In the ever-evolving landscape of medical research and practice, the application of clinical staging models has become a pivotal paradigm, notably flourishing in fields like oncology and cardiology (1). This approach, which categorizes illnesses into distinct stages according to their progression, severity, and underlying biological alterations, has demonstrated its effectiveness in directing treatment approaches and enhancing patient outcomes (2).

Clinical staging diverges from traditional diagnostic methods as it not only delineates the extent of illness severity but also situates a patient within the course of the disorder (3). This introduces the concept of early interventions to enhance recovery in the disease’s early phases, to prevent progression to later stages, and to move towards treatment personalization (3–5). While clinical staging as a model for categorizing the development of disorders was previously overlooked in psychiatry, several theories have surfaced in recent decades for major disorders (1, 2, 6).

Indeed, traditionally, psychiatry has preferred a categorical and more static classification approach to a dimensional and more dynamic one, to define mental disorders, providing clinicians with a consensus-based framework for diagnosis and treatment, derived from the main classification tools (7, 8). Nevertheless, the inherent dynamism and clinical heterogeneity of schizophrenia (SCZ) present notable challenges to these conventional traditional models (9, 10). Hence, recognizing the shortage of static categorizations to capture the evolving nature of SCZ has prompted a shift in focus towards adopting a clinical staging approach (11, 12).

The inaugural staging model arose in 1982, and subsequent decades have witnessed the development of various clinical staging concepts for several psychiatric disorders (1, 2, 13). Despite the absence of a clinical consensus designating a gold standard among these models, integrating them reveals an overarching, unified staging concept comprising four distinct stages for SCZ (14–17).

Based on the information provided, it can be inferred that the clinical staging concept commonly employed in other medical disciplines could potentially be relevant to mental and behavioral disorders (1, 2, 4). In particular, several theoretical staging models have been formulated for conditions such as bipolar disorder, depression, and SCZ over the past decade. While there are promising data for the staging of bipolar disorder (18–21), only one of these models, specifically designed for depression, has demonstrated empirical validity (22, 23). Consequently, despite ongoing efforts, it is crucial to either validate the existing models or develop new empirical staging models that have proven validity for regular clinical application.

To address this gap, our study aimed to stage psychotic disorders using a real-world clinical sample of individuals suffering from schizophrenia spectrum disorders (SSDs), with a cross-sectional observation. The primary objective was to investigate whether empirical evidence could support a staging model based on the clinical severity of psychotic disorders.

Specifically, we sought to apply a staging model for well-established psychosis, incorporating the principles of clinical staging and drawing from existing literature on the staging of psychotic disorders to a cohort of participants affected by SSDs.

Our study aimed to apply a clinical staging for SCZ based on the current main theoretical model to pave the path to verify its empirical support in clinical practice. Our methodology aligns with practices observed in other medical domains, where staging is grounded in a robust comprehension of the natural progression of illness, refined through the utilization of valid staging subtypes, and individual risk factors to formulate subject-specific predictions regarding illness risk, progression, and overall prognosis. To the best of our knowledge, this assessment stands as the sole evidence-based study explicitly evaluating a theoretical staging model for psychotic disorders, primarily derived from their current state, functioning, and outcomes.

The two main diagnostic systems in psychiatry, viz, the International Classification of Diseases version 11 (ICD-11) (41) and the Diagnostic and Statistical Manual of Mental Disorders version 5 text revised (DSM-5-TR) (42), although currently indispensable for both educational and research purposes and clinical diagnosis and economic/insurance reasons, still exhibit considerable limitations to capture the dynamic complexity of severe psychiatric syndromes and their psychopathological dimensions. Additionally, as new classification methods rooted in neuroscience and genomics evolve, there is a notable gap in the integration of crucial concepts like clinical staging, neuroinflammation, and thorough consideration of endophenotypes within both the DSM and ICD (43).

It is essential to recognize that our model differs from existing staging models of psychosis, albeit remaining within the framework of the model designed by McGorry (3, 32) and with clinical specifications introduced by Maj and colleagues (35). In particular, previous models do not provide assessment indications to be used sequentially to quantify. Conversely, our model predominantly focuses on the extended outcomes objectively assessable by the clinician through medical history collection, clinical interview, and administration of validated tests, and it further distinguishes between stable and progressing courses within the realm of non-remitting psychoses.

Particularly interesting is the role that age at onset, age at the time of assessment, and duration of illness play in the staging of SCZ. Indeed, time could be intuitively interpreted as the main determining factor in the staging of SCZ and in the overall increasing severity of psychosis since significant evidence supports the neurodegenerative hypothesis, particularly concerning the interaction between SCZ and aging (44). On this regard, there are data that correlate SCZ, like many other neuropsychiatric disorders, with accelerated brain aging (45, 46). Therefore, it may be speculated that clinical staging studies with prospective designs correlate the advancing age of patients and the duration of the disease with increasing staging (36). This perspective, however, appears limited as it does not account for the potential for improvement and recovery that does occur. Therefore, the model applied in this study also considered a cross-sectional design, giving less emphasis to the current age and duration of illness. Instead, it focused on the current and specific situation, which can be modified by therapy and rehabilitation (47).

When comparing the characteristics between different stages, it is interesting to note a clear difference in general psychotic symptomatology, as assessed through the PANSS general, which is not equally captured by the positive and negative symptom subscales. This difference becomes even more pronounced during the transition from stage 2A to 2B [i.e., from the “Episodic course with full remission (single episode)” to the “Episodic course with partial remission (multiple episodes)”]. This is unsurprising, as the repetition of several psychotic episodes leads to a notable worsening of the patient’s clinical condition. The chronic-relapsing course of SCZ results in a progressive worsening of symptoms, individual functioning, recovery capacity, and morphological and functional brain activity after each relapse (48, 49). These data support a multi-step, progressive, and prompt staging model as proposed in this work.

Recently, there has been only one attempt to perform clinical staging trials of SCZ, using a different approach with different, albeit comparable, methods and results to our study (36). In this study, Peralta et al. assessed the empirical validity of a staging model for psychotic disorders primarily grounded in their extended course delineating six consecutive stages (2A, 2B, 3A, 3B, 4A, and 4B) based on factors like symptom recurrence, persistence, and progression, including functional decline with data from 243 participants experiencing FEP (36). The authors stated that later stages consistently exhibited significantly poorer outcomes compared to earlier stages across most validators, displaying generally medium to large effect sizes and a dose–response pattern. Actually, our study represents the first real attempt to apply a staging model, albeit modified, to a real population through a prospective analysis and starting from the FEP. Our model has the ambition of being able to be applied to any patient at any time, with a transversal design and with object data from the personal and clinical history and from the psychopathological state of the patient suffering from SSDs.

Still, according to Peralta and colleagues, baseline predictors, such as familial SCZ load, neurodevelopmental impairment, childhood adversities, treatment delay, negative symptoms, neurological impairment, and poor early treatment response, independently explained 49.9% of the variance in staging (36). This result overlaps with our findings, and in some way confirms them by providing the theoretical substrate on which the additional clinical analysis is built. In fact, following the analysis suggested by Maj and colleagues, we also evaluated salient domains to be considered in the clinical characterization of a patient with a diagnosis of primary psychosis, i.e., psychopathological components, onset and course, social functioning, family history, physical and psychiatric comorbidities, childhood trauma and adversities, and protective/risk factors (35).

Therefore, a psychosis staging model primarily based on its long-term course demonstrated robust construct, outcome, and predictive validity. On the other hand, our findings have the potential to enhance our understanding of stage indicators and predictors of clinical stages with a cross-sectional viewpoint and a dynamic application of the model through multiple potentially repeated evaluation. Therefore, the use of rigid and unchangeable classification means, as well as the use of purely categorical diagnostic systems, fails to capture the nuances and dynamism of features inherent in a multifaceted disorder such as SCZ (50).

However, SCZ is a very dynamic and heterogeneous disorder in its clinical presentation. While it is possible to identify a certain cyclicality and progressiveness between the advancement of clinical stages and the clinical and functional worsening, there are also chronic patients with a long history of illness who achieve an excellent level of global recovery (51, 52). Therefore, it appears anachronistic to consider patients more serious solely based on the longer duration of their illness, especially compared to young acute patients with a dramatic decline in post-episode functioning (53). From this perspective, a transversal clinical staging model, possibly applied multiple times over the course of illness, seems better suited to the nature of the disorder than a retrospective or prospective model that does not account for the current nature of the condition.

Finally, the concept of within-stage heterogeneity explains the benefits of defining stages in terms of clinical psychopathology and stage modifiers. Using examples from medicine, we discuss the usefulness of categorizing stage modifiers into factors related to progression (i.e., potential predictors of stage transition) and extension (i.e., factors associated with the current presentation that complicate treatment selection). Finally, we propose revising the current transdiagnostic staging approach to incorporate these key concepts, and we suggest how this revised framework could be applied in both clinical and research settings.

When interpreting the results of our study, it is relevant to evaluate some limitations and strengths. First, although the overall sample size is adequate to draw global assessments, the single clinical staging levels led to a reduced sample size in certain subgroups, consequently diminishing the statistical power to identify significant differences across stages. Second, the applied clinical and assessment evaluation, despite containing tools widely used and validated in the literature, and although having been used in the appropriate population and for the intended purposes, has not been validated as a single clinical staging evaluation tool before. However, this also represents the peculiar and novel element of this work and is based on evidence-based data and clinical hypotheses that are easily reproducible in other clinical contexts and psychiatric facilities worldwide. Third, the included sample presents a wide sociodemographic and clinical heterogeneity, consistent with real-world practice. Finally, a further limitation of the study concerns the use of a cross-sectional evaluation, which returns a profile of the patient referring to the exact moment in which this is studied. Future studies should be able to carry out this type of clinical staging assessment prospectively, also including patients’ cognitive functioning evaluation.

On the other hand, the choice to apply a cross-sectional study design together with the chance to use easily accessible data from the individual history of patients and assessment tools, which are widely used and available and well-known by mental health professionals, makes this staging protocol easily reproducible and applicable in different contexts. Moreover, this is the first attempt to apply the clinical staging of SCZ in a real-world population, using both retrospective and current information, objective and feasible validated tools, and with a design that allows even a single clinical observation to place patients in a specific clinical stage.

Usually, clinical staging models place their primary emphasis on delineating distinct stages of the disorder based on severity, progression, and symptom characteristics, aiming to enhance prognostic predictions. From a clinical perspective, the definition of discrete stages establishes a framework for the development of interventions focused on prevention. In this study, we proposed a practical application of clinical staging categorizing 137 outpatients suffering from SSDs starting from the McGorry staging model and then modify it into a clinical adaptation. This classification is intended to assist clinicians in better characterizing patients and selecting treatments tailored to each specific stage. Our results show that a pragmatic use of clinical staging models in clinical practice not only is possible, but also provides a timely assessment of the stage of severity in which the patient is at the observation moment. In the future, it would be desirable to apply these staging models in the earlier stages, where intervention is more effective in curbing the progression of the illness, and with a prospective study design. Moreover, integration with biological, neuroinflammation, and neuroimaging markers is not only desirable but also necessary to improve the sensitivity and precision of a modern staging model of SCZ. Similarly, it will be necessary to dedicate more space to the evaluation of comorbidity with physical diseases and health-related quality of life, as well as social and protective factors, for a more holistic classification potentially integrated into routine clinical practice.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The study was conducted in accordance with the principles of the Declaration of Helsinki and approved by the Ethics Committee of University Hospital Mater Domini of Catanzaro (Italy) ‘Regione Calabria, sezione Area Centro’ (n. 191/2020). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

RdF: Conceptualization, Investigation, Writing – original draft, Writing – review & editing. EAC: Writing – review & editing. MR: Writing – review & editing. MA: Writing – review & editing. CSG: Formal analysis, Methodology, Writing – review & editing. PDF: Conceptualization, Writing – review & editing.