections on past accomplishments and directions for future research. Journal of Health and Social Behavior, 36, 95-114.
David, A.S. (1990). Insight and psychosis. British Journal of Psychiatry, 156, 798-808.
Drane, J.F. (1984). Competency to give an informed consent: A model for making clinical assessments. Journal of the American Medical Association, 252, 925-927.
Friedman, H. (1982). Simplified determinations of statistical power, magnitude of effect and research sample sizes. Educational & Psychological Measurement, 42(2), 521-526.
Garety, P. A., Fowler, D., & Kuipers, E. (2000). Cognitive-behavioral therapy for medication resistant symptoms. Schizophrenia Bulletin, 26, 73-86.
Gold, J.M., Queern, C., Iannone, V.N., & Buchanan, R.W. (1999). Repeatable Battery for the Assessment of Neuropsychological Status as a screening test in schizophrenia, I: Sensitivity, reliability, and validity. American Jouranl of Psychiatry, 156, 1944-1950.
Grisso, P.S. (1986). Evaluating competencies: Forensic assessments and instruments. New York: Plenum Press.
Grisso, T. & Appelbaum, P.S. (1995). The MacArthur Treatment Competence Study. III: Abilities of patients to consent to psychiatric and medical treatments. Law and Human Behavior, 19(2), 149-174.
Grisso, T., Appelbaum, P.S., & Hill-Fotouhi, C. (1997). The MacCAT-T: A clinical tool to assess patient’s capacities to make treatment decisions. Psychiatric Services, 48 (11), 1415-1419.
Grisso, T., Appelbaum, P.S., Mulvey, E.P., & Fletcher, K. (1995). The MacArthur Treatment Competence Study. II: Measures of abilities related to competence to consent to treatment. Law and Human Behavior, 19(2), 127-148.
Guy, W. (1976). ECDEU assessment manual for psychopharmacology, revised (DHEW Publication No. ADM 76-338). Rockville, MD: U.S. Department of Health, Education, and Welfare.
Hatfield, B., Bindman, J., & Pinfold, V. (2004). Evaluating the use of supervised discharge and guardianship in cases of severe mental illness: A follow-up study. Journal of Mental Health, 13(2), 197-209.
Hiday, V.A., Swartz, M.S., Swanson, J.W., Borum, R., & Wagner, H.R. (2002). Impact of outpatient commitment on victimization of people with severe mental illness. American Journal of Psychiatry, 159(8), 1403-1411.
Hoaglin, D., Mosteller, F., & Tukey, J. (1983). Understanding Robust and Exploratory Data Analysis. Wiley, New York.
Hobart, M.P., Goldberg, R., Bartko, J.J., & Gold, J.M. (1999). Repeatable Battery for the Assessment of Neuropsychological Status as a screening test in schizophrenia, II: Convergent/discriminant validity and diagnostic group comparisons. American Jouranl of Psychiatry, 156, 1951-1957.
Hofman, S.G., & Tompson, M.C. (2002). Treating chronic and severe mental disorders: A handbook of empirically supported interventions. New York: Guilford Press.
Hoge, S.K., Lidz, C.W. Eisenberg, M. Gardner, W., Monahan, J. Mulvey, E., et al. (1997). Perceptions of coercion in the admission of voluntary and involuntary psychiatric patients. International Journal of Law and Psychiatry, 20, 167-181.
Hopko, D.R., Averill, P.M., Cowan, K., & Shah, N. (2002). Self-reported symptoms and treatment outcome among non-offending involuntary inpatients. The Journal of Forensic Psychiatry, 13, 88-106.
Honigfeld, G., Roderic, D., & Klett, J. (1966). NOSIE-30: A treatment-sensitive ward behavior scale. Psychological Review, 19, 180-182.
Iyer, S.N., Rothmann, T.R., Vogler, J.E., & Spaulding ,W.D. (2005). Evaluating outcomes of rehabilitation for severe mental illness. Rehabilitation Psychology, 50(1), 43-55.
Kendall, P.C. (1998). Empirically supported psychological therapies. Journal of Consulting and Clinical Psychology.
Krampen, G. (1991). Fragebogen zu kompetenz-und kontrolluebersengungen (FKK), Handanweisdung (1 ed.). Goettingen: Hogrefe Verlag fuer Psychologie.
Lehman, A.F., Thompson, J.W., Dixon, L.B., & Scott, J.E. (1995). Introduction. Schizophrenia: Treatment outcomes research. Schizophrenia Bulletin, 21, 561-566.
Monahan, J., Hoge, S.K., Lidz, C., Roth, L.H., Bennett, N., Gardner, W., & Mulvey, E. (1995). Coercion and commitment: Understanding involuntary mental hospital admission. International Journal of Law and Psychiatry, 18, 249-263.
Lukoff, D., Nuechterlein, K. H., & Ventura, J. (1986). Manual for the expanded Brief Psychiatric Rating Scale (BPRS). Schizophrenia Bulletin, 12, 594-602.
Lyall, D., Hawley, C., & Scott, K. (2004). Nurses’ Observation Scale for Inpatient Evaluation: Reliability update. Journal of Advanced Nursing, 46 (4), 390-394.
McEvoy, J., Scheifler, P., & Frances, A. (Eds.). (1999). Treatment of schizophrenia. Journal of Clinical Psychiatry, 60 (Suppl.11), 4-80.
McKenna, B.G., Simpson, A.I.F., & Coverdale, J.H. (2003). Patients’ perceptions of coercion on admission to forensic psychiatric hospital: a comparison study. International Journal of Law and Psychiatry, 26, 355-372.
Meyers, J. E. & Meyers, K. R. (1995). The Meyers scoring system for the Rey Complex Figure Test and recognition trial: Professional manual. Odessa: Psychological Assessment Resources.
Monahan, J., Hoge, S.K., Lidz, C., Roth, L.H., Bennett, N., Gardner, W., & Mulvey, E. (1995). Coercion and commitment: Understanding involuntary mental hospital admission. International Journal of Law and Psychiatry, 18, 249-263.
Neb.Rev.Stat. §30-2601 (1997).
Neb.Rev.Stat. §30-2619 (1997).
Neb.Rev.Stat. §30-2623 (1997).
Neb.Rev.Stat. §30-2630 (1993).
Neb.Rev.Stat. §83-1068 (2004).
Palmer, B.W., Nayak, G.V., Dunn, L.B., Appelbaum, P.S., & Jeste, D.V. (2002). Treatment-related decision-making capaciety in older patients with psychosis: A preliminary study using the MacCAT-T and HCAT. American Journal of Geriatric Psychiatry, 10(2), 207-211.
Panos, P.T. (2004). The validation of the factor structure of the Brief Psychiatric Rating Scale-Expanded Version (BPRS-E) with geriatric and nongeriatric psychiatric inpatients. Research on Social Work Practice, 14(3), 180-190.
Paul, G.L. & Lentz, R.J. (1977). Psychosocial treatment of chronic mental patients: Milieu vs. social learning programs. Cambridge, MA: Harvard University Press.
Paul, G.L., & Menditto, A.A. (1992). Effectiveness of inpatient treatment programs for mentally ill adults in public psychiatric facilities. Applied and Preventive Psychology, 1, 41-63.
Penn, D., Mueser, K., Spaulding. W., Hope, D., & Reed, D. (1995). Information processing and social competence in chronic schizophrenia. Schizophrenia Bulletin, 21(2), 269-281.
Perlick, D.A., Rosenheck, R.A., Clarkin, J.F., Sirey, J., & Raue, P. (1999). Symptoms predicting impatient service use among patients with bipolar affective disorder. Psychiatric Services, 50(6), 806-812.
Petrila, J., Ridgely, M.S., & Borum, R. (2003). Debating outpatient commitment: Controversy, trends, and empirical data. Crime & Delinquency, 49(1), 157-172.
Randolph, C. (1998). Repeatable battery for the assessment of neuropsychological status. San Antonio: Harcourt Brace & Company.
Reisner, R., Slobogin, C., & Rai, A. (1999). Law and the mental health system: Civil and criminal aspect. St. Paul, MN: West Group.
Ridgely, M.S., Borum, R., & Petrila, J. (2001). The effectiveness of involuntary outpatient treatment: Empirical evidence and the experience of eight states. (MR-1340-CSCR). Retrieved September 22, 2004, from the RAND Corporation Web site:
Rey, A. (1941). L’examen psychologique dans les cas depression’encephalopathie traumatique. Archives de Psychologie, 28, 286-340.
Roth, L.H., Lidz, C.W., Meisel, A., Soloff, P.H., Kaufman, K., Spiker, D.G., & Foster, F.G. (1982). Competency to decide about treatment or research: An overview of some empirical data. International Journal of Law and Psychiatry, 5, 29-50.
Roth, L.H., Meisel, A., & Lidz, C.W., (1977). Tests of competency to consent to treatment. American Journal of Psychiatry, 134, 279-284.
Schmidt M. (1996). Rey Auditory Verbal Learning Test. Los Angeles: Western Psychological Services.
Schopp, R.F. (1998). Civil commitment and sexual predators: competence and condemnation. Public Policy and the Law, 4, 323-376.
Schopp, R.F. (1999). Therapeutic jurisprudence: Integrated inquiry and instrumental prescriptions. Behavioral Sciences and the Law, 17, 589-605.
Schopp, R.F. (2001). Competence, condemnation, and commitment: An integrated theory of mental health law. Washington, D.C.: American Psychological Association.
Shaw, J., Hatfield, B., Evans, S. (2000). Guardianship under the Mental Health Act 1983. Psychiatric Bulletin, 24, 51-52.
Sim, A., Peer, J., Ritchie, A.J., & Spaulding, W. (2003). The Rey-Osterrieth and RBANS complex figure tasks: A systematic comparison. Poster presented at the 2003 conference of the International Neuropsychological Society in Honolulu, HI.
Slobogin, C. (1995). Therapeutic jurisprudence: Five dilemmas to ponder. Psychology, Public Policy, & Law, 1, 193-219.
Spaulding, W.D. (1997). Cognitive models in a fuller understanding of schizophrenia. Psychiatry, 60, 341-346.
Spaulding, W., Garbin, C.P., & Dras, S.R. (1989). Cognitive abnormalities in schizophrenic patients and schizotypal college students. Journal of Nervous and Mental Disease, 177(12), 717-728.
Spaulding, W., Hargrove, D.S., Crinean, W.J., & Martin, T. (1981). A microcomputer-based laboratory for psychopathology research in rural settings. Behavior Research Methods & Instrumentation, 13(4), 616-623.
Spaulding, W., Johnson, D. & Coursey, R. (2001). Treatment and rehabilitation of schizophrenia. In M. Sammons & N. Schmidt (Eds.) Combined Treatments for Mental Disorders: A Guide to Psychological and Pharmacological Interventions. Washington, D.C.: American Psychological Association.
Spaulding, W.D., Poland, J., Elbogin, E., & Ritchie, A.J. (2000). Applications of therapeutic jurisprudence in rehabilitation for people with severe and disabling mental illness. Thomas M. Cooley Law Review, 17(1), 135-170.
Spaulding, W.D., Fleming, S.K., Reed, D., Sullivan, M.E., Storzbach, D., & Lam, M. (1999a). Cognitive functioning in schizophrenia: Implications for psychiatric rehabilitation. Schizophrenia Bulletin, 25, 275-289.
Spaulding, W.D., Reed, D., Sullivan, M., Richardson, C., & Weiler, M. (1999b). Effects of cognitive treatment in psychiatric rehabilitation. Schizophrenia Bulletin, 25, 657-676. THIS ONE HAS BPRS.
Spaulding, W.D., Sullivan, M., & Poland, J. (2002). Treatment and rehabilitation of severe mental illness: A comprehensive approach. New York: Guildford Publications.
Strachan, E.D. (2004). Evaluation of education and skill training regarding involuntary hospitalization for psychiatric inpatients (Unpublished doctoral dissertation, University of Nebraska-Lincoln, 2004).
Strauss, J.S. (1989). Mediating processes in schizophrenia toward a new dynamic psychiatry. British Journal of Psychiatry, 155 (suppl. 5), 22-28.
Tepper, A., & Elwork, A. (1984). Competence to consent to treatment as a psycholegal construct. Law and Human Behavior, 8, 205-223.
Twamley, E.W., Jeste, D.V., & Bellack, A.S. (2004). A review of cognitive training in schizophrenia. Schizophrenia Bulletin, xx, xxx-xxx.
Uniform Probate Code §5 (1998).
Van der Does, A.J.W., Linszen, D.H., Dingemans, P.M., Nugter, M.A. & Scholte, W.F. (1993). A dimensional and categorical approach to the symptomatology of recent-onset schizophrenia. The Journal of Nervous and Mental Disease, 181, 744-749.
Wallace, C.J., Liberman, R.P., Kopelowicz, A. & Yaeger, D. (2001). Psychiatric rehabilitation. In G.O. Gabbard (Ed.), Treatment of psychiatric disorders, 3rd edition (pp.1093-1112). Washington, D.C., American Psychiatric Publishing.
Wexler, D.B. (1988). Reforming the law in action through empirically grounded civil commitment guidelines. Hospital & Community Psychiatry, 39, 402-405.
Wexler, D.B. (ed.) (1990). Therapeutic Jurisprudence: The law as a therapeutic agent. Durham, N.C.: Carolina Academic Press.
Wexler, D.B., & Winick, B.J. (Eds.) (1991). Essays in Therapeutic Jurisprudence. Durham, N.C.: Carolina Academic Press.
Whitaker, R. (2002). Mad in America: Bad science, bad medicine, and the enduring mistreatment of the mentally ill. Cambridge, MA: Perseus Publishing.
Winick, B.J. (1995). Ambiguities in the legal meaning and significance of mental illness. Psychology, Public Policy, and Law, 1, 534-611.
Winick, B.J. (1996a). The MacArthur Competency Study: Legal and therapeutic implications. Psychology, Public Policy, and Law, 2, 137-166.
Winick, B.J. (1996b). Advance directive instruments for those with mental illness. University of Miami Law Review, 51, 57 (1996).
Appendix:
Bivariate Correlation Matrices
Table A1
Bivariate Correlations for Demographic and Clinical Characteristics Used in MANOVA Analyses with Guardianship Status
| |1 |2 |3 |4 |
|1 - Length of Stay |- | | | |
|2 - Age |.127 |- | | |
|3 – 3 - Years of Education |-.180* |.251** |- | |
|4 – 4 - Age at First |-.069 |.458** |.337* |- |
|Hospitalization | | | | |
|5 – 5 - Number of Previous |.037 |.114 |-.135 |-.190 |
|Hospitalizations | | | | |
** Correlation is significant at the 0.01 level (2-tailed).
* Correlation is significant at the 0.05 level (2-tailed).
Table A2
Bivariate Correlations for Neurocognitive Variables
| |1a |2 |3 |4 |5 |6 |
|1 - NOSIE Total |- | | | | | |
|2 - Social competence |.85** |- | | | | |
|3 - Social Interest |.62 ** |.40** |- | | | |
|4 - Neatness |.83** |.70** |.47** |- | | |
|5 - Irritability |-.63** |-.60** |-.12 |-.43** |- | |
|6 - Psychoticism |-.53** |-.49** |-.21* |-.37** |.47** |- |
|7 - Motor Retardation |-.72** |-.72** |-.49** |-.65** |.21* |.23** |
** Correlation is significant at the 0.01 level (2-tailed).
* Correlation is significant at the 0.05 level (2-tailed).
Table A5
Bivariate Correlations of BPRS Total Scores and Factor Scores
| |1 |2 |3 |4 |5 |6 |
|1 - BPRS Total |- | | | | | |
|2 - Psychotic Disorganization |.69** |- | | | | |
|3 - Hallucinations/ Delusions |.60** |.25 |- | | | |
|4 - Paranoia |.58** |.62** |-.07 |- | | |
|5 - Emotional Blunting |.19 |.15 |-.20 |.06 |- | |
|6 - Anxiety/Depression |.62** |.06 |.50** |.16 |-.22 |- |
|7 - Agitation/Elation |.48** |.37* |.24 |.38* |.07 |-.22 |
** Correlation is significant at the 0.01 level (2-tailed).
* Correlation is significant at the 0.05 level (2-tailed).
-----------------------
[1] While “autonomy” can have varying definitions, it is used here and throughout this dissertation to refer to an individual’s right to make self-regarding decisions. For a more in depth discussion of autonomy and its varying definitions as it pertains to competence and commitment, see Schopp, 2001.
[2] It is not within the scope of this study to discuss all of these concepts in detail. However, it should be noted that aspects of police power commitment (e.g., Not Guilty/Responsible by Reason of Insanity – NGRI/NRRI; competency to stand trial - CST), in particular, are out of the scope of this review in that the archival database does not come from a forensic population. It is germane, however, to discuss issues relevant to guardianship, such as parens patriae civil commitment and competency.
[3] Durable power of attorneys and advance directives (see Winick, 1996b, for a discussion) are offshoots of the mechanism of guardianship. Both serve the purpose of allowing individuals in a period of competency to plan for future periods of incompetence by expressly dictating their desires for treatment, or refusal of treatment. However, like limited guardianship, they have received much praise, but they have not been widely practiced.
[4]About ¾ of the patients hospitalized for depression performed adequately on the four measures combined.
[5] Outpatient commitment may be a confounding variable in investigation of this project goal. It is important to note that outpatient commitment has been found to be successful in some cases, but not in others. Its utility is still being examined (Ridgely, Borum, & Petrila, 2001; Petrila, Ridgely, & Borum, 2003; Hiday, Swartz, Swanson, Borum, & Wagner, 2003).
[6] In Nebraska, psychiatric diagnosis can function to disqualify an individual for developmental disability services, so the average age of onset (or age of first psychiatric hospitalization) is sometimes unusually low.
[7] Other includes diagnoses such as, but not limited to, the following: Bipolar Disorder, Dementia, Psychotic Disorder NOS, Pervasive Developmental Disorder, Impulse Control Disorder, and Asperger’s.
[8] Post hoc analyses using change scores on BPRS total from admission to 18 months were attempted, but the valid N was only 4. With all possible combinations of change scores (e.g. admission to 12 months, 6 months to 12 months), the largest attainable N was 14, with 6 in one group and 8 in the other.
[9] The difference between groups on age of onset approached statistical significance.
................
................
In order to avoid copyright disputes, this page is only a partial summary.
To fulfill the demand for quickly locating and searching documents.
It is intelligent file search solution for home and business.
Related download
- adventhealth a leader in whole person health care
- bsrad program handbook valencia college
- after visit summary tool health services research
- outline of effective practices manual
- running head intervention for involuntary inpatients
- office based opioid treatment obot policy and procedure
- template consent for minimal risk research
- documents reports all documents the world bank
- table of contents world health organization
Related searches
- why running is good for you
- academic intervention for elementary school
- intervention for health promotion
- running a business for dummies
- boars head routes for sale
- therapeutic intervention for schizophrenia
- running a corporation for dummies
- involuntary separation pay army
- navy involuntary separation pay
- involuntary movements icd 10
- running a campaign for election
- involuntary flatulence in women