Increasing the Role of Phenomenology in Psychiatric Diagnosis–The Clinical Staging Approach

Abstract

Recent editions of diagnostic manuals in psychiatry have focused on providing quick and efficient operationalized criteria. Notwithstanding the genuine value of these classifications, many psychiatrists have argued that the operationalization approach does not sufficiently accommodate the rich and complex domain of patients’ experiences that is crucial for clinical reasoning in psychiatry. How can we increase the role of phenomenology in the process of diagnostic reasoning in psychiatry? I argue that this could be done by adopting a clinical staging approach in diagnostic reasoning in psychiatry. The approach has the resources to include the progressive nature of patients’ experiences to a much greater degree than is currently practiced. It can address the recent plea for increasing the role of phenomenology in psychiatric diagnosis by offering a model for clinical reasoning that goes beyond the operationalized, static criteria of diagnostic manuals, without depriving us of their benefits.

I. INTRODUCTION

The complex nature of mental disorders makes the construction of diagnostic criteria in psychiatry particularly difficult. Recent editions of diagnostic manuals (e.g., APA, 2013; WHO, 2013) have focused on providing operationalized lists of criteria required for clinical diagnoses in order to improve their validity. Notwithstanding the many theoretical and practical improvements of this operationalization approach, several psychiatrists have raised worries about the limited role of phenomenology, that is, a careful description of patients’ experiences, in the diagnostic manuals and, consequently, in psychiatric diagnosis, treatment, and research (Parnas and Zahavi, 2002; Ratcliffe, 2015; Kendler, 2016). This recent plea for increasing the role of phenomenology in psychiatry is grounded in the idea that good clinical care should aim at exploring and understanding patients’ experiences (Parnas and Zahavi, 2002; Andreasen, 2006). Phenomenology is a rich reservoir of information, potentially indispensable for the successful diagnosing and understanding of psychiatric disorders. Although current diagnostic manuals, such as DSM or ICD, are to some extent based on and do include descriptions of patients’ experiences, several features of the operationalized approach have been argued to cause the reification of criteria in psychiatric practice and to impose important limitations on the role of phenomenology in psychiatry (Kendler, 2016; Parnas and Zahavi, 2002).

The above observations reveal an interesting tension. On the one hand, we expect psychiatrists to follow easy and effective procedures to reach a clinical diagnosis in virtue of operationalized criteria. On the other hand, the success of psychiatric diagnosis and treatment crucially depends on whether professionals can capture and draw on the rich and complex domain of patients’ experiences. It is therefore an important question whether and to what extent phenomenology can be included in the process of clinical reasoning in psychiatry. In this paper, I address this question by suggesting one model that could incorporate patterns of experience into clinical reasoning to a greater extent than by merely applying criteria from current diagnostic manuals and without revising them.

It has been argued that psychiatric symptoms often arise from the intensification of subjective experiences or behavior that have been present for some time (Eaton et al., 1995). When such changes become prominent, they can be distinguished as subclinical symptoms. I propose that one way to increase the role and progressive nature of patients’ experiences in diagnostic reasoning is to adopt a clinical staging approach (McGorry and van Os, 2013). Clinical staging is a heuristic strategy in clinical reasoning that goes beyond static diagnostic boundaries and allows for a detailed description of a variety of specific subclinical and clinical symptoms, including fluctuations in patients’ subjective experiences (McGorry et al., 2006). It is due to this feature, I argue, that the approach is well-suited to address the plea for phenomenology, that is, increasing the role of patients’ experiences in the process of clinical reasoning in psychiatry. By providing a new framework for such reasoning, the model allows practitioners to capture the dynamic nature of symptoms and to a greater extent incorporate patterns of experiences that at an early stage may have prodromal importance and later indicate the dynamic progression of disorder. Diagnostic reasoning, I argue, should not be identified with diagnostic criteria. The former is a rich context-dependent process that draws on a variety of information. The latter provides a standardized procedure of ascribing a particular diagnosis to a person. Consequently, on my account, addressing the plea for phenomenology does not require revision of diagnostic criteria as such. Rather, it could be addressed by adopting the clinical staging model in diagnostic reasoning in psychiatry. The implementation of this idea would require expanding training programs for clinicians and developing bridging techniques between the static criteria of diagnostic manuals and the clinical staging model. To this end, research within phenomenological psychiatry and psychology can prove to be of significant help.

The paper is structured as follows: in section II, I introduce the main principles behind diagnostic manuals for psychiatric disorders, focusing on the DSM-III and DSM-5 and sketch the so-called operationalization approach to psychiatric diagnosis. Next, I present two versions of the plea for phenomenology and illustrate the one that will be at the focus of the paper with a recent study from Kendler (2016) (section III). In section IV, I propose that we could address the plea for phenomenology by adopting the clinical staging approach in clinical reasoning in psychiatry. In section V, I discuss three objections to this proposal and reply to them. Section VI concludes.

II. THE OPERATIONALIZATION APPROACH IN PSYCHIATRY: PROMISES AND PERILS

Psychopathology has traditionally been a core element of psychiatry. Its goal is to provide descriptions that convert the patient’s experiences (lived in a first-person perspective) or certain aspects of their expression and behavior into specific categories of symptoms (as defined in third-person terms). In this way, psychopathology provides a means for translating the patient’s experiences into shareable information utilized further in diagnosis, treatment, and research (Parnas and Zahavi, 2002). Psychopathology as a theoretical description involves phenomenology, that is, a reflection on and description of patients’ experiences and expressions. Notwithstanding the important role that phenomenology has played in understanding mental disorders and their treatment in the course of the history of psychiatry, recent decades have brought important changes into psychiatry as a domain of practice and research. The big shift from the psychoanalytic approach dominant in the United States was spurred by extensive criticism of the low reliability of psychiatric diagnoses that was revealed in the US-UK studies carried out in the 1970s (Kendell et al., 1971; Kendell, 1975; Andreasen, 2006). The problematic findings led to the adoption of what is often called the operationalization approach in psychiatry (Andreasen, 2006; Parnas, Sass, and Zahavi, 2013).

The operationalization approach was first developed by the Task Force members of the Diagnostic and Statistical Manual of Mental Disorders (DSM), who were responsible for a radical revision of the manual in 1980. The new version (DSM-III) was meant to be evidence-based, use specific operationalized diagnostic criteria instead of broad descriptions, and aim at maximal reliability. The suggested criteria were intended to be atheoretical about the etiology of mental disorders and had a multiaxial nature, that is, they incorporated medical and psychosocial elements of clinical evaluation (Andreasen, 2006). The revised manual was set to provide an official reference point for psychiatric diagnoses. The goal was also to help clinicians to communicate and to reform professional training in psychiatry that was from then on to a large extent based on the DSM. Importantly, since the criteria used in DSM-III were not validated, it was an official recommendation that they be treated only as a frame for, and not as a final word in, the ongoing process of understanding mental disorders (APA, 1980). The publication of DSM-III and its widespread use in the United States together with parallel reforms in Europe (Bech et al., 1987; Mezzich, 1988) revolutionized psychiatry. The operationalization approach has continued to be and still is the main framework for later revisions of the manual. Few would doubt the important improvements that this approach has brought into the discipline by increasing the validity of diagnoses and structuring psychiatric practice, especially when compared to the somewhat unconstrained and, therefore, problematic therapeutic practice before its introduction (Andreasen, 2006).

Despite these genuine benefits, the operationalization approach faces several criticisms. It has been argued that our understanding of mental disorders has not improved much since the 1980s, as the psychiatric diagnosis continues to be at best syndromal. The worry is that psychiatric diagnoses and treatments that are based mostly on symptoms and stay agnostic about the biological basis and causal mechanisms of diseases may lack the specificity and effectiveness that is now commonly expected in other domains of medicine (Insel, 2014).¹ Moreover, the reliability, as well as the clinical utility and validity of diagnostic criteria, is still a concern (McGorry et al., 1995). Many of the diagnostic criteria are argued to be too broad for treatment and prognosis (schizophrenia, affective disorders, such as major depressive disorder). At the same time, we observe a rapid increase of poorly validated new categories (DSM-II–182 categories; DSM-III–265; DSM-IV–287). Other criticisms concern well-established stigmatizing effects of diagnoses (Angermeyer, 2004) and the role of bureaucratic and political factors in the processes of updating classifications (Paris and Phillips, 2013).²

One influential line of criticism against the operationalization approach has been that current psychiatric classifications, such as DSM and ICD, do not successfully capture the phenomenology of mental disorders; in this way, they impoverish our understanding of them (e.g., Parnas and Zahavi, 2002). It is worth noting that the study of phenomenology and nosology was actually very important for the movement of psychiatrists who started the revision of DSM in the 1980s (the so-called Mid-Atlantics) and who explicitly objected to a certain disregard that the psychoanalytic tradition had for nosology and the role of patients’ self-reports (Andreasen, 2006). Despite this initial appreciation for the need of drawing on the patient’s experiences in psychiatry, it has been argued that the operationalization approach and the resulting diagnostic manuals severely limit the role of phenomenology in psychiatric practice (Parnas and Zahavi, 2002; Andreasen, 2006; Kendler, 2016; Ratcliffe, 2015). In the following section, I present the two main versions of this criticism and clarify different responses to which they may lead. I focus on one of the versions and, in the second part of the paper, I propose a framework for addressing it, which arguably would not deprive us of the benefits of the operationalization approach.

III. THE PLEA FOR PHENOMENOLOGY

The plea for phenomenology in current psychiatry takes several forms. Part of the diversity stems from different uses of the term “phenomenology” in these debates. There is an ordinary sense of the word that denotes experience as such and that is often used in the jargon of philosophers of mind (Bayne and Montague, 2009). A more relevant sense of the term for this debate takes phenomenology to be a description of experiences, in this context, the patient’s experiences. This is the sense in which I have used the word so far and in which many criticisms of the operationalization approach are phrased (Andreasen, 2006; Kendler, 2016). A richer notion of phenomenology presupposes that when describing experiences, one adopts a certain theoretical and/or philosophical framework.³ This richer notion of phenomenology is present in Parnas and Zahavi’s (2002) criticism of the operationalization approach, and in Matthew Ratcliffe’s (2009) phenomenological framework for depressive disorder. The minimal and the rich notion of phenomenology allow us to differentiate between the following two versions of the plea for phenomenology that are present in recent debates:

(PP1) Current models of psychiatric diagnosis, treatment, and research should increase the role of the description of patients’ experiences.

(PP2) Current models of psychiatric diagnosis, treatment, and research should increase the role of theories and notions from the phenomenological tradition when describing patients’ experiences.

Although the two versions invoke different notions of phenomenology and, if accepted, would have different consequences, they are also importantly connected. On the one hand, (PP1) seems to follow from (PP2) in that increasing the role of phenomenological philosophy in current psychiatry is likely to imply increasing the extent to which the patient’s experience is included. On the other hand, (PP1) invites (PP2), in that the inclusion of detailed description of the patient’s experience will likely require that we use a set of notions that would make such a description possible and minimally consistent, and frameworks from philosophical phenomenology are the first natural candidates to consider for that role. (PP1) seems to be more basic, in that it comprises common complaints about the exclusion of several aspects of patients’ experiences. It is also a less committing version of the plea for phenomenology in that it does not postulate applying any specific approach from philosophical phenomenology, but merely postulates greater inclusion of description of patients’ experiences in current psychiatric practice. Therefore, the main focus of this paper is to discuss (PP1). In this section, I explain the criticism involved in (PP1) and illustrate it with an example from Kendler (2016).

Although the authors of contemporary psychiatric classifications claim that the classifications are “phenomenologically descriptive” (DSM-III, 1980), the critics of the operationalization approach see them as mostly behavioral. The criticism typically involves the following two complaints (Parnas and Zahavi, 2002; Parnas, Sass, and Zahavi, 2013; Kendler, 2016). First, one of the key postulates of the operationalization approach has been that descriptions of mental or subjective phenomena should be spelled out with the use of descriptions of external, observable behavior. To this end, psychiatric classifications rely on a mixture of simple lay and technical language. The lay language is prescribed in order to describe mental phenomena in a nonjudgmental, atheoretical, and arguably reliable way. The critics have argued that this is in itself problematic, because ordinary language may not be suitable to capture the rich and detailed experience in mental disorders, and it often carries implicit assumptions about their nature (Parnas and Zahavi, 2002; Kirmayer, 2008; Peled, 2018).⁴

Second, the operationalization approach explicitly recommends that diagnostic criteria should be defined at the “lowest possible level of inference” (Andreasen, 2006). As a result, the complexity of the description is intentionally reduced. The structured interview technique based on classification manuals is a key element of current diagnostic process in psychiatry (Nordgaard et al., 2013). This, according to the critics, leads to a systematic deemphasizing of the patient’s subjective experiences in psychiatric diagnosis and treatment. In effect, practitioners have a scant framework to rely on when, for example, interviewing patients (Parnas, Sass, and Zahavi, 2013; Kendler, 2016). If psychiatrists follow merely a simplified set of criteria in their diagnostic reasoning, then their understanding of patients is likely to be dramatically reduced, which would affect both treatment and research in psychiatry. The two criticisms lead to a deeper, ontological worry about the object of psychiatric study. The critics of the operationalization approach seem to agree that the conception defended in diagnostic classifications makes the object of psychiatric study and treatment vastly oversimplified and does not provide adequate tools for psychiatric practice and research (e.g., Parnas, Sass, and Zahavi, 2013; Kendler, 2016).⁵

To show that the above criticism is based on more than theoretical reflection, Kenneth Kendler (2016) provides a useful case study that illustrates the criticism. Kendler points to two important problems. One is that diagnostic manuals leave out important information about patients’ experiences. Another is that the description of experiences they actually provide is reduced and inadequate. Regarding the former, Kendler provides evidence that eighteen symptoms experienced in major depressive disorder, as indicated by descriptions from psychiatry and psychology textbooks, are not included in diagnostic manuals such as DSM-III or DSM-5. Kendler’s sample consists of nineteen textbooks published between 1956 and 1988 in five countries that represent the post-Kraepelinian Western psychiatric tradition in descriptive psychopathology prior to the operationalization approach. He identifies eighteen symptoms that are missing from current diagnostic manuals, such as experienced changes in volition and motivation, experiences of anxiety, depression-associated depersonalization, and derealization. Regarding the latter problem, Kendler observes that the description of the core mood symptoms present in the DSM-5 is narrower than the ones that can be found in the vast majority of the literature. While DSM-5 describes the mood in major depressive disorder with adjectives such as “depressed,” “sad,” “empty,” and “hopeless,” a whole variety of other emotions and feelings that are often used to characterize it and their corresponding terms (e.g., “painful,” “miserable,” dull’, etc.) are left out. Furthermore, he observes that the cognitive content of the mood characterized in DSM-5 by feelings of worthlessness and guilt often involves several more fine-grained and subtle emotions, such as gloom and various self-accusatory and self-derogatory experiences that are not mentioned.

With this evidence, Kendler argues that, at least in some domains of psychiatry, phenomenology is lacking in diagnostic reasoning. On the basis of this case study, Kendler argues for a more general claim about how current diagnostic criteria relate to mental disorders. He acknowledges that in order to be effective, diagnostic criteria should be succinct and ideally require minimal inference. To this end, the criteria currently present in diagnostic manuals may be generally well motivated, that is, he does not suggest that his findings should lead us immediately to revise the criteria for major depressive disorder. His criticism is concerned not with the revision of diagnostic criteria but with their use. Kendler argues that although generally effective, operationalized diagnostic criteria are often mistakenly identified with psychiatric disorders per se, a problematic phenomenon he calls the reification of diagnostic criteria. The practitioners and researchers in psychiatry tend to identify what is merely an index for a mental disorder with the disorder as such. This tendency in clinical practice, teaching, and research leads to an impoverished understanding of psychiatric disorders that typically involve a variety of subtle phenomena, much of which occurs in the sphere of patients’ experiences. According to Kendler, while the criteria for psychiatric diagnosis may need to remain succinct, what should be made more inclusive is the process of reasoning in psychiatric diagnosis and treatment as carried out by practitioners. There are strong reasons, he argues, to evaluate more than the DSM criteria: “Part of the process of good clinical care is to explore the experiences of our patients. . . . This cannot be done without knowledge of the world of psychopathology outside of DSM.” (Kendler, 2016, 779). Therefore, we can read Kendler’s suggestion as an invitation to reform the process of diagnostic reasoning in psychiatric diagnosis, even if we keep the DSM criteria intact and treat them merely as a guide. This was, indeed, the initial intention of the authors of DSM-III, as mentioned above. Yet, there is a strong tendency to keep diagnostic process as close to the operationalized criteria as possible, for example, via structured interviews. Hence, we have the tension between the need for succinct operationalized criteria and the important role of the rich and complex domain of patients’ experiences.

Unfortunately, Kendler’s account does not provide any specific recommendation for how we could avoid reification and resolve the above tension. It is still an open, unaddressed question whether and how a rich description of patients’ experiences informed by our psychopathological tradition could be included in the process of diagnostic reasoning in psychiatry and how to reconcile a richer model of diagnostic reasoning with the succinct diagnostic classification criteria. In the next section, I propose one model that could address this question and provide a framework for increasing the role of phenomenology in diagnostic reasoning in psychiatry. Below, I focus on Kendler’s version of the plea for phenomenology (PP1) and stay largely neutral with respect to whether in the course of that attempt we should adopt a specific framework from philosophical phenomenology (PP2). The question of how a richer description could be accommodated in diagnostic reasoning in psychiatry can, I think, be addressed independently of whether we settle on a particular theoretical framework for its description. However, given that research in phenomenological psychiatry and psychology can be particularly useful when developing specific strategies for implementing my proposal, I briefly return to (PP2) in section 5.

IV. CLINICAL STAGING TO THE RESCUE

How can we increase the role of phenomenology in the process of diagnostic reasoning in psychiatry? I now argue that the plea for phenomenology could be addressed by adopting the clinical staging model in diagnostic reasoning in psychiatry and without revising the operationalized diagnostic criteria. The model incorporates early occurring and subtle patterns of experience (McGorry and van Os, 2013) and because of that it is well suited to provide a structure for the process of clinical reasoning that includes a richer description of patients’ phenomenology and goes beyond static diagnostic boundaries of manuals. My proposal draws on the following, important distinction between the process of diagnostic reasoning as practiced in psychiatry and diagnostic criteria as exemplified in manuals. There is more to the former than can possibly be included in the latter. The process of clinical reasoning is involved at various stages of diagnosis, treatment, and research and is dependent on factors that go well beyond criteria in diagnostic manuals, including, among others, the experience and training of the practitioner, and the healthcare context. Diagnostic reasoning in psychiatry is a type of cognitive process that involves various types of knowledge and is constrained by contextual factors. Therefore, it cannot be equated with the procedure of following structured interviews based on diagnostic manuals.

The distinction speaks directly to Kendler’s version of the plea for phenomenology that focuses on the reification of the criteria in their actual use. It is the process of clinical/diagnostic reasoning, I argue here, that can be revised with the application of the clinical staging model to address the plea for phenomenology. As I explain in section 5, we have strong reasons not to apply this framework to revise diagnostic criteria as such. In what follows, I first present the clinical staging model, then I explain how it can be used to address the plea for phenomenology and illustrate it with examples.

Independently of the plea for phenomenology, McGorry and colleagues (2006; McGorry and van Os, 2013) have argued that we should adopt a clinical staging approach in psychiatric diagnosis in order to better capture the timing and progression of illness and to improve the timing of interventions in psychiatry. Their goal is to improve the overall clinical utility and predictive validity of psychiatric diagnoses, which is at best syndromal and may hinder both treatment and research progress in our understanding of mental disorders (e.g., Insel, 2014). In this way, the approach can bring psychiatry closer to other branches of medicine and health care. McGorry and colleague’s (2006) proposal is neither concerned with nor intended to increase the role of phenomenology in current diagnostic practice. Below, I argue that it could actually be a solution to this problem.

Clinical staging is a refined form of diagnosis that tries to define the extent of progression of a disease. This is done by carefully differentiating between initial, milder phenomena of subclinical nature, and those that occur as the illness progresses and are central to its fully developed form. The aim of the clinical staging approach to diagnosis is to provide a detailed description of where a person currently lies along the continuum of the course of illness (McGorry et al., 2006). As a diagnostic model, clinical staging has been used widely in the treatment of malignancies that require early intervention and where both the quality of life and patients’ survival are at stake early on. Those include, for example, sarcoidosis, myelofibrosis, autoimmune diseases, and Hodgkin’s disease (Mader et al., 1997; Hasselbach, 1993). McGorry and colleagues argue that clinical staging may be useful in any disease or disorder that tends to progress over time or that may progress. The majority of psychiatric disorders are like that, so they can be seen as amenable to clinical staging and, relatedly, their treatment is argued to benefit from it.

The clinical staging approach is a preventively oriented framework, that is, it is assessed by its success in preventing the progression of an illness into more advanced stages or the regression to an earlier stage. In the case of psychiatry, this has often been an issue. On the one hand, underdiagnosing and undertreating are costly. Timing is important in psychiatry because it is believed that an early intervention in major mental disorders such as schizophrenia or major depressive disorder can be crucial for preventing potential deficits and modulating the severity of disorders at later stages. On the other hand, because of the stigmatization related to psychiatric illness, practitioners do not want to run the risk of overdiagnosing and overtreating. Notwithstanding these difficulties, McGorry and colleagues argue that clinical staging can be used in at least some mental disorders where we have knowledge and understanding of the prodromal phenomena and the dynamic of disease at later stages. For example, anxiety disorders often precede the onset of mood disorders, specifically major depressive disorder, so we can develop a clinical staging model for mood disorders, including major depressive disorder, with anxiety symptoms as one of the prodromal elements (2006). Extensive work on clinical staging in psychiatry has also been done in the domain of psychotic disorders. A commonly discussed staging model in early psychosis is based on duration and relapse criteria, rather than anatomic extent and impact.

In a series of papers, McGorry and colleagues (2006; 2010; 2013) synthesize our current knowledge of staging in psychiatry and outline a general framework of staging for major depressive disorders and psychotic disorders. Their model suggests which groups can be targeted with specific types of preventive interventions. Ideally, clinical staging models should involve both clinical features (i.e., symptoms and signs) and objective measures that link them to psychopathology. Although we are still lacking a well-developed clinicopathological model for both disorders, research in the neurobiological basis of schizophrenia and major depressive disorder has rapidly expanded in recent years, and there is hope that in the future we will be able to develop a more comprehensive clinicopathological account of clinical staging in psychiatry. The above research suggests that we have reasons to be optimistic about the possibility of developing clinical staging models for at least some of the main mental disorders. The optimism is moderate, since one should still doubt whether it will be possible to apply a fully individualistic approach to all mental disorders. In some cases, psychiatric illness is tightly interwoven with a variety of circumstances: historical, current interpersonal, and social ones. This may make the trajectory importantly less predictable than the trajectories of many diseases. To avoid these limitations in the psychiatric case, McGorry and colleagues focus on major mental illnesses such as mood disorders and psychotic disorders and group many symptoms and operating factors together.

Can the clinical staging approach address the recent plea for phenomenology in psychiatry? I argue that it could provide a solution to the problem, at least in its less committing form (PP1). This is because the approach provides a systematic framework for increasing the role of phenomenology in the process of diagnostic reasoning. The claim is compatible with a possible role for other complementary approaches. While current diagnostic manuals involve some description of patients’ experiences (e.g., of mood in major depressive disorder), they need to be succinct in order to fulfill the main tenets of the operationalization approach. To this end, most of the diagnostic criteria present in manuals are static in nature. The criteria average not only over types of experiences and potentially idiosyncratic groups of patients, but, importantly, they typically average over the time and progression of mental disorder as well. There are a few exceptions, however. A dynamic approach can be found in the DSM-5 criteria for neurocognitive disorders. For example, Mild Neurocognitive Disorder might be viewed as a “stage” preceding Major Neurocognitive Disorder. Perhaps Acute Stress Disorder might be viewed as a stage preceding PTSD (although this progression is not invariable).⁶ The presence of these dynamic elements, especially in the case of neurocognitive disorders, where the basis is normally biological, is consistent with the observation that the criteria for several main mental disorders, such as major depressive disorder or schizophrenia, are not dynamic. This is because they do not involve information about subliminal and prodromal symptoms that tend to precede the onset of a disorder or information about subtle changes that often signal its progression. As a result, a lot of information potentially crucial for diagnosis and treatment does not enter operationalized criteria, and in effect, is often lost in the process of diagnosing, when diagnosing is carried out by merely following the criteria.

Clinical reasoning in psychiatry is more challenging than in other domains of medicine, due to its highly subjective nature and often questioned reliability (McGorry and van Os, 2013). The clinical staging model tries to address this challenge by incorporating even early occurring and subtle patterns of experience into the reasoning process in diagnosis, treatment, and research. It has been argued that psychiatric symptoms often arise from the intensification of subjective experiences or behaviors that have been present for some time (Eaton et al., 1995). When such changes become prominent, they can be distinguished as subclinical symptoms. Some elements in these patterns may disappear with time, while others may be preserved. Various elements of symptoms predictably cohere in time and can often first be observed in adolescence (Paus, Keshavan and Giedd, 2008). For example, depressive mood disorder is often preceded by symptoms and milder forms of the anxiety disorder (Cummings et al., 2014), a fact which is of great diagnostic value for early prevention programs (e.g., Balazs et al., 2013; Schetter and Tanner, 2012). Describing various specific experiences involved in anxiety and depressive mood, and their co-occurring patterns (e.g., Ratcliffe, 2015; Starr and Davila, 2012) is critically important in the process of understanding the patient’s condition, diagnosis, and treatment. Another clinically important element of patients’ phenomenology that has received relatively little attention in manuals are various experiences of depersonalization, alienation, and derealization occurring in adolescence that have been shown to correlate with a later onset of psychotic disorders (Nelson et al., 2012). A careful description of such experiences is crucial for understanding the prodromal significance of these episodes and for early prevention programs (Sass et al., 2013).

The clinical staging approach provides a model of clinical reasoning that can allow psychiatrists to go beyond the static diagnostic criteria of current manuals by incorporating the above mentioned important information about patients’ phenomenology into the process of diagnostic reasoning. The main focus of this approach is on placing a person in the evolution of the clinical phenotype. In this model, the early stages are not classified as schizophrenia or major depressive disorder. Nevertheless, the above-described occurring symptoms receive clinical attention and care. The threshold for receiving a specific diagnosis and the need for more advanced psychiatric care has to be set according to the severity and persistence of these early subclinical symptoms. The clinical staging model allows clinicians to capture these subtle phenomena. For example, on the clinical staging model, the reported abnormal experiences in prodromal stages of major depressive disorder would not immediately lead to the diagnosis of major depressive disorder as such. They will, nevertheless, be treated as clinically important elements of the patient’s state at the current moment or later if the disorder progresses. Although experiences occurring in prodromal early stages need not be part of strict diagnostic criteria for major depressive disorder, they can provide important input into the clinical reasoning process at several stages of diagnosis and treatment. In this way, early interventions could perhaps be made without the problem of stigma and overdiagnosing.

The above-discussed examples of prodromal stages and symptoms in major depressive disorder and schizophrenia illustrate that a careful description of patients’ rich experiences is an integral part of the clinical staging approach and can benefit diagnostic reasoning and psychiatric practice in general. Such early occurring experiences may be of varied nature; sometimes, they need not point to one specific diagnostic unit. Rather, they can be treated as an indication that early support and/or intervention, as well as further monitoring, are needed. In the first case, gathering information about the co-occurring experiences of anxiety may be important in the process of reasoning about future episodes of depressive disorder or its relapse in various populations (e.g., Schetter and Tanner, 2012). In the second case, information about various experiences of derealization occurring in adolescence may be important for both the prevention and future monitoring of adolescents and for a more fine-grained description and understanding of patients that receive a diagnosis of one of the psychotic disorders (Schultze-Lutter et al., 2012; Fux et al., 2013).

The clinical staging approach to psychiatric reasoning can be applied also in common cases where practitioners and researchers confront patients with full-blown symptoms of an illness. On the one hand, information about early, prodromal stages before the onset of the illness, if available, can be very useful for building a profile of the patient’s illness, further predictions about the progression, and more fine-grained research on patients who are diagnosed at the stage when they are manifesting full-blown symptoms (McGorry and van Os, 2013). On the other hand, the clinical staging model can be particularly useful in understanding and treatment of the dynamics and progression of a full-blown illness. Rich and detailed descriptions of patients’ experiences at the stages of progressing disorder should be incorporated in the clinical staging model of clinical reasoning, because they may be indicative of the patient’s entering into a different stage of her disorder or of an upcoming relapse (Eisner et al., 2013). Phenomenology is a key element required for capturing the dynamic nature of mental disorders. Following only succinct criteria from diagnostic manuals in the process of clinical reasoning would fail to take these subtle phenomena into account The above two examples illustrate quite nicely that the clinical staging model to clinical reasoning provides a structure for including them in the process of reasoning in diagnosis, treatment, and research.

The clinical staging model of diagnostic reasoning goes beyond static criteria of current diagnostic manuals. It is due to this feature that the approach can capture the dynamic nature of symptoms and incorporate patterns of experiences that at an early stage may have prodromal importance and later indicate the dynamic progression of disorder. By applying the clinical staging approach to clinical reasoning in various domains of psychiatric practice (early intervention, treatment, research), we can hope to avoid the danger of reification and address the plea for phenomenology without modifying the diagnostic manuals as such. The proposal provides a way of resolving the tension between the operationalization approach and the plea for phenomenology, by offering a specific pragmatic suggestion for how to increase the role of phenomenology in psychiatry. In this way, it goes one step further than acknowledging the need for increasing the role of phenomenology (cf. Kendler, 2016; Parnas and Zahavi, 2002). The proposal is also consistent with the goal of reforming psychiatry within the medical model and increasing the impact and reliability of psychiatric health care.

V. CLARIFICATIONS, OBJECTIONS, AND REPLIES

To what extent could the clinical staging approach address the plea for phenomenology? In this section, I clarify several aspects of my proposal by considering and responding to three objections. First, one could argue against drawing a close connection between phenomenology and the dynamic process of staging by observing that the static nature of the currently used diagnostic model cannot per se be identified with the low inclusion of phenomenology. One could, after all, adopt a dynamic approach, while still failing to do justice to the relevant phenomenology at some or all stages. Conversely, one could adopt a phenomenologically sophisticated approach that is too static (something we often find in phenomenological psychopathology).⁷ Call this the independence objection.

Although the static nature of the current diagnostic model and the low inclusion of phenomenology in psychiatric diagnoses are in principle two independent issues, there is an important and close connection between the two. The static approach to diagnostic reasoning based solely on the criteria available in current diagnostic manuals imposes structural limitations on how much information concerning phenomenology can be included in the process of diagnostic reasoning. Adopting the dynamic approach to diagnostic reasoning in psychiatry would therefore provide a new and more inclusive framework for expanding the relevant descriptions of patients’ experiences. Importantly, there is a sense in which the two projects—of providing a dynamic staging model for psychiatry and of increasing the role of phenomenology—are tightly interwoven. The success of the former depends on arriving at the staging criteria that are both indicative and fine-grained enough to distinguish between experiences typical of stages of a certain mental disorder. This is what is required of the McGorry and van Os’s model (2013) in order to accurately describe the trajectory of a disorder. As a result, the dynamic approach not only allows for but also requires addressing the phenomenology in a more detailed way than is currently done. Developing a functional clinical staging model implies greater interest in the phenomenological research. On the other hand, a good phenomenological account of a mental disorder requires accommodating the dynamic nature of experience at various stages of progression and explaining dynamic transitions between mental phenomena.⁸

Second, one could argue in reply to my proposal that if clinical staging is the right kind of response to the diagnostic crisis, then a successful change and the inclusion of phenomenology in psychiatry requires that we actually replace current diagnostic classifications (DSM; ICD) with those based on the clinical staging models. One could argue that such replacement could be a possible and welcome change, given that some parts and features of current diagnostic manuals may already involve or allow for including elements of the staging approach. For example, it could be argued that some of the information about experiences in prodromal and progressing stages can be included in the “Unspecified” diagnostic categories in which such “stages” could perhaps be coded. Alternatively, such information could perhaps be included in the multidimensional structure of criteria for some disorders, as for example in the Alternative DSM-5 Model for Personality Disorders (e.g., Widiger, 2011; Oldham, 2015). The following two reasons seem to speak against the replacement idea. First, the above example of multidimensional framework is limited in scope and as such it cannot support the idea that similar revisions would be applicable in cases of other major mental disorders. But suppose that the implementation of clinical staging model into the manuals’ diagnostic criteria were possible, at least for some major psychiatric disorders, such as affective disorders and schizophrenia. Then we encounter the second problem for the replacement idea: the clinical staging diagnostic manuals would still have to meet some constraints on the amount of information they include. For example, information about the full spectrum of prodromal experiences in schizophrenia could not be easily fitted into the operationalized standard of current manuals that aim at easy and standardized procedures. Rather, the manuals would have to involve a simplified description of staging. Once published as strict diagnostic criteria, the clinical staging model would likely suffer from a similar reification as the one discussed above in the case of current diagnostic manuals. Thus, the argument goes, the clinical staging proposal would not be able to avoid the reification problem. Call this the back to reification objection.

One reply to this could be to argue that the reification of a richer and more sophisticated model is still better than the reification of an oversimplified model, but this is not the line I would like to take. I believe that the above worry is genuine and that it actually points to an important problem with reification and supports the above idea that we should focus on improving the quality of diagnostic reasoning, not on the revision of diagnostic criteria. As I have argued in section 4, we should not identify diagnostic criteria from manuals with the process of clinical reasoning in psychiatry. The latter is a rich, cognitive, and context-dependent process that goes well beyond even the expanded and enriched proceduralized diagnostic criteria found in manuals. On the account that I propose, we need to accept the practical limitations on diagnostic classifications in psychiatry, but nevertheless try to reform and improve clinical reasoning. The above worry about a continuing reification provides, I think, a strong reason for adopting the above strategy that addresses the plea for phenomenology outside psychiatric classifications. This move is compatible with Kendler’s suggestion as well, for he merely objects to reification, not to the basic principles that require that classifications be succinct and effective.

This is where the third objection to my proposal comes into the picture. One could argue as follows: if, as I argue, we address the plea for phenomenology by adopting the clinical staging model in clinical reasoning and keep the classifications succinct, then the staging approach will not be properly constrained and may lead to chaotic practices in psychiatry with a low degree of reliability. Placing phenomenology and staging outside procedures from diagnostic manuals would lead us back to unrestricted diagnosing and can result in both overdiagnosing and underdiagnosing. Call this the back to chaos objection.

The back to chaos objection presents a genuine concern for my proposal. My tentative reply to it is that we should think of the clinical staging model as a proposal for systematic revision of education and training in clinical reasoning that is offered to practitioners and researchers in psychiatry. A helpful example comes from the training that many psychologists receive in the course of their education. An important part of this training consists in acquiring knowledge about psychiatric classifications and diagnostic criteria. An even more important part consists in a detailed study of clinical symptoms and patients’ experiences that go beyond diagnostic manuals. This is done with a careful examination of case studies, internship training, and sometimes involves familiarization with different frameworks from philosophical phenomenology. This kind of training is certainly less constrained than knowledge of diagnostic criteria, but when cashed out within the clinical staging model, it would be constrained enough to become part of a systematic approach to clinical reasoning in psychiatry.

The structured nature of the clinical staging model makes it a particularly suitable framework for addressing the plea for phenomenology. On this proposal, expertise in clinical reasoning in psychiatry cannot be reduced to the propositional knowledge of a set of rules that manuals deliver. Instead, it involves skill and what may be labeled as procedural knowledge of both applying the criteria and of going beyond them in one’s clinical reasoning.⁹ It is likely that in many educational contexts such training is to some extent already available to psychiatrists (for what they call the phenomenological toolkit, see Carel, 2012; Carel and Kidd, 2014). What may be still missing is training in connecting one’s knowledge of diagnostic criteria from manuals with the model of clinical staging that comprises a rich phenomenology. This kind of bridging in clinical reasoning may require developing new kinds of training programs. This is where the second construal of the plea for phenomenology (PP2) becomes highly relevant. The ongoing research within the phenomenological tradition in psychiatry and psychology is the first place to look for resources and inspiration to develop such training programs. Since different phenomenological frameworks and notions may be differently suitable to contribute to our understanding of psychopathological experience, their application seems to constitute a separate theoretical task. The task is already taken up by psychiatrists and psychologists working within the phenomenological tradition.

One example comes from Matthew Ratcliffe’s (2015) account of experiences of time in depression. Ratcliffe’s analysis draws on the phenomenological tradition (e.g., Minkowski, 1970; Fuchs, 2002) in characterizing existential changes in experiencing time that often occur in depression. According to him, depression (or major depressive disorder) often involves alterations in the overall structure of temporal experience, including the experience of loss of possibilities, drive, and personal projects (2015, ch. 7). By analyzing patients’ experiences of time, psychiatrists can enrich their understanding of affective disorders such as major depressive disorder at various stages of its progression. Another example of a useful phenomenological analysis that could serve the purpose of developing strategies to combine the clinical staging model with the diagnostic criteria comes from Fuchs’ (2005) comparative account of the nature of experiences of embodiment in affective disorders, such as major depressive disorder and schizophrenia. Fuchs provides a detailed analysis of alterations in experiences of embodiment that occur in major depressive disorder and schizophrenia. In the former, the body is often experienced as a material obstacle, while in the latter, the body is often experienced as separated from the mind (Fuchs, 2005, 99–102). To this end, Fuchs’ account can provide resources for, for example, interpreting prodromal symptoms and distinguishing between the early stages of both disorders. By drawing on these and similar accounts, the bridging techniques could be developed and become part of training programs in applying the staging approach to diagnostic reasoning. This could allow phenomenology to enter back into the diagnostic reasoning process which would now be only partly constrained by classification criteria.

VI .CONCLUSIONS

One of the main criticisms of current diagnostic practice in psychiatry is that classification criteria do not represent the rich and extensive experience of patients. The lack of proper phenomenological description is claimed to result in an oversimplified conception of mental disorders and to affect treatment and research. In this paper, I have argued that the plea for phenomenology need not require revision of diagnostic criteria as such; rather, it can be addressed by adopting the clinical staging model in diagnostic reasoning in psychiatry. The model goes beyond the static criteria in diagnostic manuals such as DSM and ICD and can capture the rich and dynamic nature of mental disorders. Its implementation may require expanding training programs for clinicians and developing bridging techniques between static criteria of diagnostic manuals and the clinical staging model.