Evidence on the capacity of severe mental disorder patients to make well-founded decisions about their healthcare: a meta-review

Background Determining the mental capacity of psychiatric patients for making regarding acceptance refusal is crucial in clinical practice. This comprehensively examines the current evidence on the capacity of patients with a mental illness to make well-founded medical care decisions. Methods Systematic review of literature review articles following PRISMA recommendations. PubMed, Scopus and CINAHL were electronically searched up to 30 September 2019. Free text searches and medical subject headings in English were combined. Publications were selected as per inclusion and exclusion criteria. The AMSTAR 2 tool was used to assess the quality of reviews. Results Thirteen publications were reviewed. In one review, up to 67% of patients in a mixed psychiatric population had capacity to decide about admissions; 71% (median) had capacity for making decisions about treatments. In another, community-dwelling or clinically stable psychiatric outpatients were close to non-psychiatric subjects in decision capacity performance. In a third review, people with psychosis had moderately impaired risk-reward decision-making ability compared with healthy individuals (g =-0.57, 95% CI: -0.66 to -0.48; I 2 45%), and were more likely to value rewards over losses (k = 6, N = 516, g = 0.38, 95% CI 0.05 to 0.70, I 2 64%) and to base decisions on recent rather than past outcomes (k = 6, N =516, g = 0.30, 95% CI: -0.04 to 0.65, I 2 68%). In a fourth review, future care (crisis) planning led to a 40% reduction in the use of compulsory inpatient treatment over 15 to 18 months. In other reviews, patients with mental illness were able to provide valid preference measures and gave sufficiently consistent answers regarding their preferred treatments. Decision-making responded favourably to interventions. The publications complied satisfactorily with the AMSTAR 2 critical domains. Conclusions Whilst impairments in decision-making capacity may exist, most patients with a severe mental disorder are able to make rational decisions about their care. Best practice strategies should help mentally ill patients grow into voluntary and safe users of medications, enabling them to keep a sense of control over their lives and enhancing their health-related quality of life.

more likely to value rewards over losses (k = 6, N = 516, g = 0.38, 95% CI 0.05 to 0.70, I 2 64%) and to base decisions on recent rather than past outcomes (k = 6, N =516, g = 0.30, 95% CI: -0.04 to 0.65, I 2 68%). In a fourth review, future care (crisis) planning led to a 40% reduction in the use of compulsory inpatient treatment over 15 to 18 months. In other reviews, patients with mental illness were able to provide valid preference measures and gave sufficiently consistent answers regarding their preferred treatments. Decision-making responded favourably to interventions. The publications complied satisfactorily with the AMSTAR 2 critical domains. Conclusions Whilst impairments in decision-making capacity may exist, most patients with a severe mental disorder are able to make rational decisions about their care. Best practice strategies should help mentally ill patients grow into voluntary and safe users of medications, enabling them to keep a sense of control over their lives and enhancing their health-related quality of life.

Background
In 1995, Appelbaum and Grisso stated that competence to consent to treatment relied on four legal standards: the ability to communicate a choice; the ability to understand relevant information; the ability to appreciate the situation and its likely consequences; and the ability to manipulate information rationally (1). In healthcare, the capacity to make decisions regarding treatment is closely related to the autonomy, the exercise of self-governance, and the ability of an individual to take intentional actions (2). The capacity to consent to treatment is often used in the clinical assessment of the ability to engage in authentic autonomous decision-making, a fundamental element of a person's dignity and rights (3).
Assessment of mental capacity has become a key component of daily clinical practice (4) (5). Mental health legislation and medical ethics increasingly require physicians to empower patients to make decisions, and also to respect the patient's wishes with regard to accepting or refusing therapy (4) (6). However, it has been reported that coercive treatment, involuntary hospitalisations and medications are currently overused (7); this has a direct negative impact on patients' adherence to treatment and on their engagement and participation in shared decision-making with their healthcare professionals (8).
An increasing number of publications are assessing decision-making capacity in mental health. Systematic reviews bring together published studies in a single report in order to appraise the evidence; systematic reviews of reviews are the logical next step for comparing and contrasting the findings of individual reviews, so as to summarise the evidence and provide stakeholders with the information they need (9). This systematic review of review articles was designed as a comprehensive examination of the current state of knowledge in the field, with the aim of assessing the current evidence on the decision-making capacity of patients with various mental illness (especially schizophrenia, psychosis and bipolar disorder) with regard to their treatment and disease management. The review compares the conclusions of various comprehensive publications, discusses the strength of these conclusions, and identifies existing gaps in the evidence.

Methods
The review of the literature was conducted in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines (10). A series of steps, including the 4 definition of the search strategy, identification and selection of publications, data extraction and synthesis, and quality assessment was followed.

Search strategy for identification and selection of publications
The aim of the search strategy was to provide a comprehensive list of published literature reviews that met the inclusion criteria. Free text searches and medical subject headings were combined ( Selection of publications was carried out as per inclusion and exclusion criteria ( Table 2). Potentially relevant abstracts were assessed by two reviewers to identify all papers suitable for inclusion. Full text copies were requested. Reviews which were identified after mutual agreement were included and data were extracted. A third reviewer was involved in the process to resolve any disagreements on the selection of publications.

Data extraction and quality assessment
Data extraction was carried out by one researcher. A data extraction form that covered citation, country, population, interventions, comparators, outcomes, settings, review type, aims, literature review size, strengths and limitations and key findings of the review as stated by authors was used to extract data (Tables 3 and 4). The AMSTAR 2 (A MeaSurement Tool to Assess systematic Reviews) (11) assessment tool was used to assess the quality of reviews.

Results
A total of 714 publications were initially identified; 692 were either duplicated or deemed not relevant for the review based on the assessment of titles and abstracts; 22 full text publications were initially considered valid and retrieved for closer examination; 9 were excluded because they referred to diseases excluded from the review, 13 publications were finally included in the review (Figure 1).
The number of studies included in each review varied between seven (12) (13) and 63 (14).
The general healthy population or groups of patients with a non-mental disorder were the comparators (Table 3). In four reviews (12) (15) (19) (21) no comparisons between groups were made. The number of patients with a mental disorder included in the reviews ranged between six (18) and 4,273 (16).
Four reviews focused on schizophrenia or schizoaffective disorders only (13) (17)  was the primary assessment instrument in eight reviews, while two reviews used the Iowa or Canada Gambling Tasks to assess capacity to make decisions. Nine of the reviews included in the present report were conducted by authors in European countries (12) (15) (17) (19)(20)(21)(22)(23)(24), two were based in the United States (14) (18), one study was conducted in Australia (16), and another in several countries (13) (Table 4).

Prevalence of decision-making capacity
One systematic review of 37 empirical, quantitative studies of mental capacity in a mixed population of psychiatric patients reported that up to 67% of participants had the capacity to decide whether to be admitted to a psychiatric unit while a median of 71% had capacity for making treatment decisions (a median of 29%, interquartile range (IQR) 22-44, lacked capacity) (21). Another systematic review found that 26% (95% confidence interval (CI): 18 to 36) to 67% (95% CI: 35 to 88) of people with schizophrenia or other non-affective disorders were able to make medical decisions related or unrelated to the management of their condition (22). Overall, the definitions and measurement of capacity varied widely in the studies (21) (22).

Decisional capacity in different clinical settings
Capacity to consent to treatment or admission differed between patients in medical (non-psychiatric) settings and those in psychiatric settings. Lepping et al (2015) (20) reported that 55% of patients in psychiatric and 66% of patients in non-psychiatric settings had the capacity to make medical decisions. However, most patients in psychiatric settings were inpatients. Appreciation of the problem and necessity for treatment were more frequently compromised in psychiatric patients, while nonpsychiatric patients struggled primarily with reasoning. The authors found a significant variation between studies due to heterogeneity in designs and methods that reached 86% in psychiatric settings and 90% in non-psychiatric settings. Jeste et al (2006) (18) found a 48% to 79% overlap between people with schizophrenia and non-psychiatric patients on the MacArthur subscales, which indicated that most patients with schizophrenia had adequate decision-making capacity (18).
However, the proportion of inpatients and outpatients in the samples was highly heterogeneous.
Community-dwelling or clinically stable outpatients were much closer to non-psychiatric subjects in terms of the capacity for decision-making, but psychotic inpatients had several characteristics which distinguished them from outpatients and temporarily limited their capacity. Greater severity of positive and negative symptoms, experiencing a stressful life event (e.g., hospitalisation), and often receiving higher doses of medication adversely impacted cognition among psychiatric inpatients (20).
The authors concluded that similar proportions of non-psychiatric and psychiatric outpatients either had or lacked capacity to consent to treatment or to hospital admission, and that impairment in the capacity to make decisions was not a distinguishing feature of schizophrenia patients (18) (20).
Another meta-analysis of ten studies showed that compared to healthy controls, patients with schizophrenia or schizoaffective disorder were significantly more likely to have impaired decisionmaking capacity in terms of understanding, reasoning, appreciation and expression of a choice in clinical research and treatment, as measured by the MacArthur Competence Assessment Tool (MacCAT) instruments (13). However, some of the studies included in this meta-analysis found decisional capacity to be improved in patients with schizophrenia following intensive educational interventions. Furthermore, the standardised mean differences were more significant in older than in younger age subgroups, suggesting that, compared to their healthy counterparts, the impairment of decision-making capacity could be more obvious in older patients than in younger patients.
Another systematic review explored the degree of impairment in each dimension of decision-making capacity in schizophrenia patients compared to non-psychiatric controls, as assessed by the MacCAT 7 (17). The odds for a decreased understanding and a decreased appreciation were some five times higher in individuals with schizophrenia than in non-mentally ill controls, those for decreased reasoning almost four times higher, and those for a decreased aptitude to express a choice was over six times higher. The use of an enhanced informed consent form contributed to significant improvements in decision-making capacity compared to the use of standard forms. Decision-making capacity responded favourably to interventions, such as the simplification of the information, shared decision-making, and metacognitive training (19). The authors concluded that even if patients with schizophrenia have a significantly decreased decision-making capacity, they should be considered to be as competent as non-mentally ill controls unless very severe changes were identifiable during the clinical examination (17).

Determining factors of decisional capacity
In a systematic review and meta-analysis of factors that help or hinder treatment decision-making capacity in psychosis (23 studies, n= 1823), psychotic symptoms were found to have small, moderate and strong associations with appreciation, understanding and reasoning respectively (19). Better decisional capacity in psychiatric patients was associated with higher insight, better metacognitive ability, higher anxiety and lower perceived coercion. Psychosis, symptom severity, involuntary admission into hospital and refusal of treatment were the strongest risk factors for psychiatric patients to be judged as lacking capacity (21).

Capacity of people with mental illness to make value-based and risk-reward decisions
Since decision-making is the process of forming preferences for possible options, selecting and executing actions, and evaluating the outcome, the Iowa and the Cambridge Gambling Tasks (IGT and CGT) have been widely administered to measure decision-making capacity in mentally ill individuals, even though they are not specifically designed for use in psychiatry (25). These tasks simulate reallife decision-making situations by manipulating the possibilities and magnitudes of potential rewards and punishments in a series of hypothetical scenarios presented to the patient (26).
A systematic review and meta-analysis explored the factors which may help or hinder the ability to make risk-reward decision making in a pooled sample of 4,264 individuals with psychosis, based on their performance on the IGT and the CGT. Compared with healthy individuals, people with psychosis had moderately impaired risk-reward decision-making ability (g =-0.57, 95% CI -0.66 to -0.48; I 2 45%; moderate quality) (23). They were also more likely to value rewards over losses (k = 6, N = 516, g = 0.38, 95% CI: 0.05 to 0.70, I 2 64%), and to base decisions on recent rather than past outcomes (k = 6, N =516, g = 0.30, 95% CI: -0.04 to 0.65, I 2 68%). Analysis of the positive or negative influence of the type and dose of antipsychotics on decision-making capacity was inconclusive. The authors suggested that, although people with non-affective psychosis may make less effective decisions than healthy individuals in the IGT and CGT, their difficulties were moderate and comparable with those observed in other clinical groups.
Mukherjee and Kable (2014) (14) calculated that around 27% of patients with various mental disorders did not differ with respect to healthy individuals when deciding about losses and rewards on the IGT. Furthermore, individuals with mental illnesses had fewer deficits than individuals with frontal lobe lesions. The assessment of the severity of impairment across types of mental illness did not demonstrate any significant differences according to specific psychiatric diagnosis.

Capacity of people with mental illness to choose treatments
Eiring et al (2015) (15) investigated the relative value adults with a mental illness place on treatment outcomes, including the attributes of particular medications or medication classes and the consequences and health states associated with their use. It reported that patients were able to provide valid preference measures with the different methods applied, generally understood the tasks, and gave sufficiently consistent answers. Among patients with schizophrenia, positive, acute or psychotic symptoms appeared consistently among the least desirable outcomes. Negative symptoms, such as reduced capacity for emotion, were found more desirable or less important than positive symptoms. Independence received high ratings and inpatient status low ratings. Overall, patients with schizophrenia tended to value disease states higher and side effects lower than other groups and perceived side effects more negatively than their therapists. Patients with bipolar disorder gave low values to mania and severe depression and reported weight gain to be important.

Quality assessment
Reviews presented well-framed research questions based on the evidence-based PICOS model (27) ( Table 3) and were high quality according to the AMSTAR 2 assessment tool (Table 4) (11). AMSTAR 2 was developed to evaluate systematic reviews of randomised trials or non-randomised studies of healthcare interventions, or both. Publications included in this review complied satisfactorily the AMSTAR 2 critical domains. No critical weaknesses were identified in the assessment. Therefore, the reviews provided an accurate and comprehensive summary of the results of studies of decisionmaking capacity in mental disorder patients.

Discussion
This systematic review brings together a set of high-quality reviews on the capacity of individuals with a severe mental illness to make decisions about their healthcare. The findings reveal that patients with psychotic disorders or other severe mental illnesses have the capacity to decide about their treatments and can make complex risk-reward decisions in usual clinical practice. Small deviations from optimal performance may arise due to deficits in the ability to fully represent the value of different choices and response options; similar results have been found in experimental research in patients with schizophrenia (28).
Most of the reviews appraised addressed the capacity to make decisions in people with severe mental disorders either requiring hospital admission or already hospitalised. This means that most studies included patients with severe symptoms or more severe mental disorders. Some required electroconvulsive therapy, which is most commonly used in patients with more severe symptoms that have failed to respond to other treatments (29). Even in these more severely ill psychiatric populations, between 60% and 70% had the capacity to make some treatment decisions (1).
Hospitalised patients usually have greater care needs, even when their psychiatric symptoms are controlled, exhibit significantly more severe negative, positive and manic symptoms, and have lower global functioning than outpatients (30). Therefore, decisional capacity can be expected to be higher amongst outpatients than among inpatients, and higher in everyday life than the rates reported in the studies included here.
This review shows that people with schizophrenia have the capacity to make difficult decisions related to hospitalisation, the type of treatment they prefer to receive, and their care plans. Patients with schizophrenia or bipolar disorder are able to describe prodromal symptoms of relapse and to suggest a treatment and the need for hospitalisation in advance; they can request or refuse medications and state their preferences for pre-emergency interventions, non-hospital alternatives and non-medical personal care (31)(32)(33). It has been shown that advance directives are important to ensure the timely provision of medical treatments, thus minimising decisional impairments in the acute stages of psychosis (34).
Beyond acute episodes, the findings also support the notion that continued training and learning, simplification and enhancement of the information improve the capacity of patients with severe mental disorders for decision-making in everyday life (35). The results of various studies demonstrate that brief interventions aimed at recovering capacity for understanding can help schizophrenia patients to perform very much like healthy people in the four dimensions of decisional capacity (understanding, appreciation, reasoning and expression of a choice) (36). Regular information reinforcement, strengthening neurocognitive functioning and training are important to maintain longterm levels of competence and to maximise decision making capacities of patients (37) (40).
The MacCAT-T was the tool most frequently used to evaluate decision-making capacity in the studies reviewed. This instrument measures the individual capacity to consent to treatment with a semistructured interview tailored to the patient's specific disorder and treatment decision. Responses are rated by the clinician as 2, 1, or 0 (adequate, questionable, and inadequate), according to the specific situation or condition. There is no total score; each summary rating can be discussed separately and there are no cut-offs for individual summary scores (38). Therefore, the final decision depends entirely on clinical judgement, based on the practitioner's knowledge of the patient and of the course of the disease.
Small sample size, heterogeneity, language and selection bias of participants were among the limitations frequently reported by the authors of the studies reviewed. However, since the publications included in the present report were systematic reviews and meta-analyses, the risk of bias was minimised.
Overall, this review provides solid support for the idea that the autonomy of severe psychiatric patients is preserved, and successfully challenges the idea that people with severe mental illnesses lack capacity and are unable to make their own choices (8). It contributes to the growing body of evidence suggesting that best practice strategies should help severe mentally ill patients to grow into voluntary, safe users of medications and to recover a sense of control over their own lives, as preventing agitation and other acute symptoms and enhancing their health-related quality of life (39).
It also reveals the gaps in the evidence regarding the capacity of people with severe mental illness to make appropriate clinical decisions in everyday life in the community.
Authors across studies coincide in emphasising that most patients with a severe mental disorder are able to make rational decisions about their medical care and participate in decision-making regarding treatments despite the impairments in their decisional capacity. Impairment does not constitute incapacity to make decisions. Furthermore, advance directives are important to prevent crises and to minimise the impact of these negative events on patients' health. Clinicians play the crucial role of United Kingdom Psychosis (non-affective psychotic disorder) sizes/associations) Eiring Ø et al 2015, (15) Systematic review To investigate patients' preferences for outcomes associated with psychoactive medications. , -a small association with general symptoms: k = 5, N = 169, r = -0.13, 95% -0.25, -0.00, I2 = 0% (low quality) -no association with positive symptoms: k = 10, N = 512, r = -0.01, 95% CI -0.11 to 0.08 (moderate quality) -no association between overall psychotic symptoms: k=6, r=−0.10, 95% CI −0.21 to 0.02, I2 = 0% (very low quality).
-no association with current antipsychotic doses: N=171, r=−0.02, 95% CI −0.17 to 0.13, I2 =0% (low quality) **AMSTAR 2 Score, interpretation: High-Zero or one non-critical weakness: The systematic review provides an accurate and question of interest; Moderate-More than one non-critical weakness*: The systematic review has more than one weakness, available studies that were included in the review.; Low -One critical flaw with or without non-critical weaknesses: The revie summary of the available studies that address the question of interest; Critically low -More than one critical flaw with or wit should not be relied on to provide an accurate and comprehensive summary of the available studies.