REFLEX, a social-cognitive group treatment to improve insight in schizophrenia: study protocol of a multi-center RCT

The percentage of persons with schizophrenia who have only limited insight into their illness is large, ranging from 50-80% [1]. Insight is considered a combination of a number of dimensions, that can fluctuate independently of each other, including awareness of mental illness, relabeling of symptoms and awareness of need for treatment [2]. Insight in schizophrenia is usually measured with a semi-structured interview, such as the SAI-E [3], SUM-D [4], and item G12 of the PANSS-interview [5], or self-rating questionnaires, such as the Beck Cognitive Insight Scale [6], and the Psychosis Insight Scale [7].

Poor insight has a negative impact on relevant outcomes of the disorder [see for a review: [8]]. Poor treatment compliance in patients mediates this relationship, but there is also a direct association between insight and outcome [9]. Limited insight has been associated with more positive and negative symptoms [10], more relapse and rehospitalizations [9], lower GAF-scores [11], and better observer quality of life and social functioning [9]. However, good insight may also have unfavorable consequences. Several studies have shown better insight to be associated with more depressive symptoms [8]. The exact nature of this relationship remains unclear [8, 10]. The relationship between depression and insight is thought to be mediated by internalized stigma: insight is only associated with depression in patients who hold stigmatizing beliefs about mental illness [12, 13].

Given the negative impact of limited insight on the outcome of schizophrenia, insight is a logical target for treatment. However, treatment options to enhance insight are limited. Psycho-education does not necessarily lead to better insight [14], neither does psycho-dynamic psychotherapy [15]. Turkington et al. [16] developed a treatment program that combines psycho-education on medication with cognitive behavioral therapy. Treatment adherence improved and patients were better able to label their symptoms as psychotic both immediately and one year after treatment. In others studies, no clear effects of cognitive behavioural therapy on insight was found [17].

Kemp et al. [18] demonstrated that therapy adherence and insight in symptoms improved after a brief intervention based on the principles of motivational interviewing. Others studied the same intervention, with inconsistent results [19, 20]. Two smaller studies showed that when patients are confronted with video images of themselves during a psychotic episode, their insight improves [21, 22]. In sum, although there are several interventions aiming to enhance insight in schizophrenia, there is still a need for improvement.

Three types of models have been put forward to explain this variance: the clinical model, the neuropsychological model, and the psychological denial model [23]. The clinical model suggests that poor insight is a primary symptom of schizophrenia, analogous to delusions and hallucinations. The neuropsychological model argues that specific cognitive impairments are responsible for poor insight in schizophrenia [24, 25]. Finally, the psychological denial model explains poor insight as the outcome of a coping strategy that is used to reduce the distress associated with a diagnosis of schizophrenia [26].There is limited support for the clinical model, partly because of the lack of testable hypotheses. Literature does provide evidence for the neuropsychological model and some preliminary support for the psychological denial model [23], but none of these models alone can account for the variance in insight.

Recent evidence [27] suggests that one aspect of cognitive functioning may have been overlooked in insight literature: social cognition. Social cognition refers to "the mental operations underlying social interactions, like the ability and capacity to perceive the intentions and dispositions of others" [28]. In particular, the ability to take perspective has been linked to insight[[29]; Pijnenborg, Spikman, Jeronimus and Aleman: Insight in schizophrenia: the role of affective perspective taking and empathy, submitted]. In other words: the ability to infer mental states was associated with the tendency to take another person's perspective on oneself.

Based on these findings, we propose a model that integrates elements from previous models and combines them with recent findings on the role of social cognition in insight [30]. According to this model (see Figure 1) self-reflection moderates the relationship between the prerequisites for insight on the one hand and insight on the other. Self-reflection is considered a meta-cognitive process that concerns the ability to reflect upon thoughts and feelings [31]. Self-reflection is thought to be impaired in schizophrenia; patients demonstrate difficulties in generating personal narratives that link the past with the present [32, 33]. The model explains why schizophrenia patients with poor insight erroneously hold on to their pre-morbid self-image. Because these patients do not adjust their self-images to changing circumstances, they implicitly assume that functioning and future perspective are still the same as before their illness started. In other words: they make too few self-corrections. A number of processes are thought to hamper self-reflection in schizophrenia. First of all, poor insight is associated with a lack of mental flexibility [25].We propose that this relationship is mediated by self-reflection. A lack of mental flexibility will hamper the capacity to consider alternatives and make complex inferences about oneself, which will inevitably result in poor insight. Second, recent evidence shows that insight in schizophrenia is associated with Theory of Mind (ToM) and in particular the ability to take the perspective of others [34]. ToM refers to the ability to interpret mental states of others, or the notion that mental representations of the world do not necessarily reflect reality, and can be different from one's own [35]. According to David [36], insight requires a capacity for self-reflection and the ability to make self-evaluations. David quotes 18^th century Scottish poet Robert Burns to illustrate that the ability to 'see oursels as others see us' helps people in making these evaluations about themselves. 'Seeing yourself through the eyes of others' is a process that overlaps with ToM, and in particular with the ability to take the perspective of another person to evaluate your own mental state. Indeed, schizophrenia patients are found to recognize symptoms of mental illness in others, but not in themselves [37]. However, a direct link between insight and perspective-taking is thought to be unlikely, as perspective-taking is not primarily intended for self-evaluation [29]. In line with our model, Langdon and Ward suggest self-reflection as a mediator in this relationship. Indeed, Dimaggio et al. [38] described an association between self-reflection and ToM. The last precondition of insight in our model is stigma-sensitivity. Schizophrenia is associated with a heavy stigma. There is evidence that some patients cope with the threat that stigma poses to their self-esteem by denying the illness [39, 40].

Psychological defensiveness in psychosis is associated with unawareness of having a mental disorder, unawareness of the effects of antipsychotic medication and inability to attribute symptoms to a mental disorder [26]. Cooke et al. [23] also reported an association between better insight and lower self-esteem (but not depression) and implied the influence of a psychological mechanism that preserves self-esteem. In addition, unawareness of having a mental disorder is associated with more denial of common personal failings [41]. Apparently, some patients are reluctant, at an unconscious level, to reflect upon themselves in the light of a severe mental illness. In line with our model, patients with lesser abilities for self-reflection and patients who endorse stereotypes about mental illness tell more impoverished stories about themselves [12]. The model is in line with Beck's [6] concept of cognitive insight. Cognitive insight is seen as a prerequisite of insight and encompasses the capacity of patients with psychosis to distance themselves from their psychotic experiences, reflect on them, and respond to corrective feedback. This concept clearly overlaps with the concepts self-reflection, non-social cognition and perspective-taking in our model.

We used this model to develop an intervention to improve insight in schizophrenia. This group-based intervention, from now on referred to as REFLEX, consist of three modules of four sessions each. The central theme of the first module is dealing with stigma. The second and third module aim to stimulate self-reflection through structured exercises. These exercises facilitate mental flexibility and perspective-taking. In the second module, patients reflect upon differences between their past and present circumstances and attributes. In the third module, patients are required to reflect upon their thoughts and feelings in the present. The present paper presents the design of a randomized controlled multicenter trial aiming to evaluate the efficacy of REFLEX.

The study is funded in part by a European Young Investigator (EURYI) Award from the European Science Foundation to AA. Other contributions (in terms of personnel involved) are from the mental health centers involved. The research has been approved by the medical ethical board of University Medical Center Groningen, Groningen (number: NL2714604209; date: 13-10-2009), and is conducted in accordance with the principles of the Declaration of Helsinki.

REFLEX may have a potential for improving insight in patients with schizophrenia. Our design offers the opportunity not only to examine the results of REFLEX at a behavioral level, but takes underlying changes in brain activation into account as well. With the study, we hope to contribute to the existing knowledge of what mechanisms are underlying changes in insight in schizophrenia. Improving insight in schizophrenia is a challenging task that needs careful consideration. During the development of the intervention, some clinicians addressed that patients might become more depressed if their insight would improve. Although the evidence for the development of depression is not conclusive [23], we paid special attention to this issue. As was explained in the introduction of this paper, there is evidence that this risk concerns patients with internalized stigma. By including a module that aims to reduce internalized stigma, we feel we have minimized this potential risk. Second, by definition, patients with impaired insight often do not feel they need treatment and will not spontaneously enroll in a therapy trial to improve their insight. Therefore, a common language needed to be developed. We cannot simply give patients a phone call, tell them that they are mentally ill and not fully aware of this and expect them to participate in our study. Instead, patients will be explained that being under treatment in a mental health institution brings about a lot of changes in their daily lives. REFLEX may help them to recognize these changes and gain more control over their lives. In a previous and unpublished pilot study we found that explaining the aim of REFLEX in comparable phrasing was acceptable to most patients and made them consider participation. Third, a methodological problem that is associated with insight is that most assessment instruments are based upon the traditional medical model: patients have to use the same terminology as their psychiatrists to be considered insightful. This approach ignores the insight some patients demonstrate, when they are able to describe the problems they experience in daily life and to attribute these problems to a mental illness. If they do not use the term "psychosis" or "schizophrenia" to describe their mental health, their insight is rated as impaired by most of the current instruments. However, this may be just a discussion about labels and not about actual insight. We hope to have solved this problem by including scales that measure the preconditions of insight, such as the BCIS that measures cognitive insight, to traditional insights scales. The BCIS does not measure agreement insight in and medical way, but takes into account how patients perceive their own thoughts and feelings. Finally, care as usual for people with schizophrenia in the Netherlands is extensive. Psycho-education, Liberman training modules (social skills), and CBT are accessible for most patients. Because of this extensive care, it is very hard to obtain treatment effects over and above the effect of care that is already provided. However, several studies have shown that specific interventions can make a significant contribution to relevant outcome measures [55, 56]. Through the unique focus in REFLEX on the improvement of self-reflectivity and perspective taking abilities and by that improving insight we expect to make a significant contribution to the well-being of people with schizophrenia.