A Clinical Staging Model for Early Intervention Youth Mental Health Services

Mental disorders most often have their onset in youth and represent a significant portion of the disease burden of young people (1). These findings logically lead to calls for safe interventions to be delivered as early as possible in the course of mental ill health so as to avoid the direct and indirect damage to young lives caused by mental disorders (2,3) and to avoid illness progression and improve functional outcomes (4,5). In response to this challenge, a developing network of early intervention primary-level mental health services for individuals aged 12 to 25 has recently been established across Australia—called “headspace” and funded by the federal government through the National Youth Mental Health Foundation (6). The program is just one example of international innovations in youth mental health care (7,8).

A dual challenge for any early intervention mental health service is to develop the capacity to respond to very large numbers of young people with subsyndromal disorders (9,10) while also deciding which interventions to implement in the relative absence of evidence for approaches that are effective in this group (10). For this cohort of young people, the current psychiatric classification system (11) remains limited in regard to early and nonspecific forms of mental disorders (12–14). Interventions should be guided by diagnosis, yet current psychiatric diagnostic systems are limited in their application to young people with evolving, unclear, or mixed subsyndromal presentations, which results in many young people receiving no care at worst or suboptimal care at best and which leaves many clinicians in primary care ambivalent about the utility of diagnosis (14,15). In our experience at the early intervention centers that we manage, young people often present for care with significant functional impairment and distress, even though their symptoms are subsyndromal (16,17). Further, various studies have shown that young people with subsyndromal depression (18), bipolar disorder (19), and psychosis (20) are at much higher risk of developing full-blown disorders in young adulthood. Rates of transition from subthreshold to full-blown disorders vary between syndromes and studies. Fergusson and colleagues (18) found that 27.4% of study participants transitioned from subsyndromal depression to major depression with one to two years and 33.3% within three to seven years. Axelson and colleagues (19) demonstrated that 45% of study participants converted from subthreshold bipolar disorder to either bipolar I or II within about one year. Despite the diagnostic ambiguity, young people with subthreshold disorders demonstrate a clear need for intervention. At present, however, there is a real tension between this clear need for intervention and a lack of knowledge in regard to the safest and most effective treatments and service structures.

The clinical staging framework may assist in developing this knowledge base. Clinical staging is widely used in general medicine and is defined as a more refined form of diagnosis, which serves as an adjunct to the traditional psychiatric diagnostic system (21). It seeks to place a person on a continuum of illness course so as to enable clinicians to select more effective and less harmful treatments. Its ultimate aim is to reduce the risk of progression to more severe forms of illness, or better still to bring about full remission (21). The clinical staging framework also proposes that the emergence of earlier stages of mental ill health represents a modifiable risk factor to progression to later and more serious stages of a mental disorder, thereby changing the focus of the clinician from treating the current episode to being mindful of the longer-term trajectory of illness.

Recently, various staging models have been proposed for single DSM-IV disorders (22). However, such approaches may not fully account for a critical aspect of clinical staging—that is, earlier stages of illness are often characterized by less severe and less specific clinical phenotypes. Common symptoms of anxiety and depression typically dominate these presentations, often at levels of symptomatology or impairment that do not reach thresholds for the assignment of specific mental disorders. The single-disorder approach, therefore, risks excluding many earlier presentations of illness that fail to meet criteria for a disorder or those that are characterized by very mixed symptomatology. Therefore, the utility of such models for health services planning is likely to be extremely limited. In addition, first-line (typically benign) interventions for early-stage illnesses would not differ greatly enough to warrant the complexity of a single-disorder staging approach, again severely restricting utility in real-world clinical settings.

Hickie and colleagues (13) have recently developed an expanded staging model with detailed consensus criteria to incorporate a range of syndromes within a single model. The model comprises six stages: stage 0, asymptomatic individuals at risk of a disorder who have not yet presented for care; stage 1a, help-seeking individuals with mild symptoms and mild functional impacts; stage 1b, those with attenuated syndromes, often with mixed or ambiguous symptomatology and moderate or severe functional impacts; stage 2, those with discrete disorders—that is, clear episodes of psychotic, manic, or severe depressive disorders; stage 3, those with a recurrent or persistent disorder; and stage 4, those with severe, persistent, and unremitting illness (13). Interrater reliability of this model has been shown to be good, with 90% concordance between experienced raters (κ=.72) (13).

Hickie and colleagues (12) built upon this staging framework and added another dimension associated with determining the type of pathophysiological pathway (for example, nonspecific anxiety, circadian anxiety, developmental anxiety, or a combinations of types that progresses, respectively, to discrete anxious-depression, bipolar spectrum disorder, psychotic disorder, or combinations). It is proposed that these pathways overlap substantially in early stages and may become more independent at later stages, which may lead to more specific and targeted interventions. Interventions can be somewhat stratified on the basis of what is seen to be minimally required at each stage, while also being tailored to the individual depending on what is known about a person’s biological, social, and psychological risk and protective factors (3) that may influence progression.

The number of clients assigned to each stage in our early intervention clinic settings has varied across our studies (16,17,23), perhaps because of factors such as service location (inner city versus outer suburban) and differences in the range of services offered. At the outer-metropolitan, suburban Sydney site (headspace Campbelltown), which serves a population of more than 30,000 in the 12–25 age group and assists approximately 1,500 young people per year, at intake between 33% and 41% of young people were assigned to stage 1a, 38%−40% were assigned to stage 1b, 11%–14% were assigned to stage 2, and 7%−8% were assigned to stages 3 and above (16,23). Notably, a marked increase in symptom severity was observed between stage 1a and stages 1b and above, highlighting the clinical significance of better identifying this high-risk group (13).

The follow-up data in the study by Hickie and colleagues also showed that despite receiving standard care in a multidisciplinary early intervention environment, a significant number of young people transitioned from lower to higher stages over time: 11% of stage 1a clients transitioned to 1b, 19% of stage 1b clients transitioned to stage 2 or greater, and 33% of clients progressed from stage 2 to a later stage (13). For young people in stage 1b who transitioned to a higher stage, 75% of all transitions occurred within 12 months of initial contact with the service. The observed key clinical features associated with transition from stage 1b to stage 2 were depressive disorders to bipolar-type disorders; attenuated psychotic syndromes to first-episode psychosis; and severe depression to the development of additional psychotic or hypomanic features, severe psychomotor change, or other significant psychological features, such as body image disturbance and abnormal eating behavior (13). The number of clients who achieved remission was not reported. “Transition down” rates were also not reported, because, as in all staging models, transition is unidirectional in that an individual may attain symptomatic or functional recovery but remains at the highest stage ever reached.

Currently, the evidence base is incomplete for models of care in youth mental health services (10). In recognition of this and to test the model more thoroughly, we propose a service model with a new framework that uses clinical staging and is based on preliminary data obtained from these and other studies. Despite a relative lack of evidence regarding youth mental health care models, a number of integrated components have been identified and recommended (24) and have been incorporated into the overall model as outlined below.