Abstract
In the current debate about ‘prodromal diagnosis’, ‘risk syndrome for psychosis’ and ‘attenuated psychosis syndrome’ there are two questions that need to be clearly articulated: is it justified to treat people who may be at risk of schizophrenia? And should these at-risk individuals be given a formal diagnosis?
Is it justified to treat people who may be at risk of schizophrenia?
There are two aspects to this question that require further examination. First, how we conceptualise risk for schizophrenia and second, what is meant by treatment.
Risk
It is now largely recognized that risk for schizophrenia occurs on a continuum, with certain factors such as family history of the illness [1], pre and perinatal brain insults [2,3], delayed developmental milestones [4,5] and abuse during childhood [6] placing the individual at increased risk. However, all these factors have low specificity. That is, although they are associated with increased rates of schizophrenia compared to the general population, the majority of people who have these factors do not develop the disorder. Additionally, they are all distal to the onset of illness. That is, they precede the characteristic onset phase of schizophrenia, the first episode of psychosis, by many years if not decades.
More proximal to the first episode of psychosis are psychiatric symptoms and alterations in behaviour. This prodromal phase occurs in the majority of people who develop schizophrenia and is generally not longer than a few years, with an average duration of about one year [7]. Thus it has been seen as a legitimate target for detection and intervention, with the aim of prevention of onset, or at least amelioration or delay of onset [8]. This is the background to the ‘prodromal intervention’ field.
However, even apparent prodromal symptoms lack specificity. Characteristic prodromal features include sleep disturbance, irritability, lowered mood, anxiety and decline in school or work performance [7,9]. This clinical picture could be the result of a number of conditions, such as major depression, substance abuse and physical illness, as well as a psychotic prodrome. Even attenuated or isolated psychotic symptoms may not necessarily progress to a frank psychotic disorder, as these are now known to be quite common in the general population [10–14].
There are two implications of this lack of specificity. First, is the recognition that a prodrome cannot be diagnosed as such prospectively, and that the terms ‘prodrome’ and ‘prodromal’ should not be used to refer to this clinical picture or those apparently presenting with it. This is more than pure semantics. Adoption of the term ‘prodromal’ to refer to individuals implies that they are on the path to schizophrenia. It suggests that the ‘process’ has already begun. Instead, as McGorry [15] has pointed out, terminology such as ‘at risk mental state’ should be used, to highlight that a person presenting with apparent prodromal symptoms may be at increased risk of onset of psychotic disorder within a brief time frame (1–2 years), but that onset is not inevitable.
The second implication is that there needs to be some narrowing of the definition of ‘at risk’ so that those so identified truly have a high likelihood of onset of first episode psychosis within a brief period of time. One strategy to achieve this aim has been the development of the ultra high risk (UHR) criteria. These have been described elsewhere [8,16,17]. Essentially, there are three groups identified: (i) attenuated psychotic symptoms group (APS): those who have experienced subthreshold, positive psychotic symptoms during the past year, (ii) brief limited intermittent psychotic symptoms group (BLIPS): have experienced a brief episode of frank psychotic symptoms that has spontaneously resolved; and (iii) trait and state risk factor group: have a first degree relative with a psychotic disorder combined with a significant decrease in functioning during the previous year. The UHR criteria have been tested in a number of settings worldwide and have been found to be associated with an increased risk of psychosis estimated to be between 200 and 400 times that of the general population [18,19]. Initial rates of onset of psychotic disorder were as high as 40% within one year [8]. More recently rates have reduced [20], with recent reports of rates of between 14 and 18% within 12 months [21–24] but higher over the longer term [25].
Treatment
It should be recognized that these criteria provide a method of identifying those at risk of psychotic disorder, but clearly an assessment of an individual patient requires more than just determining whether the criteria are met. People rarely present to clinical services concerned about the risk of developing schizophrenia. More often they seek help for a range of psychiatric symptoms, difficulties with relationships and role functioning and other problems with living. Thus there is a dual approach to treating UHR individuals. This involves both monitoring for signs of deterioration and evidence of onset of psychotic disorder, and managing current needs [26–28]. Far from setting ‘to one side’ ‘non-predictive symptoms’, as Rosenman and Anderson [29] suggest, these presenting problems outside the UHR criteria are usually the focus of clinical attention. Indeed if we were to ignore the problems for which UHR patients want help they would soon cease attending the service.
In terms of specific treatment that attempts to prevent, delay or ameliorate onset of psychotic disorder, five randomized control trials (RCTs) have now been conducted in the UHR population. Low dose antipsychotics [30–32], cognitive therapy [31–33] and fish oil [34] have all been found to be effective at reducing symptoms, improving functioning and decreasing risk of development of psychotic disorder, at least in the short term. An important finding has been that antipsychotics are no more effective than more benign treatments, including supportive therapy [32]. More trials are needed to determine optimal interventions. Fish oil in particular is a promising therapy that is likely to have a wide therapeutic effect with few if any side effects. This study needs replication, and ongoing research is needed into other non-pharmacological interventions such as supportive therapy and problem solving. These types of interventions have the advantage that they can be applied to manage current difficulties that people present with as well as potentially having a preventative effect.
Should at risk individuals be given a formal diagnosis?
Applying a formal Diagnostic and Statistical Manual of Mental Disorders (DSM) diagnosis to UHR individuals is a separate issue from whether they should be treated. Rosenman and Anderson [29] have unfortunately confused the two. The merits and otherwise of creating a new category in the forthcoming edition of the DSM (DSM-V) have recently been debated in the literature [35–46].
To put the significance of the proposal for a new diagnosis into context, it is necessary to understand the importance of the DSM. The DSM is the most widely used diagnostic classification system in the USA. It heavily influences the way in which health services are delivered in the USA, where a diagnosis is a necessary, but not sufficient, condition of health insurance reimbursement [38]. The introduction of the new diagnosis of ‘risk syndrome for psychosis’ or ‘attenuated psychosis syndrome’ has the potential to affect hundreds of thousands of people potentially diagnosed with this new disorder, as well as health policy, funding and services.
Advocates of inclusion of this new diagnosis in DSM-V argue that clearly defining the syndrome will highlight the potential for early intervention and will encourage attention and resources to be directed to an important clinical population [38,47]. Another driving force is that of access to reimbursement of health care costs through insurance companies in the US health care system [48]. This has led to the perception that a diagnosis is necessary for treatment. For example, Carpenter states in relation to the ‘attenuated psychosis syndrome’, that ‘without a diagnosis, the consequence is that they cannot get help’ [47, p.461]. However, in psychiatry, treatment commonly commences before diagnosis. Treatment really begins on first encountering a patient. Even the assessment process can be therapeutic as patients realize that they are being listened to, that others have had similar problems and that there may be solutions. The strength of this therapeutic alliance is associated with improved outcomes across a range of disorders and therapies [49,50]. Although a provisional diagnosis is important and hypothesis testing needs to occur, the toleration of ambiguity, phenomenological exploration, ongoing evaluation and treatment of current needs are all essential features of psychiatry. Initial impressions may need to be modified as more information becomes available or other clinical features emerge. This is often the case in general medicine as well. Rheumatology patients sometimes need to be treated with anti-inflammatory medication for weeks or months before a definitive diagnosis of rheumatoid arthritis can be made or ruled out. Again a provisional diagnosis is made in this case, but ongoing evaluation and thinking about the patient are required.
In contrast to this broad and longitudinal approach to assessing psychiatric patients, the DSM system focuses on categorizing patients according to a predetermined set of criteria that purport to encapsulate a disorder [51]. It suggests that criteria and diagnoses are immutable [52]. Thus the requirement for a DSM diagnosis in order to gain insurance coverage and initiate treatment could lead to premature closure of thinking. The danger of reifying ‘attenuated psychosis syndrome’ in the DSM is that it will be seen as a final diagnosis, with a consequent lack of re-evaluation of a patient's problems and it may encourage an unhelpful algorithmic approach to treatment. Such an algorithmic approach neglects a more rich understanding of an individual and his or her social environment [53,54]. A checklist diagnosis may actually hinder our ability to understand what is happening for that person.
Furthermore, the vast majority of people meeting UHR criteria have other comorbid disorders [55,56] that could be diagnosed in order for insurance access [38]. So even in the USA, symptomatic help-seeking people meeting UHR criteria would not be denied psychiatric care.
However, there are likely to be several unintended negative consequences of creating a new diagnosis of ‘attenuated psychosis syndrome’. For example there is the danger that risk will be seen as disorder [40]. That is, people meeting the UHR criteria will be conceptualized as being on the schizophrenia spectrum and that a lifelong underlying ‘process’ has started. While at least some people meeting UHR criteria will have symptoms that are on a continuum with those seen in full blown disorder, translating this into a formal DSM diagnosis may cause problems clinically. This may lead to them being treated as if they have a form of schizophrenia, including exposure to antipsychotic treatment, despite this not being recommended in clinical guidelines for treatment of UHR individuals [57,58]. Interestingly, such problems with the schizophrenia spectrum concept were articulated decades ago [59].
Other consequences of a diagnosis of mental illness include labelling, stigma and discrimination. Stigma involves a negative perception of the person, and can come from both others and the self. A damaged view of the self may be internalized, with consequent loss of confidence and hope [42,60]. Discrimination is related to labelling and stigma. Being labelled with a psychiatric diagnosis can potentially lead to exclusion from a range of activities, social roles and expectations [61]. Discrimination from health insurance companies may also occur as they may be reluctant to cover a person who has risk of a potentially costly disorder [62,63].
A final issue is that of ‘diagnostic creep’, in which the threshold for a diagnosis gradually changes in response to clinical practice, political lobbying and other social forces [48]. It is a particular problem when the boundaries of the syndrome or disorder are ill defined or open to interpretation. With the attenuated psychosis syndrome diagnosis a person whose symptoms lie just below the threshold for diagnosis may be labelled as having the disorder in order to access treatment and gain insurance coverage. Thus a progressively larger pool of people may be treated as having this diagnosis and gradually the conceptualization of the disorder could expand. This could result in lobbying to officially change the criteria, and a larger pool of people, many of whom are not at risk, will be diagnosed.
Conclusions
The UHR concept remains a useful one. It identifies people with a significantly higher likelihood of developing a psychotic disorder (mostly schizophrenia) within a few years. From a research perspective it identifies a group worthy of study to increase our understanding of psychotic-like symptoms and their trajectories and the emerging phase of psychotic disorders. The UHR concept is consistent with the continuum view of psychosis [13] and is probably a reflection of biological reality. This expanded way of thinking about psychotic disorders may help us to understand pathways to schizophrenia and risk and resilience factors. However, creating a new diagnosis in the DSM could be associated with problems. Many people not truly at risk of psychotic disorder will be diagnosed as having this new entity. The potential costs of stigma and discrimination must be considered. More research is needed to refine risk factors and continue to test interventions. Until we have identified a group with much higher short term risk and developed effective and tolerable interventions, the costs of this new diagnosis outweigh the benefits. This should not preclude those help-seeking and symptomatic UHR individuals from getting the help they need.
Acknowledgements
I would like to thank my father, Dr Allen P Yung, for his advice about diagnosis and treatment in general medicine.