Current status and challenges of the psychiatric nosography. Is there a conflict between expectations and evidence?

1. Introduction

Nosology is the medical discipline that deals with the systematic exploration, classification and definition of diseases. Otherwise formulated, it is a scientific approach focused on exploring the characteristics of nosographic entities and organizing them according to clearly predefined criteria, usually involving data regarding their ultimate explanations. While in different fields of medicine, this approach is not a usual subject of controversy (for example, the classification of infectious diseases according to their pathogens), when discussing psychiatric nosology, multiple epistemological conundrums have to be confronted. This approach is made especially difficult by the fact that any unique theory about the underlying nature of psychiatric conditions is far from being unanimously accepted, thus raising an essential challenge for defining the criteria according to which mental disorders may be classified⁽¹⁾. Although confronting such a formidable obstacle may seem futile in the absence of clear, incontrovertible data derived from research, the number of taxonomies suggested throughout history by diverse schools of thought in psychiatry can be considered quite high, even if very few of them pass the time test. For example, diverse authors have modeled nosological systems in psychopathology by focusing on categorizing conditions according to their presumed causes (these are considered “pathogenetic classifications”), clusters of symptoms (a “semiological or syndromal classification”), or population distribution (e.g., supported by a cultural perspective, based on epidemiological data etc.).

Since the topic of the pathogenesis of most mental illnesses is still controversial, this narrative review mainly explores the current status of psychiatric nosography, which is defined as the scientific description of specific mental health and behavioral conditions, not focusing on their origins. However, due to their historical importance, nosological models based on pathogenetic theories will also be mentioned, at least those that could be retrieved by a nonsystematic approach to the topic. This review focuses only on the strengths and limitations of the most known classification systems in psychiatry and highlights their overlapping domains, as well as their specifics, and does not approach in detail the theories on which such taxonomies are based.

From a historical perspective, the first classification of mental disorders was inspired by botanical nosology, and it was created by the French physician and botanist François Boissier de Sauvages de Lacroix in 1763⁽²⁾. This seminal work, named Nosologia methodica, was initially published in Latin, then posthumously in French, and included 10 distinct classes of diseases, divided into 2400 unique conditions⁽³⁾. Out of these, four categories of “vesaniae”, or mental diseases, were listed: (1) hallucinations (e.g., vertigo, diplopia, hypochondriasis, and somnambulism), (2) “morositates” (e.g., pica, bulimia, polydipsia, satyriasis), (3) “deliria” (e.g., “paraphrosine” – a type of temporary delirium induced by substances/medical conditions, or demonomania – a type of melancholia generated by the devil), and (4) “folies anomales” (i.e., amnesia and agrypnia – a type of insomnia)⁽²⁾. Boissier employed the observation method formulated by Thomas Sydenham and considered mental illnesses to be of the same rank as any natural elements; therefore, he used the same botanical classification, structured according to the species and genera⁽⁴⁾.

Later, Pinel published a classification of mental diseases in his Nosographie philosophique ou méthode de l’analyse appliquée à la médecine (1798) and Traité médico-philosophique sur l’aliénation mentale ou la manie (1801), where he proposed distinct categories of “mental alienation” – melancholy or exclusive delirium, mania with delirium, mania without delirium, dementia or abolition of thought, and idiotism or obliteration of the intellectual and emotional faculties⁽⁴⁾. Pinel took an etiological approach to nosography and considered a moral intervention the appropriate therapeutic method for these diseases⁽⁵⁾.

Treatise of Mental Disorders (1860), authored by B.A. Morel, continued the etiological approach started by Pinel and suggested a classification with two major categories: (1) the particular forms of insanity (alienations or hereditary forms of insanity, forms of mental insanity due to intoxication, hysterical, epileptic, hypochondriacal forms of insanity, and forms of alienation or idiopathic insanity) and (2) dementias⁽⁵⁾. Wilhelm Griesinger (1817-1868) was interested in pathophysiological research and considered that many mental illnesses originate in disorders of the brain. He is considered a pioneer of biological psychiatry, although he admitted that psychogenesis is a possible explanation for some psychiatric disorders⁽⁶⁾.  Griesinger and other German psychiatrists who followed (Karl Kahlbaum, Richard von Krafft-Ebing) added important knowledge to the foundation of three diagnoses: melancholia (“psychotic melancholia” was delineated from “simple depression”), catatonia (defined as a collection of movement disorders), and hebephrenia (psychosis with onset in adolescence)‌⁽⁶⁾. Also, they formulated diagnoses as clusters of related symptoms and structured such clinical representations more clearly than their predecessors⁽⁶⁾.

The seminal work of Emil Kraepelin (1856-1926) consisted of classifying mental disorders not strictly on biological factors but on clinical elements, thus avoiding the theoretical pitfalls some of his predecessors shared⁽⁶⁾. He grouped catatonia, hebephrenia and paranoid delirium in the category of “dementia praecox” and distinguished it from “manic-depressive insanity” (which included all types of mood disorders, regardless of polarity), while constantly revising this nosography during successive editions of his Psychiatrie. Ein Lehrbuch für Studierende und Ärtze^(5,6). “Manic-depressive insanity” had, according to Kraepelin, a fluctuating evolution without significant deterioration, while “dementia praecox” presented an early onset and progression toward severe dysfunctions and institutionalization⁽⁶⁾. Therefore, the diseases were defined not only by certain clusters of symptoms but also by their evolution and prognosis, which was a new epistemological perspective on categorizing mental illnesses. In the post-Kraepelinian era, Manfred Bleuler defined distinct trajectories of schizophrenia, enriching the nosography of this pathology⁽⁶⁾. Also important, the psychoanalytical approach to nosography and nosology conceptualized by Sigmund Freud (1856-1939) was based on a new theory about the pathogenesis of mental disorders. Sometimes compared to a revolution in psychiatry and presumably able to explain the onset and evolution of these disorders, this conceptual model created by Freud also reframed previous diagnoses while creating new ones, like anxiety neurosis and obsessional neurosis^(5,7).

The physician Jacques Bertillon (1851-1922) can be considered the founder of a completely novel nosography and nosology, because he presented a revolutionary way to construct a nomenclature of diseases based on statistical principles, starting in 1985⁽⁵⁾. He suggested that, to be valid and useful, such a taxonomy should be built on homogenous data and clearly predefined principles that could be replicated worldwide⁽⁵⁾. Bertillon participated in the first three revisions of this nomenclature, in 1900, 1910 and 1920, and after the founding of the League of Nations in 1920 in Geneve, the revisions were supervised by a health office within this institution⁽⁵⁾. The World Health Organization (WHO), founded in 1948, took on these revisions and published the sixth edition of the Manual of the International Statistical Classification of Diseases, Injuries, and Causes of Death (ICD), which included, for the first time, a section dedicated to mental disorders, further divided into psychoses, psychoneurotic personality disorders, and character and behavior disorders⁽⁵⁾.

The 7^th (1955), 8^th (1965) and 9^th (1975) editions elaborated further on the nosography of mental disorders by adding more details and increasing the complexity of the coding system, with the 10^th edition appearing in 1992⁽³⁾. The ICD-10 is used in more than 100 countries and includes 155,000 codes for all diseases versus only 17,000 codes of the ICD-9⁽³⁾. There is a multiaxial system of diagnosis in ICD-10, although less well known than its counterpart in the American Psychiatric Association’s (APA) Diagnostic and Statistical Manual of Mental Disorders (DSM). The ICD-10’s multiaxial diagnosis consists of axis I clinical syndromes = psychiatric disorders, including personality disorders and somatic diseases; axis II = disabilities, which are evaluated by WHO-DAS (Disability Assessment Schedule) 2.0; and axis III = environmental factors and lifestyle factors⁽⁸⁾.

APA published its first edition of the DSM in 1952, with 102 broadly constructed nosographic categories, which differentiated between organic brain dysfunctions and functional disorders, the last being created by environmental stressors (psychoses and psychoneuroses); the descriptions were short and focused on causes, less on clinical manifestations, and the nosography was based on a psychodynamic model^(9,10). DSM-II (1968) did not record significant changes compared to its previous version, and the psychodynamic fundament was preserved^(9,10). The DSM-III (1980) excluded etiological specifications, became more rooted in the phenomenological aspects, and increased the number of diagnostic categories to 265 versus 182 in the previous edition^(9,10). DSM-III-R (1987) also increased the number of diagnostics to 292^(9,10). DSM-IV (1994) preserved the multiaxial diagnosis format initiated by DSM-III, with axis I = mental disorders, including those induced by organic causes or substances, axis II = personality disorders and mental retardation, axis III = organic diseases, axis IV = psychosocial and environmental problems, and axis V = Global Assessment of Functioning Scale (GAF) score^(9,10). It is easy to observe the similarity of construct between the two multiaxial systems (ICD-10 and DSM-III), although there is a different assessment instrument (WHO-DAS versus GAF) and a lower number of axes in the WHO taxonomy (but the same level of information can be retrieved in either of the axial systems). The nosological framework of the DSM-IV-TR (2000) is not much different from its previous version, with both strengths and limitations preserved, as the total number of disorders remained the same (297)⁽¹¹⁾.

The exploration of nosographic systems also presents therapeutic significance since case management involves not only a list of diagnoses but also a comprehensive perspective on each patient’s specifics – i.e., resources, limitations and personal perspective on the disorder he/she is confronted with, personal lifestyle, personality features, genetic background, etc.^(12-15) Therefore, making the diagnosis is only a part of a more complex and structured investigation of the patient, which mixes psychological, psychosocial, medical, pharmacological, neurobiological, economic and other related factors. In order to respond to such practical challenges, scientists have tried to work on more complex, nosological-driven taxonomies, which can be of use in formulating a treatment plan (some examples will be presented below). However, even when atheoretical nosographic systems are used, the clinician integrates the data (i.e., the diagnosis/diagnoses in this case) into his/her own experience and resources, and formulates the case management plan by summing up an important number of extranosographic variables.

2. Current classifications of mental disorders based on phenomenological elements

Capitalizing on the experience of previous editions, both ICD-11 (2018) and DSM-5 (2013) continued to evolve, and a trend toward harmonization of the WHO and APA classifications can be observed – this tendency is still true for the latest DSM-5-TR (2022)^(1,16-18). Since both nosographic systems are atheoretical and based on statistical principles, it is not difficult to understand how the compatibility between the two could further increase. A joint effort from the DSM-5 Task Force and the ICD-10 Revision Advisory Committee was essential to develop such harmonization between the two systems⁽¹⁹⁾.

The 11^th edition of the ICD introduced many changes compared to the previous edition, starting from structural modifications by reformulating the coding system and up to a new conceptualization of some previous diagnoses or nosographic categories^(1,20). The sections dedicated to psychiatric disorders followed this perspective of reconstruction, based on international expert cooperation, and new subchapters were added⁽²⁰⁾. Currently, there are 23 subchapters in the mental disorders section, with updated definitions and a more sophisticated way to integrate multiple diagnoses (comorbidities) due to the novel coding technique^(1,20,21). Examples of newly introduced disorders in the 11^th edition of the ICD are type-II bipolar disorder, complex post-traumatic stress disorder, intermittent explosive disorder, binge eating disorder, gaming disorder, or disorders associated with the use of new psychoactive substances. Rebranded disorders are compulsive sexual behavior disorder (formerly known as “excessive sexual drive”), bodily distress disorder and body integrity dysphoria (overlapping with former “somatoform disorders”), and “autism spectrum disorders”. Other ICD-10 chapters were completely restructured – for example, the “neurotic, stress-related and somatoform disorders”, while the chapter dedicated to “obsessive-compulsive and related disorders” was newly created in the ICD-11 starting from the obsessive-compulsive disorder in ICD‑10. Catatonia may be associated, according to the latest ICD edition, with other mental disorders, eliminating the restrictions to organic disorders and schizophrenia, which were present in the previous edition.

At a general level, however, both ICD-10 and ICD-11 remain categorical systems, but an important change is the transition to a dimensional approach in interpreting the personality disorders (according to severity and the presence of one or multiple maladaptive personality traits) and to a dimensional perspective in diagnosing severity, course and specific symptoms of various mental disorders⁽²¹⁾.

Regarding the DSM-5 and DSM-5-TR, multiple changes may be observed by comparison to the previous version. The multiaxial system was eliminated, and so was the GAF scale^(16,17). The exclusion of the GAF was due to the lack of conceptual clarity and questionable psychometrics in daily routine, while the multiaxial system was dropped for less clear reasons⁽¹⁶⁾. In the latest DSM editions, the GAF was replaced by WHO-DAS 2.0, and the diagnoses are recommended to be listed successively instead of using the multiaxial format. Also, some disorders were reclassified, with six classes being added and four removed⁽²²⁾. The disorders met for the first time during infancy, childhood and adolescence are not a single chapter anymore, those nosographic entities being redistributed to other chapters. Some sections and diagnoses were re-named – for example, “substance dependence” and “substance abuse” were replaced by “substance use disorders”, dementias were transformed into “neurocognitive disorders”, mental retardation became “intellectual disability (intellectual developmental disorder)”, pervasive developmental disorders became “autism spectrum disorder”, etc. New entities were introduced, like behavioral addictions (“gambling disorder”), “excoriation (skin-picking) disorder”, “hoarding disorder”, “binge-eating disorder”, etc. Although the criteria for personality disorders did not change from the previous edition, an alternative dimensional model was developed and presented for further study.

The advantages of the DSM/ICD nosography are: (1) these systems have large clinical applicability; (2) structural simplicity (a predefined set of phenomenological criteria, usually coupled with functional impairments and a temporal criterion); (3) the categorical perspective, which allows for multiple diagnoses; (4) the atheoretical perspective, which eliminates the possibility of their rejection based on preferences for a certain pathogenetic perspective; (5) the statistical-driven orientation and epidemiological foundation.

The main limitations of the DSM/ICD nosography are: (1) the heterogeneity of the populations defined by the same diagnostic criteria; (2) one symptom may be found in various diseases, which means there is a poor construct specificity; (3) excessive reliance on subjective data reported by the patients or observed by clinicians at a certain moment; (4) the main focus is on symptoms, therefore prophylaxis cannot be conceptualized in this framework (it does not include risk factors, data on pathogenesis, etc.);  (5) ignoring the specifics of each patient’s clinical presentation and encouraging stigmatization by placing “labels” instead of formulating personalized diagnoses.

3. Psychodynamic-oriented nosographic systems

The Operationalized Psychodynamic Diagnosis (OPD)

The first edition of the OPD was authored by a group of German psychodynamically oriented researchers and clinicians (Arbeitskreis OPD), founded in 1992, who wanted to create an alternative multiaxial diagnosis system to ICD-10’s Chapter V and to the DSM-IV^(23,24). According to this OPD Working Group, the nosographic approach of ICD and DSM was insufficient for clinicians interested in planning a psychotherapeutic intervention and in monitoring the potential directions of evolution during the therapy in a certain patient⁽²⁵⁾. For example, a symptom is associated by a psychodynamic therapist with certain conflicts that have an etiological relationship with the clinical manifestations but also with dysfunctional relationships and the patient’s life history⁽²⁵⁾. Therefore, new descriptive levels have appeared in this nosographic system to make the formulation of a psychodynamic diagnosis possible.

The second edition (OPD-2) was published in 2007, and the latest edition was published in 2023^(25-27). The second edition places more emphasis on the process than on the structure, while the third edition formulates the axes more clearly and is more focused on dimensionality⁽²⁷⁾. A distinct version of OPD, the Operationilisierte Psychodynamische Diagnostik des Kindes- und Jugendalters (OPD-KJ and OPD-KJ-2), has been created to address specific problems in diagnosing children and adolescents using a psychodynamic approach⁽²⁸⁾.

According to OPD-1 and OPD-2, there are five axes: (I) Experience of illness and the prerequisites for treatment; (II) Interpersonal relationships; (III) Conflict; (IV) Structure;  (V) Psychosomatic and psychiatric disorders according to ICD-10, Chapter V (F codes)⁽²³⁾. The first four axes correspond to core psychodynamic concepts – i.e., personality structure, intrapsychic conflict(s) and transference, while the fifth axis is a descriptive-phenomenological one⁽²³⁾. The first axis has a modular structure and assesses the current severity and duration of the illness, the patient’s experience, presentation and concept of illness, and his/her resources and impediments for change; this axis allows scoring the GAF scale and the EQ-5D, thus improving the compatibility of this system with other types of structured evaluations; based on this axis, the clinician may differentiate between patients who may benefit from psychosomatic interventions and those who may be more fitted to psychotherapy programs; the same axis allows for adding supplementary modules, focused on psychotherapy or judiciary contexts⁽²³⁾. The second axis assesses the patient’s experience of him/herself and others (the investigator included), and a formulation of the relationship dynamic; grounded on the transference-countertransference analysis, this axis evaluates the repetitive patterns of dysfunctional relationships by assessing the way the patient perceives others, the way the patient perceives him/herself, the way the others (clinician included) perceive the patient, and the way the others perceive themselves in rapport to the patient⁽²³⁾. The third axis evaluates the repetitive-dysfunctional conflicts (there are seven conflicts listed; for example, individuation versus dependency, or submission versus control), and the mode of processing the main conflict; these conflicts are scored on a Likert scale, from “absent” to “very significant”; the most important two conflicts are chosen for each patient, if they were present during the last two months, and integrated in the final rapport⁽²³⁾. The fourth axis includes details on self-perception and object perception, but also on self-regulation, regulation of object relationships, internal communication, communication with others, specific areas of attachment, and the total structure; the operationalization of this axis helps the therapist formulate a plan for working with personality^(23,29).

The third edition of the OPD includes the phenomenological diagnosis (axis V in OPD-2) into axis I, and the GAF (maximum value in the last week) is distributed on three dimensions: symptoms, working activity, and social functioning; other minor redefinition of isolated criteria exist, and two distinct versions, for clinical use and for research, exist for axis II and IV⁽²⁷⁾.

The axial structure of OPD-2 has confirmed validity and reliability when administered concomitantly with SCID-Interviews for DSM-IV and other questionnaires, like Brief-Symptom-Checklist (BSI), Inventory of Personality Organisation (IPO), and Borderline Personality Inventory (BPI)⁽³⁰⁾. The interrater reliability of OPD-3 was tested, and the diagnosis of conflicts, levels of structural integration and dysfunctional relationship patterns were confirmed as presenting high reliability when used in parallel with SCID-I and SCID-II⁽³¹⁾.

The strengths of OPD are: (1) its axial structure, which allows for circumscribing the focus of the therapeutic intervention; i.e. – dysfunctional patterns of relationships, specific intrapsychological conflicts, or the structure of personality; (2) although a psychodynamic-oriented psychotherapist has a certain advantage in understanding the rationale and theoretical background of the OPD, for the practical use of this instrument, only 60 hours of training (three separate seminars) are needed, as shown by the Working Group for OPD; (3) it integrates clinical, phenomenological diagnosis/diagnoses, into a psychodynamic oriented treatment plan; (4) it also includes validated scales like GAF or EQ-5D, which allows for compatibility of research using OPD with other nosographic systems^(28-30).

OPD limitations are related to: (1) possible changes in time when assessing the importance of a certain conflict or structural vulnerabilities, which may be subject to a dynamic development; (2) also, a psychodynamic diagnosis is based on a theory of another individual’s inner world; (3) not to be ignored is the fact that psychotherapists may bring to the OPD formulation of the case, personal considerations and convictions derived from own clinical experience, focusing on oedipal conflicts, individuation, or identity crisis⁽²⁵⁾. 

The Psychodynamic Diagnosis Manual (PDM)

The first edition of this manual was created by a collaborative task force (The Alliance of Psychodynamic Organizations), which included five institutional members, with authors mostly from the American Psychoanalytic Association and International Psychoanalytical Association^(32,33). The first section of PDM is dedicated to adult mental disorders and includes three axes: P for “personality patterns and disorders”, divided into healthy, neurotic and borderline (with the notable exception of psychotic personality), and further subcategories (such as schizoid, paranoid, antisocial, narcissistic, somatizing, masochistic, etc.); M for “profile of mental functioning”, which evaluates nine essential categories of inner experience – e.g., capacity for regulation, attention, and learning, or capacity for relationships and intimacy, or capacity for differentiation and integration; and S for “symptom patterns – subjective experience”; the subjective experience is conjugated with affective patterns, mental contents, and accompanying somatic symptoms, leading to constellations of symptoms described on the second axis of DSM-IV-TR^(32,33). The P-M-S algorithm is a hybrid categorical-dimensional concept, since the P and S axes are categorical disordered personalities and patterns of clinical manifestations, while the M axis is a dimensional formulation of individual emotional experiences, from “optimal” to “severe dysfunctional”⁽³²⁾.

The second chapter of PDM is dedicated to infancy and early childhood mental disorders, and includes three categories: interactive disorders (e.g., anxiety, depression), regulatory-sensory processing disorders (e.g., inattention, sensory seeking), and neurodevelopmental disorders of relating and communicating (e.g., self-absorption, repetitive behaviors)^(32,33). These disorders are established based on the analysis of three axes: the MCA axis is dedicated to the profile of mental functioning, the PCA is dedicated to the personality patterns and disorders, and the SCA is dedicated to the symptom patterns in children and adolescents⁽³²⁾.

The second edition, PDM-2, was published more than 10 years after the first one, and preserves the same core objectives as the previous edition – i.e., to support the diagnosis and case formulation of psychiatric disorders based on psychodynamic principles^(29,34). The first part of the PDM-2 is also dedicated to the disorders observed in adulthood and preserves the P-M-S axes, while the second part addresses the mental disorders observed in adolescence (preserving the corresponding axes:  MA, PA, SA), the third part approaches the disorders of childhood (the axes are MC, PC and SC), the fourth section is focusing on the disorders detected in infancy and childhood (0-3 years old) and, as a novelty, a distinct section is dedicated to disorders in later life (with axes ME, PE and SE)⁽³⁴⁾. This edition also includes a section of tools for clinicians to help them better understand the PDM-2 approach – i.e., psychodiagnostic charts for each age interval and clinical illustrations based on the use of these tools^(34,35).

More specifically, the P axis allows for the assessment of the personality organization levels and personality syndromes, and it is considered the main axis for adults, because personality functioning influences mental functioning (M) and subjective experience of symptoms (S), a phenomenon that is not observed in other age groups⁽³⁵⁾. The authors of PDM-2 considered that, in adults, personality has become stable and should be addressed first, while in adolescents and children, the developmental aspects (on axis M) will require more attention⁽³⁵⁾.  For example, a patient who has a fear of close relationships and avoids contact with their own feelings will experience depression in a particular way, different from a patient who is overinvolved in their own emotions and scrutinizes them carefully⁽³⁵⁾.  In PDM-2, the psychotic level of personality was added to the healthy, neurotic and borderline, and it is defined by features such as overgeneralized concrete or bizarre thinking, socially inappropriate behaviors, severe annihilation anxiety, or various convictions refractory to others’ counterarguments; identity diffusion, poor discrimination between fantasy and external reality, reliance on primitive defenses are also attributes of a psychotic level of personality⁽³⁵⁾. At a clinical level, such psychotic personalities may be seen in patients with anorexia nervosa (the belief of being overweight in spite of the evidence of the contrary), extreme obsessive-compulsive disorder (not being able to resist their compulsions although they are self-harming), or dissociative disorders – impossibility to distinguish past traumas from present reality⁽³⁵⁾. Regarding the personality syndromes, PDM-2 lists depressive (central preoccupation – self-criticism, self-punitivenss, loss), dependent (central tension – keeping versus losing relationships), anxious-avoidant and phobic (central preoccupation – safety versus danger), obsessive-compulsive (central tension – submission versus rebellion against controlling authority), schizoid (central preoccupation – fear of closeness versus longing for closeness), somatizing (central tension – integrity versus fragmentation of bodily self), hysteric-histrionic (central preoccupation – fear/envy of opposite gender, unconscious devaluation of own gender), narcissistic (central preoccupation – inflation versus deflation of self-esteem), paranoid (central tension – attacking versus being attacked by humiliating others), psychopatic (central tension – manipulating versus fear of being manipulated), sadistic (central tension – suffering indignity/inflicting such suffering), and borderline (central preoccupation – self-cohesion versus fragmentation, engulfing attachment versus abandonment dispair) personalities⁽³⁵⁾.

The M axis allows for the description of 12 categories of basic mental functions, which are considered useful instruments for clinicians to formulate complex descriptions of each patient⁽³⁵⁾. This axis is built on a synthesis of psychodynamic, cognitive and developmental models, and its 12 areas are grouped into four domains: (1) cognitive and affective processes; (2) identity and relationships; (3) defense and coping; (4) self-awareness and self-direction⁽³⁵⁾. PDM-2 recommends a set of tools validated for the assessment of these constructs – for example, scales for the evaluation of defense mechanisms or for attachment styles, but also projective tests are considered to be useful (e.g., Rorschach test)⁽³⁵⁾.

Finally, the S axis refers to the descriptions of symptoms included in DSM-5 and ICD-10, but with a focus on the subjective experience of each patient⁽³⁵⁾. Affective patterns, mental content, accompanying somatic states and associated relationship models are explored on this axis, and the symptom patterns are listed under eight categories: predominantly psychotic disorders, mood disorders, disorders related primarily to anxiety, event and stressor-related disorders, somatic symptom and related disorders, specific symptom disorders, disorders related to addiction and to other medical conditions, and psychological experiences that may require clinical attention⁽³⁵⁾.

Regarding the specifics of PDM-2 in the elderly, the section dedicated to this age group mentions: the ME axis should take into account: (a) the optimal mental abilities, despite aging, (b) mental abilities impaired by aging, and (c) low-level mental abilities as part of the individual’s developmental history, not affected by age. There is a 5-point scale for assessing the level at which each of the 12 mental function is evaluated, similar to adults;  caution should be exercised when evaluating PE axis in elderly, since age-related behavioral features may confound the diagnosis of a personality disorder/pattern. The SE axis should take into account the subjective experience of loneliness and associated poor quality of life, multiple medical conditions, dependency, cognitive impairment, higher use of drugs and higher mortality rate; also, when conditions like late-onset schizophrenia, depression due to medical illnesses, physical disabilities or social isolation, anxiety related to medical illnesses, hoarding disorder, neurocognitive disorders, etc. are present, extra care is needed when assessing the subjective experience in this age group⁽³⁵⁾.

The advantages of PDM-2 are related to: (1) its simplicity – i.e., it has only three axes (unlike OPD); (2) the age-focused analysis and operationalized scoring; (3) the hierarchical format according to demographic characteristics – for example, the primacy of the P axis in adulthood and of M axis in the elderly population. Also, (4) by integrating the diagnoses of DSM-5 and ICD-10, this manual ensures a level of compatibility with the research based on these classifications. The existence of (5) clinician-friendly tools that may enhance the applicability of this system is another advantage for OPD-2. There are studies that confirmed the operationalizing of the PDM Adult section is valid by comparing with the Psychodiagnostic Chart (PDC), Minnesota Multiphasic Personality Inventory-2 (MMPI-2), the Karolinska Psychodynamic Profile (KAPP), and the Operationalized Psychodynamic Diagnosis (OPD)‌^(36,37). The PDM-2 was also proven a reliable assessment tool, with good inter-rater reliability and convergent validity with the Level of Personality Functioning Scale (LPFS); it was important to note that it did not require extensive training, as a recent study showed⁽³⁸⁾.

There are also several shortcomings related to (1) the need for further validation studies, especially on the P axis, and (2) the exploration of particular aspects related to each age group, especially to the elderly, which is a newly introduced section of the manual.

4. A biological-oriented nosographic system – the Research Domain Criteria (RDoC)

The RDoC initiative was launched in 2009 as a reaction to the increasing criticism addressed to the current phenomenological nosographies. For example, although many patients have the same diagnosis, the underlying neuropathology and genetic factors may be quite different, which explains why the RDoC framework refers to neurobiology, neurophysiology and observable behavior. This project was created and coordinated by the US National Institute of Mental Health⁽³⁹⁾.

The main differences between RDoC and the DSM/ICD nosography are: an important translational perspective (it is based on genetics and neurobiology); a dimensional approach to psychopathology; it envisages a structured approach to psychopathology based on psychometrically optimized measures; it is an integrative approach; it is not limited in any way to the currently defined diagnostic categories⁽³⁹⁾.

The major domains of RDoC, which represents areas of human functioning and behavior, are: (1) the negative valence system, which involves responses to aversive stimuli; (2) the positive valence system, which is involved in anticipating, obtaining and responding to rewarding stimuli; (3) cognitive systems; (4) systems for social process; (5) arousal/regulatory systems; (6) sensorimotor systems. All of these have references to genetics, molecular and cellular mechanisms, circuits, physiology, behavior and self-reports, as units of analysis for each domain⁽⁴⁰⁾. Further, each domain incorporates a variety of more detailed constructs. The neurodevelopmental axis follows the individual from conception through the stages of adulthood, and the environmental axis is used for all aspects that may model the research, such as family, school factors, neighborhood and culture, accidents, assaults, etc.⁽⁴⁰⁾

Clinicians and researchers explored the RDoC model for various psychopathological applications. In a review focused on the preventive interventions for alcohol/drug use in adolescents, the 22 retrieved interventions were distributed according to the RDoC domains, and distinct neurocognitive trajectories were defined based on the risk factors for substance use disorders (SUDs)⁽⁴¹⁾. Such risk factors were maladaptive emotional reactivity, altered reward sensitivity and maladaptive habit formation, short-sighted and impulsive preference, limited social network and low metacognitive awareness, and heightened stress reactivity and insomnia, corresponding to the core domains of RDoC⁽⁴¹⁾. Protective factors, established according to the same model, were adaptive emotional regulation, adaptive reward learning and healthy resources of rewards, far-sighted and goal-directed preference, healthy social network and high metacognitive awareness, and resilience, adaptation to stress and good sleep⁽⁴¹⁾. The practical relevance of these results is the possibility of therapeutically approaching the risk factors while enhancing the protective factors for SUDs in adolescents, based on the RDoC framework.

Another review explored the neuroimaging findings in adolescents with depression and formulated these findings within the RDoC model, the negative valence system being associated with dysfunctions in the amygdala, insula, fornix and hippocampus, while the cognitive system domain with amygdala and insula, followed by striatum and precuneus; for the arousal and regulatory systems domain, the amygdala was the most invoked, followed by insula, striatum and putamen⁽⁴²⁾.

An example of analyzing the risk for depression using the RDoC model includes the following elements and associated instruments: (1) positive valence systems – event-related potentials on EEG during specific tasks, startle eyeblink reflex, cardiovascular activity and skin conductance; (2) negative valence systems – event-related potentials, startle eyeblink reflex, skin conductance; (3) arousal and regulatory systems – sleep quality (actigraphy, EEG), cardiovascular activity and skin conductance, cortisol levels; (4) cognitive systems – event-related potentials; (5) sensorimotor systems – actigraphy, EEG spectral features of motor activity; (6) social processes – event-related potentials, social cognition tasks⁽⁴³⁾.

The advantages of RDoC are: (1) its construct simplicity; (2) the neurobiological and behavioral basis (quantifiable variables); (3) its value as a more homogeneous model than the currently available nosographic systems, but also (4) its continuously developing nature, which allows the integration of newly generated data from computational sciences, genomics, metabolomics, etc.

The limitations are related to (1) the lack of validated methods to support its dimensions and (2) the measurement systems which need further exploration to confirm their reliability⁽⁴⁰⁾. Also, (3) RDoC was initially designed for research purposes, not clinical use; therefore, translating this conceptual framework into a clinical setting is still challenging.

5. A clinical-oriented new taxonomy –
the Hierarchical Taxonomy of Psychopathology (HiTOP)

Observing that traditional taxonomies have the disadvantage of placing arbitrary boundaries between psychopathology and normality, and frequently vaguely defined limits between disorders, allowing for high rates of comorbidity and diagnostic instability⁽⁴⁴⁾, a consortium of clinicians and researchers started in 2015 to work on a new dimensional approach to psychopathology. HiTOP is a data-driven strategy that proposes a hierarchical grouping of psychopathological conditions in super-spectra, spectra, and subfactors, at the upper levels, and homogenous symptomatic components and maladaptive traits, at the lower levels^(45,46). In this perspective, psychopathology is a set of dimensions that are clustered into progressively broad and transdiagnostic categories, named “spectra”, and the basic rule of constructing this taxonomy is observing the covariation of symptoms, thus reducing heterogeneity; combining co-occurring syndromes in spectra helps in eliminating, or at least reducing the comorbidity rate^(44,46). Therefore, psychopathology is no longer distributed in different categories of diagnoses but along a set of dimensions, each with its own degree of severity, thus decreasing the diagnostic instability^(44,46). The declared purpose of this taxonomy is to improve clinical outcomes due to the focus on symptom severity instead of heterogeneous diagnoses and to the possibility of targeted treatment across various levels of hierarchy⁽⁴⁵⁾.

At the lowest level of the hierarchy, there are different signs, symptoms, clinical components and traits, such as euphoric activation, hyperactive cognition and reckless overconfidence, which are considered to possess the minimum heterogeneity⁽⁴⁶⁾. The next level is represented by syndromes, formed by grouping smaller components; for example, bipolar type I and II disorders are the results of assembling previously mentioned components⁽⁴⁶⁾. Further, at the more complex level of the taxonomy, a number of subfactors appear (i.e., mania, fear or substance abuse)⁽⁴⁶⁾. There are four spectra above the subfactorial level: i.e., “somatoform”, “internalizing”, “thought disorder”, “externalizing, disinhibited”, “externalizing, antagonistic”, and “detachment”, and above this level, the super-spectrum is represented by “general psychopathology”, reflecting an overall dysfunctional level⁽⁴⁶⁾. Each level, from the symptoms/signs to the super-spectra, is evaluated dimensionally and receives a score of severity⁽⁴⁶⁾. Therefore, the clinician may choose to address therapeutically the higher level of the hierarchy – for example, the spectrum level, thus implicitly addressing all the subjacent symptoms and isolated manifestations that constitute the respective spectrum^(46,47).

The assessment of the functional impairment is not related to each syndrome in this taxonomy but reflects the global dysfunction – for example, by using the World Health Organization Disability Assessment Schedule (WHO-DAS)^(46,47).

The RDoC, as a taxonomy conceptualized within a neuroscience framework, may be explored simultaneously with HiTOP, which is based only on clinical features. Several authors suggested such crosstalk between HiTOP and RDoC could be relevant for both systems in order to support their perspectives reciprocally^(47,48). If combined, RDoC may help clarify the substrate of the clinical dimensions in the HiTOP, while the last taxonomy could offer clinical targets, psychometrically defined, for the RDoC research⁽⁴⁸⁾.

However, it is important to note that, while the RDoC domains are the products of experts’ consensus regarding the biobehavioral systems involved in mental health, the HiTOP dimensions are empirically based on covariation among signs, symptoms and diagnoses⁽⁴⁸⁾. 

Other studies that focused on the validation of HiTOP concepts provided conflicting results. For example, the construct and criterion validity of the HiTOP spectra determined by the Personality Inventory for DSM-5 (PID-5) in 257 patients with somatic symptoms and related disorders showed that spectra are related to the categorical and dimensional measurements of somatoform disorders, but both the construct and criterion validity of the HiTOP spectra have problems^(47,49,50). In order to improve the clinical translation of this model, the Measures Development Workgroup of HiTOP is engaged in the formulation of questionnaires and interviews that are specifically tied to the elements of this taxonomy, thus providing a valid means to test specific HiTOP components^(47,49,50). However, a meta-analysis of structural evidence for the HiTOP model (n=35 studies; N=23 DSM diagnoses) confirmed that five transdiagnostic dimensions fit the DSM nosographic entities well, with most diagnoses presenting good factor loading on the expected factors, and congruence coefficients between factors showed a hierarchical structure consistent with the HiTOP model⁽⁵¹⁾.

The advantages of HiTOP are: (1) a potentially positive effect on improving the communication between mental health specialists by using dimensional rating with population-based normative; (2) a possible effect of improving the prognostic power due to the superiority of dimensional scores versus categorical diagnoses in predicting outcomes such as chronicity, functional impairment or comorbidities; (3) utility for planning the therapeutic approach, but this still has to be demonstrated; (4) avoids reification of categories by enhancing the dimensional approach to psychopathology^(45,49-51).

Limitations of this model are related to: (1) its novelty and lack of research that could confirm its validity; for example, there is no consensus on the pathophysiology of each dimension; (2) it has a rather complex terminology, that should be explained to all users; (3) the assessment tools matched to this model are still far from being validated; (4) the compatibility of HiTOP and DSM/ICD nosographic models is not yet validated.

6. Blending cultural-restricted
and internationally-validated diagnoses –
the Chinese Classification of Mental Disorders (CCMD)

The first edition of CCMD was published in 1985, and its last edition, CCMD-3, was released in 2001^(52-55). Although the CCMD-2-R (1989) was less harmonized with ICD-10, the last edition was presumably more influenced by the APA and WHO classifications^(52-55). Several entities were added – e.g., “traveling psychosis” and “culture-related mental disorders”⁽⁵⁶⁾. Regarding the compatibility of CCMD and DSM diagnoses, a study compared the results of administrating the Adult Diagnostic Interview (ADIS-2), an instrument able to generate diagnoses according to DSM-III-R and CCMD-2 systems (N=254 Chinese patients), and concluded on the reliability and validity of both models in most diagnostic categories⁽⁵⁷⁾. However, differences were noted in this study regarding neurasthenia and hysterical neuroses, indicating the need for cross-cultural studies focused on these discrepant disorders⁽⁵⁷⁾.

CCMD-3 is considered a medical classification system based on both symptoms and etiology, the last aspect becoming evident when cultural pathologies related to vital energy disequilibrium are presented, for example^(54,57). In the chapter dedicated to “psychotic disorders”, a specific pathology is defined – i.e., “traveling psychosis”, which is an acute psychosis observed in individuals who travel long distances in trains, buses, ships, or planes, due to overcrowding, lack of sleep, lack of nutrition and water, chronic hypoxia, mental stimulation, etc.; the patients present impaired consciousness, fragmentary delusions, hallucinations and behavioral disorders. The duration is short, varying from a few hours to one week after stopping the travel. For the diagnosis of schizophrenia, at least two criteria out of nine (mainly positive symptoms) should be present for more than one month⁽⁵⁴⁾. CCMD-3 acknowledges the existence of “recurrent episodic mania” (unlike ICD-10) and “minor depression”, as well as the other ICD-10 entities within this chapter, but there is no “moderate depression⁽⁵⁵⁾. 

The chapter “Hysteria, stress-related disorders, neurosis” includes a set of hysteria disorders – e.g., hysterical amnesia, hysteria fugue, hysterical psychosis, hysterical stupor, and transient hysterical disorders occurring in childhood and adolescence⁽⁵⁴⁾. The neuroses in the CCMD-3 are almost all anxiety disorders and somatoform disorders acknowledged by the DSM-IV-TR framework. Eating disorders, sleep disorders and sexual dysfunctions are grouped in the chapter “Physiological disorders related to psychological factors”⁽⁵⁴⁾.

A sub-chapter dedicated to “Mental disorders related to culture” includes syndromes that are induced by disequilibrium of vital energy (Qigong), mental disorders due to witchcraft, and koro syndrome⁽⁵⁴⁾, all of which would be classified as unspecified dissociative disorders in the ICD-10 framework. 

The personality disorders in the CCMD-3 are paranoid, schizoid, dissocial, impulsive, histrionic, anankastic, anxious, dependent and other/unspecified⁽⁵⁴⁾. Pathological gambling, pathological stealing, pathological fire-setting, and trichotillomania are included in the “Habit and Impulse Disorders” chapter. The “Psychosexual disorders” mentioned by CCMD-3 are gender identity disorders, disorders of sexual preference, and sexual orientation disorders (including homosexuality and bisexuality)⁽⁵⁴⁾.

There is an optional section of CCMD-3 dedicated to treatment principles and another section dedicated to nursing diagnosis and measures⁽⁵⁵⁾. Also, similar to the ICD-10 structured interview (CIDI), CCMD-3 has its own diagnostic scale – i.e., Rating Test for Health Problems and Diseases (RTHD), with seven axial diagnosis systems⁽⁵⁵⁾.

The strengths of this classification are: (1) its comprehensiveness due to the inclusion of both culturally specific syndromes and internationally acknowledged disorders; (2) the correspondence between the ICD main chapters and its own nosographic categories; (3) concise definitions that approximate the ICD style.

The limitations of CCMD-3 may be formulated in relation to: (1) its limited acknowledgment to the specialists in other countries where large Chinese migrant populations may exist; (2) it has not been yet translated into other languages; (3) the statistical criteria that certain diagnoses have been founded on are not available for independent research and validation; and (4) there is a limited body of good-quality research about the some of the diagnoses included in CCMD that are not part of ICD, at least studies published in other languages than Mandarin.

7. Conclusions

As shown in Table 1, each nosographic model presents its own strengths and limitations; therefore, clinicians should be aware of the specific purposes for which they are using these instruments. However, regardless of the theoretical orientation psychiatrists adhere to, for practical reasons, they may have to use some of these classification systems more than others for daily purposes. For example, certain nosographic models have a larger area of coverage in clinical domains and are more grounded in objective data (epidemiology, statistics), being acknowledged at an institutional level, as is the case for ICD-10 in the European countries (at least in relation to the healthcare insurance companies) and DSM-5-TR in the United States of America (for similar purposes). To increase their flexibility of use, some of these classification systems have developed compatibility modules with larger recognized nosographic models, as in the case of OPD-3 with DSM (by including the GAF score and psychiatric diagnoses) and ICD (for psychatric and soamtic diagnoses), for example. Also, the task force that tried to ensure compatibility between DSM-5 and ICD-11 reflects the same effort toward creating common diagnostic categories that may facilitate cooperation in both the clinical and research fields of psychiatry.

Still, important challenges exist in the domain of psychiatric nosography, and these are related to psychosocial, economic and cultural factors. Regarding the question included in the title, it is important to have reasonable expectations when approaching the topic of how evidence-based psychiatric nosographic models are. This also means that it would be unrealistic to expect any time soon a unitary nosography since inherent variations across the biological, psychological, social and cultural characteristics of the patients with psychiatric disorders are obvious and reported by psychiatrists since the beginning of the nosography. These variations are a constant challenge to any psychiatric nosography and nosology, and they still preclude the formation of a unitary pattern of classification criteria. A more reasonable approach would be to adequately assess the most appropriate nosography for the purpose of the psychiatrist-patient interaction in order to avoid an expectation/result conflict. For example, using OPD-3 when having to decide if a patient is more suited for psychosomatic or psychotherapeutic intervention seems fair, but when planning to conduct a pharmacological trial, using the same nosography could drastically reduce the validity of the monitoring visits results, by inserting supplementary subjective variables. Another example of misusing these systems would be to rigidly apply the framework of a cultural-based nosography to non-native populations, to use HiTOP for monitoring psychotherapy evolution in a patient in the absence of prior research in this field, or to squeeze into the RDoC framework complex phenomenological concepts without correlating them first with quantifiable neurobiological or behavioral outcomes.

Not being a systematic review limits the generalizability of this article’s conclusions. Therefore, further research in this field should aim to find relevant data that support the validity and reliability of each nosographic system when used alone or in combination with others in clinical activity, pharmacological research, epidemiological investigation, judiciary environment, and therapeutic contexts.    

Corresponding author: Octavian Vasiliu E-mail: octavvasiliu@yahoo.com

Conflict of interest: none declared.

Financial support: none declared.

This work is permanently accessible online free of charge and published under the CC-BY licence.