Approaches to psychiatric nosology: A viewpoint.

Psychiatric nosology is required for communication among clinicians and researchers, understanding etiology, testing treatment efficacy, knowing the prevalence of the problems and disorders, healthcare planning, organizing the services, and reimbursement purposes. Many approaches have been used for psychiatric nosology, including categorical, dimensional, hybrid, and etiological. The categorical approach considers illness as being either present or absent, and similarity with prototypical description of a disorder is taken as a marker for the disorder. The dimensional approach regards that symptoms of disorder exist on a continuum from normal to severely ill. The hybrid approach combines categorical and dimensional approaches, with categorical diagnosis for broad diagnostic group and dimensional indicator for severity. The etiological approach tends to find "reason" for the set of symptoms, which could be biological, psychological, or social. In this article, certain critical issues about the different nosological approaches are discussed. Hybrid approach currently seems to be the most preferred for widespread usage. In conclusion, psychiatric nosology needs to evolve through epistemic iteration leading to successive changes and devising a more refined and useful system with time.

INTRODUCTION

One needs nosology for a variety of purposes. It is required for communication among clinicians and researchers about what constitutes a particular disease and what does not. It is needed to further the research for understanding etiology, by having a set of symptoms being called as a disorder. It is needed to understand the treatment efficacy of various interventions, as one set of symptoms may respond to an agent while another set may not. Nosology is also useful to understand the prevalence of the problems and disorders, so that appropriate healthcare planning can be done. Moreover, it is required for insurance purposes while deciding what kind of disorders should be under the ambit of reimbursement. Hence, nosology cannot be done away with, and is in fact required for scientific application of the field.

EVOLUTION OF PSYCHIATRIC NOSOLOGY

Before one takes a critical look at the different approaches to psychiatric nosology (a classification or list of diseases), it would be worthwhile to look at the evolution of nosology and the purposes served by it. The nosology of psychiatric disorders has evolved over time. Beginning from the times of Hippocrates and Aristotle, attempts have been made to categorize and delineate various psychiatric disorders.[12] Hippocrates and his followers classified mental illnesses into various disorders, including mania, melancholia, paranoia, and phobias; which were assumed to occur due to imbalance in four humors. Pinel reduced all mental illnesses to four basic types.[3] Karl Kahlbaum and Ewald Hecker developed a descriptive categorization of syndromes, employing terms such as dysthymia, cyclothymia, catatonia, paranoia, and hebephrenia.[3] Emil Kraepelin grouped together a number of existing diagnoses and proposed 15 categories of mental illnesses.[4] He suggested the dichotomy of dementia praecox and manic-depressive insanity, and proposed disorders like psychogenic neurosis, psychopathic personality, and syndromes of defective mental development. The initial versions of Diagnostic and Statistical Manual and International Classification of Diseases (DSM-I and DSM-II, and ICD-7 and ICD-8) were based upon clinical description of cases. There was a psychodynamic tilt in the classification system. The DSM-III (1980) and ICD-9 (1975) represented a shift from the previous classificatory systems to a criteria-based diagnosis, which led to widespread usage of the classification systems in research and clinical practice. Subsequently, classification systems incorporated new entities in the later editions (DSM-IV, DSM-IV TR, and ICD-10).[567] The DSM-IV TR and ICD-10 are the currently most widely used nosological systems in psychiatric practice.

APPROACHES TO PSYCHIATRIC NOSOLOGY

Many nosological approaches to diagnosis have been considered for the understanding of the disorders. These prominently include categorical approach, dimensional approach, and hybrid approach among others. Other approaches that have been considered from time to time include etiological approach to classification. Herein we consider the various approaches and examine their uses and pitfalls [Table 1].

Table 1

Psychiatric nosology systems: Summary

We take a sample case for exploring the various approaches. Mr. A, a 40-year-old banker, has symptoms of persistent and pervasive sadness for about a month now. He fatigues easily and has less interest in activities, as compared to his previous self. His appetite has reduced and he has decrement in sleep. He expresses ideas of guilt and pessimistic views of future, but denies any suicidal ideas. He manages to go to work, but has poor efficiency. He has no significant medical illnesses. This is a case of what psychiatrists would usually label as a depressive disorder as per the current nosological systems.

CATEGORICAL APPROACH

This approach considers illness as being either present or absent. Depending upon the similarity with the prototypical description of a typical case, the disorder is labeled to be either present or absent. For a particular case, the set of symptoms either constitutes a disorder or not, and there are no “in between” diagnoses. According to this approach, there may be a large number of disorders depending on the prototypes and descriptions. If symptoms of two disorders are present, then the two disorders are said to be present as comorbid disorders (e.g., generalized anxiety disorder and depressive disorder existing as comorbid disorders). According to the categorical approach, disorder is distinct from normalcy (i.e., either one is ill or not ill).

Conceptualizing our case according to categorical approach, Mr. A has depressive disorder which is distinct from “normal” functioning. It is an entity separate from dysthymia, depression not otherwise specified, and anxiety disorders which are taken as other disorders. With treatment if he improves adequately, he would not be depressed (i.e., depression will either be present or absent).

Many problems are encountered with the categorical approach. One of the major issues faced is determining the threshold of symptoms that determines the diagnosis. The threshold at which a diagnosis is made depends upon the individual assessor's conceptualization and sensitivity to individual symptoms. Considering our case, should depression be diagnosed when patient is sad 8 h a day, 14 h a day, or 20 h a day? Should depression be diagnosed when patient is somewhat sad, moderately sad, or intensely sad?

Another problem with the categorical approach is that of sub-syndromal symptoms. Symptoms of a disorder usually present at a variety of levels of severity. It has been seen that sub-syndromal symptoms are associated with dysfunction and disability, and treating them leads to improvement.[8] But as they fall below a diagnostic threshold, no diagnosis would be made according to categorical approach, and consequently there would be no intervention or reimbursement for the same.

What constitutes a prototypical case may also be a matter of contention. The conceptualization of prototypical case may vary between experts and centers.[910] Whether a case fits a diagnosis depends upon what cluster of signs and symptoms are taken for making the diagnosis, and the agreement between assessors (as measured by kappa) may not concur on whether a patient has a psychopathology or diagnosis.

Challenges may also be faced while delimiting from other disorders. If two sets of symptoms (of two disorders) are present, then are they two different syndromes (e.g., depressive disorder and generalized anxiety disorder) or the combination should be considered a different syndrome (e.g., schizoaffective)? Often psychiatric disorders are comorbid with others, raising the issue that whether segregation into pigeonholes is the right approach.[111213]

Since some of the symptoms of a disorder are frequently encountered in another disorder, hierarchy of diagnosis has been proposed.[14] Presence of a disease higher in the hierarchy subsumes symptoms of disorder lower in the classification. This leads to diagnostic quandaries. For example, if a patient has symptoms of both schizophrenia and depression, then should depression be coded separately as a diagnosis or should only diagnosis of schizophrenia be considered (being higher in the hierarchy of diagnosis)? It has been suggested that the evidence for hierarchy is minimal,[15] implying that disorders should be coded as comorbid.

DIMENSIONAL APPROACH

According to this approach, symptoms of disorder exist on a dimension which is a continuum from normal to severely ill. Dimensionality can be envisaged in terms of number (count) of symptoms (e.g., five out of eight symptoms to diagnose major depressive disorder) and severity of each symptom group (e.g., positive, negative, disorganized, cognitive, affective dimensions of schizophrenia).

A disorder with three or more ordinal categories (e.g., mild, moderate, and severe) can be said to have a dimensional approach. Scales such as Hamilton Depression Rating Scale (HDRS) represent a dimensional evaluation. The dimensional approach suggests that symptoms may be present in normal as well as in ill. A cutoff is used to ascertain the threshold of diagnosis (which transforms the disorder into categorical).[16] The diagnostic threshold is determined by the expert's opinion.

Conceptualizing the present case of Mr. A according to the dimensional approach, sadness exists on a continuum with severity ranging from normal to severely pathological. The grading of individual symptoms and overall sadness can be done using HDRS, and a cutoff score of >7 on HDRS is used for diagnosing depression. The treatment response can be monitored by using the severity scales.

Challenges are faced while using the dimensional approach as well. Can we describe all mental illnesses on few dimensions or do we need to have different dimensions for each disorder? A related question posed is that for a particular disorder, whether to take only one dimension of severity or multiple dimensions of symptom sets. In our case, the question is whether to have a separate dimension for depression or a combined set of dimensions for all disorders (like depressive symptoms, psychosis, biological functioning, etc). Also, whether to use one dimension of severity (e.g., HDRS) or to use multiple dimensions describing different symptoms like sadness, cognitive symptoms, and biological function.

Also, when symptoms are on continuum, then when should they be treated or intervened with? Improvement in low-grade symptoms (when they approach normalcy) is less with medication.[8] Treating everyone with any degree of symptoms is not cost-effective from public health viewpoint. Hence, using a purely dimensional approach may not offer assistance in making healthcare and clinical decisions.

HYBRID APPROACH

Hybrid approach represents a combination of categorical and dimensional approaches. It utilizes categorical approach for broad diagnostic group and dimensional approach for severity of the disorder. The present-day ICD and DSM use this approach to some extent. The diagnosis of depressive episode suggests a broad categorical diagnosis, while ratings of severity into mild, moderate, and severe represent dimensionality. However, both ICD-10 and DSM-IV TR tend to use categorical approach to a larger extent.

The upcoming DSM-5 is likely to use hybrid approach more effectively.[1718] A categorical diagnosis will still be made (e.g., schizophrenia spectrum and other psychotic disorders), but an additional dimensional construct would be applicable for most disorders to convey severity (ratings on a 0-4 scale on various symptoms cross-sectionally, with severity assessment based on past month). The DSM-5, which is in field trials at the time of writing, retains the categorical diagnosis but introduces dimensionality on a larger scale.

Conceptualizing the case of Mr. A from the hybrid perspective, categorical diagnosis would be of major depressive disorder – single episode, while severity would be measured by clinical global impression (CGI) of the past 1 week (rated from 0 to 7). Progress can be measured by change in CGI over time. Even when episode is in remission, sub-syndromal symptoms can be measured using CGI.

The hybrid approach also has its share of problems. It combines the issues of dimensional and categorical systems. It can be seen as a compromise approach that does not have clear biological reasoning. For practical purposes, the cutoffs agreed for a scale to make diagnosis are based on a decision of experts, and they do not reflect presence or absence of illness in a true sense. Also, there is difficulty in reliably measuring psychiatric constructs, which may lead to inaccuracies in diagnosis and severity assessment.

OTHER APPROACHES

The etiological approach tends to find “reason” for the set of symptoms. The reasons could be biological, psychological, or social. This approach has been suggested time and over in the past. It has been propounded persuasively for some psychiatric disorders like autism.

However, the etiological approach is difficult to use for psychiatric disorders with the present understanding of etiology (which is usually multifactorial). Interplay of genetic and environmental factors is implicated in development of most disorders. Role of each factor can be hardly quantified with certain degree of precision. Also, it may be unclear whether a factor causes the disorder or merely unmasks it in susceptible individuals.

WHICH APPROACH IS THE BEST?

Hybrid approach seems to the most useful currently from various standpoints. It is useful to the clinicians using the categorical aspects of the classificatory system for making a diagnosis. It helps researchers who are more interested in the dimensional aspects to assess treatment response and ascertain etiology. It helps administrators and healthcare planners who use both categorical and dimensional aspects to make decisions. It also helps patients and caregivers to imbibe proper information (both categorical and dimensional aspects of the disorders) to understand what kind of illness is present and to what degree?

It seems that psychiatric nosology needs to evolve through a process called as “epistemic iteration.”[1] Changes would occur to improve the nosological system with time, keeping in view the gained scientific knowledge over the period. However, changes in psychiatric classification are likely to be influenced by views of experts and political influences. The most illustrative example is the inclusion of homosexuality as a disorder and then its removal as a disorder. Hence, psychiatric classification systems are likely to have “wobbly” (back and forth) iterations. The present system has evolved over the ages, incorporating the scientific knowledge gained and practicalities adhered to. The system is not a perfect one, but serves a lot of purpose. It is expected that with time, a more refined and useful system will evolve on the strength of further research findings.