Advantages and Disadvantages of the DSM‐5 Classification System

With any system of classification like DSM‐5, there will always be strengths and weaknesses. One of the major strengths of the DSM system is the wide acceptance and use of the system. Mental health professionals in the United States routinely utilize diagnostic systems in their work if for no other reason than to allow their clients to receive treatment in hospitals and reimbursement from health care providers. Each category of disorder has a numeric code taken from the ICD coding system, used for health service (including insurance) administrative purposes.

The limitations of the DSM system are reflected in the terminology related to diagnosis itself. Since the DSM-3, the goal was to improve the uniformity and validity of psychiatric diagnosis in the wake of a number of critiques, including the famous Rosenhan experiment (see box below). There was also felt a need to standardize diagnostic practices within the United States and with other countries, after research showed that psychiatric diagnoses differed between Europe and the United States.^[1]

Reliability and Validity

The revisions and refinements in the DSM classification system have been largely driven by the need to improve reliability and validity. Reliability measures how consistent a diagnosis is and how reliably the categories can be judged. Validity looks at how accurate the diagnosis is and how valid the categories are in the sense of discriminating among disorders that have distinctive etiologies and possibly require different treatments.

To be useful, any diagnostic system must demonstrate reliability and validity. The revisions of the DSM from the third edition forward have been mainly concerned with diagnostic reliability—the degree to which different diagnosticians agree on a diagnosis. If clinicians and researchers frequently disagree about the diagnosis of a patient, then research into the causes and effective treatments of those disorders cannot advance. To be considered reliable, or consistent, different evaluators using the system should arrive at the same diagnoses when they evaluate the same people. For example, a diagnosis of major depressive disorder, a common mental illness, had a poor reliability kappa statistic of 0.28, indicating that clinicians frequently disagreed on diagnosing this disorder in the same patients. The most reliable diagnosis was major neurocognitive disorder, with a kappa of 0.78.^[2]

Critics assert, for example, that many DSM-5 revisions or additions lack empirical support; inter-rater reliability, or the degree of agreement among raters, is low for many disorders; several sections contain poorly written, confusing, or contradictory information; and the psychiatric drug industry unduly influenced the manual’s content (many DSM-5 workgroup participants had ties to pharmaceutical companies).footnote]Welch, Steven; Klassen, Cherisse; Borisova, Oxana; Clothier, Holly (2013). “The DSM-5 controversies: How should psychologists respond?”. Canadian Psychology. 54(3): 166–175. doi:10.1037/a0033841[/footnote]

Diagnostic validity concerns whether the diagnosis measures what it claims to measure. In psychiatry, there is a particular issue with assessing the validity of the diagnostic categories themselves. In this context,

• content validity may refer to symptoms and diagnostic criteria.
• concurrent validity looks at whether the markers that indicate a disorder actually correlate with the disorder, and if these markers are true over time and across multiple people and measures. It may also look at the appropriate alignment of a treatment response to a disorder.
• predictive validity may refer mainly to diagnostic stability over time and the relationship between the diagnosis and the ability to predict people’s behavior in certain situations.
• discriminant validity may involve delimitation from other disorders.

Robins and Guze proposed in 1970 what were to become influential formal criteria for establishing the validity of psychiatric diagnoses. The five phases of diagnostic validation used by Robins and Guze were (1) clinical characteristics of the syndrome and of the patients who develop it (including core symptoms, demographic characteristics, and precipitating factors); (2) exclusionary criteria differentiating the syndrome from other known disorders; (3) family studies; (4) laboratory data (radiological, chemical, pathologic, and psychological evidence); and (5) follow-up studies (for diagnostic stability, course, and treatment response).^[3] Once the characteristics of disorder, the exclusionary criteria, family, lab, and follow-up studies are completed, there should be enough information to deem the DSM content as a valid description of the mental disorder.

Figure 1. Radiology laboratory studies like MRIs are included in the criteria for validity.

These were incorporated into the Feighner Criteria and Research Diagnostic Criteria that have since formed the basis of the DSM and ICD classification systems. ^[4]

Other critics also believe the DSM needs to become more sensitive to the importance of cultural and ethnic factors in diagnostic assessment. They encourage us to consider and understand that the symptoms or problem behaviors included as diagnostic criteria in the DSM were largely determined by a consensus of mostly U.S.-trained psychiatrists, psychologists, and social workers. Imagine that the American Psychiatric Association had asked Asian-trained or Latin American–trained professionals to develop their diagnostic manual? Might there might have been different diagnostic criteria or even different diagnostic categories? ^[5]

Overall, the bulk of limitations of the DSM system listed so far are related to the diagnostic system in general. It is worthwhile to be alert to the criticisms of the DSM system, particularly since they serve as a reminder that mental health professionals aim to help the individual and not the disorder. Furthermore, the DSM system has been useful to insurance companies who adopt its use to establish coverage for certain clinical disorders, and has been helpful in allowing researchers and clinicians to have a common language with which to discuss clients. 

Differences between Categorical and Dimensional Understanding of Behavior

Both the DSM and ICD systems represent a categorical approach to the description of mental disorders. As the name implies, a categorical approach attempts to categorize mental disorders into distinct diagnoses. The categorical approach is based on the idea that a person either meets criteria for a specific disorder or they do not. Traditional categorical models of classification, which are based on the presence or absence of symptoms, do not take into account levels of expression of a characteristic or the presence of any underlying dimension. This conceptualization allows a mental health professional to make a dichotomous decision (i.e., yes, the person meets criteria or no, the person does not meet criteria for that disorder). A dimension refers to a continuum on which an individual can have various levels of a characteristic, in contrast to the dichotomous categorical approach in which an individual does or does not possess a characteristic. 

The DSM-4 organized diagnoses using five separate axes (clinical disorders, personality disorders, general medical disorders, and then sections on psychosocial and environmental factors, and the global assessment of functioning (which measured the severity of a disorder). An axis is defined as a category of information regarding one dimension of an individual’s functioning. In the DSM-IV-TR system, an individual was diagnosed on five different domains, or “axes.” The DSM-5 got rid of the axis system, as there was no scientific justification for grouping things by axis. Originally, the multiaxial system was intended to allow professionals to characterize clients in a multidimensional way (medical, psychological, developmental, and environmental/social status), but rationale for removal of the multi-axial system was based on unclear boundaries between medical and psychiatric diagnoses, inconsistent use of Axis IV by clinicians and researchers (psychosocial and environmental problems), and poor psychometric and clinical validity of Axis V (Global Assessment of Functioning).^[6]

The most important change in DSM-5 was the inclusion of dimensions in diagnoses; for example, how severely ill is a patient with schizophrenia or depression? The dimensional approach focuses on varying levels of different behaviors that a person exhibits, rather than whether or not a person meets criteria for a particular disorder. The dimensional approach is also included in Section III (“Emerging Measures and Models”). This section includes assessment measures and diagnoses not considered well-established enough to be part of the main system. For example, an “Alternative DSM-5 Model for Personality Disorders” is described.^[7] The decision to retain the old DSM-4 categorical model for personality disorders in DSM-5 was controversial (currently the ten personality disorders are grouped into three general categories), and efforts continue to persuade the American Psychiatric Association to replace it with the dimensional model in DSM 5.1. ^[8]

Since the categorical model is widely used in clinical practice and has a significant body of research supporting it, its common usage is compelling to laypeople when they are judging the credibility of professional opinion. Therefore, the dimensional approach is often further criticized for being difficult to interpret and less accessible. It is, however, widely used in some professional settings as the established approach, for example by forensic psychologists. ^[9]

Regardless of whether you use a categorical or dimensional approach to understand a person’s functioning, it is crucial to remember that behavior can change over time.

Problems Associated with Classification and Labeling

Finally, but probably most importantly, any diagnostic system such as the DSM system allows for individuals to be labeled for behavior that may or may not be an important part of their character. Labeling occurs when information about a person’s diagnostic classification is communicated in a negative manner that leads to stigma for the individual with a mental disorder.

The “Mentally Ill”

The social construction of deviant behavior plays an important role in the labeling process that occurs in society. This process involves not only the labeling of criminally deviant behavior, which is behavior that does not fit socially constructed norms, but also labeling that which reflects stereotyped or stigmatized behavior of the “mentally ill.” The labeling theory was first applied to the term “mentally ill” in 1966 when Thomas J. Scheff published Being Mentally Ill. Scheff challenged common perceptions of mental disorders by claiming that mental disorder is manifested solely as a result of societal influence. Scheff argued that society views certain behaviors and actions as deviant and, in order to come to terms with and understand these actions, often places the label of mental disorder on those who exhibit them. Certain expectations are then placed on these individuals and, over time, they unconsciously change their behavior to fulfill them. Criteria for different mental illnesses are not consistently fulfilled by those who are diagnosed with them because all of these people suffer from the same disorder, they are simply fulfilled because the “mentally ill” believe they are supposed to act a certain way so, over time, come to do so.^[10]

Scheff’s theory had many critics, most notably Walter Gove who consistently argued against Scheff with an almost opposite theory; he believed that society has no influence at all on mental disorder. Instead, any societal perceptions of the “mentally ill” come about as a direct result of these people’s behaviors. Most sociologists’ views of labeling and mental disorder have fallen somewhere between the extremes of Gove and Scheff. On the other hand, it is almost impossible to deny, given both common sense and research findings, that society’s negative perceptions of “crazy” people has had some effect on them. It seems that, realistically, labeling can accentuate and prolong the issues termed “mental illness”, but it is rarely the full cause.^[11]

Labeling theory posits that self-identity and the behavior of individuals may be determined or influenced by the terms used to describe or classify them. It is associated with the concepts of self-fulfilling prophecy and stereotyping. However, the label of “mentally ill“ may help a person seek help, for example, psychotherapy or medication. Labels, while they can be stigmatizing, can also lead those who bear them down the road to proper treatment and (hopefully) recovery. If one believes that “being mentally ill” is more than just believing one should fulfill a set of diagnostic criteria, then one would probably also agree that there are some who are labeled “mentally ill” who need help. It has been claimed that this could not happen if there were no way to categorize (and therefore label) them, although there are actually plenty of approaches to these phenomena that don’t use categorical classifications and diagnostic terms, such as spectrum or continuum models. Here, people vary along different dimensions, and everyone falls at different points on each dimension.

The issue at stake is that DSM-5 may lead to the increasingly widespread “medicalization” of psychology. It is suggested that – also due to its impact via the social media – DSM-5 is likely to turn into a true “social representation” (Moscovici et al., 2001) with the power to strongly influence clinical practice, pushing it in the direction of the large-scale prescription of drugs.^[12] Allen Frances, Chair of the DSM-IV Task Force, came to be a remarkably prolific and vocal critic of the proposed changes that came out in the DSM-5. In a BMJ editorial, he described the ‘grave’ consequences of ‘false positive epidemics’ of disorders that would be constituted through inappropriate usage of new diagnostic entities; in so doing, DSM-5 would ‘expand the territory of mental disorder and thin the ranks of the normal’. In other words, it would help to further ‘medicalize’ society.^[13]