|| Human Nature Review ISSN 1476-1084 | Table of Contents | What's New | Search | Feedback | Daily News | Submit A Manuscript ||
PDF of this article Download Adobe Acrobat Reader Email the reviewer Publisher's web site Send a response to this article Search the web for related items Contact the Editors
Human Nature Review 2002 Volume 2: 515-517 ( 21 November )
URL of this document http://human-nature.com/nibbs/02/boggs.html
by Anne Farmer, Peter McGuffin and Julie Williams
Oxford University Press, 2002
Reviewed by Barry Boggs, Ph.D., ABC Sciences PC, Memphis, TN, USA.
In Measuring Psychopathology, Farmer, McGuffin and Williams have written a concise, practical introduction to a topic that can be unwieldy and not without controversy, namely, what is psychopathology and how can it be reliably measured.
Twentieth century psychiatry has seen competing theories of psychopathology, competing classification systems, competing diagnostic manuals and a vast body of empirical research findings that cannot always be reconciled. Farmer, McGuffin, and Williams bring order to this situation by confining themselves to definitions and diagnoses that can be operationalized. Those who have followed developments in psychiatry since the publication of DSMIII will not find this to be a novel approach but Farmer, McGuffin, and Williams demonstrate just how far it has come, especially in their discussion of converting operational diagnostic criteria to algorithms that would allow a computer to “make diagnoses.”
But, Farmer, McGuffin and Williams remain sensitive to the historical roots of current diagnostic approaches and in the opening chapter provide a brief history of earlier definitions of psychopathology offered by Freud, Meyer, Szasz and others. Their discussion of Szasz was a little too brief for this reader. Szasz still has a following in the United States and proponents of a “nondisease” or “nonlabeling” approach to psychopathology are not few in number, especially among clinical psychologists, who often reluctantly provide diagnoses to satisfy the demands of third party payers. But, like Szasz, Farmer, McGuffin and Williams do reject the disease model of psychopathology, albeit for different reasons. They prefer the concept of disorder, more specifically, they prefer the concept of mental disorder contained in the ICD-10 that, in part, reads, “Disorder is not an exact term but is used here to imply the existence of a clinically recognizable set of symptoms associated in most cases with distress and with interference of personal functions.” Disorder can be operationalized and measured while disease is illusive except in clear cases of organic syndromes.
After dealing with definitional issues in chapter 1, in chapter 2, Farmer, McGuffin, and Williams turn to signs and symptoms of psychopathology which they call the "basic building blocks" of diagnostic categories. Again, the discussion is quite concrete and the symptoms are largely observable behaviors or subjective reports of distress. The discussion is structured around the conventional components of the mental status exam – appearance and behavior, abnormalities of speech, abnormalities in mood, content of thought, abnormal perceptions and abnormalities in cognition.
Chapter 3 focuses on classification systems, principally, the two most widely used ones in Europe and North America -- the International Classification of Diseases, tenth edition (ICD10), produced by the World Health Organization and the Diagnostic and Statistical Manual, fourth edition (DSM IV) produced by the American Psychiatric Association. The format of this chapter is quite helpful in that various aspects of the ICD10 and DSM-IV are laid out in comparison charts. The criteria for depressive episode (ICD10) and major depressive disorder (DSM-IV) are compared. One is impressed more by the similarities than the differences suggesting, perhaps, the continuing convergence of the ICD10 and DSM-IV. The authors remind us of the value of reliable, operationalized diagnostic criteria and standardized diagnostic procedures by citing the now famous epidemiological studies in which “huge differences” were found in first admission rates for schizophrenia and manic-depressive disorder between New York and London with the former being much higher than the latter. In follow-up studies using standardized interviewing procedures and computerized scoring conducted by trained researchers in the UK and USA, the difference in rates of the disorders were significantly reduced, while among practitioners in the field the difference in rates between the UK and USA remained the same
Chapter 4 reviews basic concepts of validity, reliability and utility as pre-requisites for accurate measurement of psychopathology. This will be familiar territory for experienced researchers and clinicians but might be a useful introduction to those just beginning their training (or a review for those taking their first round of graduate examinations!).
From this point on, Measuring Psychopathology is largely a discussion of the relative merits of specific techniques, procedures and tests used to assess psychopathology. This discussion occurs at two levels – a comparison between procedures within a category, for example, the respective merits of structured and semi-structured interviews, and between categories, for example, the respective merits of rating scales and interviews. In one of the book’s more fascinating discussions the authors note that the received wisdom in the field seems to be that the “gold standard” of diagnosis is the “best estimate” approach whereby a group of researchers review all the information on a subject and then arrive at a consensus diagnosis (this is not an uncommon procedure among clinicians who work in hospital or clinic settings). But, this is costly and time consuming, they argue. They suggest that “An alternative method is to devise a method of summarizing all the available information in a way that can be fed into a computer and since operational criteria should be readily translatable into algorithms, allow a computer program to ‘make a diagnosis.’” The authors argue that computer programs have accompanied structured and semistructured interviews since their inception, thus, this is not an entirely new approach. They set about comparing computer generated diagnoses and the “best estimate” gold standard.
They cite a study by Craddock et al (1996) which suggested “that good to excellent agreement can be achieved between OPCRIT (a diagnostic procedure that uses a computer program for generating diagnoses) and those made by best estimate procedures.” (Quote is from Farmer et al.) Similarly, they cite a study by Azevedo et al (1999) using a slightly different approach but arriving at a similar conclusion, namely, “that the OPCRIT approach provides an efficient alternative to best estimate procedures.” (Quote is from Farmer et al.).
But, this evidence seems inadequate to support the conclusion that the authors offer at the end of this discussion. They conclude that, “There is therefore a strong argument at this stage that computers that can be relied upon to give one hundred reproducibility are far superior to the best trained humans.” If they are referring to the process of reproducibility then there can be little argument that computers are superior; however, if they are referring to the entire diagnostic process this seems a strong conclusion relative to the amount of evidence offered. There are other discussions in the book about the value of computers for generating diagnoses but this reader was unable to find other discussions comparing computers and best estimate procedures in the head to head fashion that this discussion addresses. It is quite likely that there are other studies on the topic. Nonetheless, over time the “one hundred per cent reproducibility” of computers can only add reliability to the diagnosis of psychopathology if the human input of signs and symptoms is accurate.
According to Farmer, McGuffin and Williams one of the unresolved issues in the measurement of psychopathology that researchers and clinicians alike will carry into the twenty-first century is whether psychopathology should be conceptualized as categorical or dimensional. Is it more accurate to say that a person has a disorder, yes or no, the way we characterize diseases, such as cancer and emphysema, in other branches of medicine, or is better to think of psychopathology as occurring along a continuum of, for example, subjective distress and interference with personal functions? But, as the authors ably point out, researchers, clinicians and students will carry two powerful tools in their diagnostic repertoire – one is neuroimaging and the correlation of neuroimaging with both cognition and affective states and the other is a vastly expanded knowledge of the human genome and human genetics. It may very well be that through media such as these some psychopathology, for example, various types of bipolar disorder, will become firmly established as categorical, while other forms of psychopathology, such as dysthymia or certain anxiety states, are more appropriately described along dimensions of subjective distress and deterioration in adaptive behavior.
Farmer, McGuffin and Williams have produced a compact, eminently readable and practical guide for researchers, students and clinicians whose everyday business is to diagnose and measure psychopathology. It would likely be a useful text for courses in psychopathology, measurement theory, research methods, psychological assessment or epidemiological research. General audience readers with some background in psychology would likely find select chapters to be quite interesting such as the opening chapter on defining psychopathology and the chapter on polydiagnostic approaches utilizing the computer. Parts of the text in which specific tests and procedures are reviewed might prove a little tedious for a general audience but even these might be helpful for persons who find themselves or someone they know in the role of consumer of mental health services.
Buy Measuring Psychopathology from:
Computer-generated translation of this page français deutsch español português italiano ― also try this alternative fast translation service.
© Barry Boggs.
Boggs, B. (2002). Review of Measuring Psychopathology by Anne Farmer, Peter McGuffin and Julie Williams. Human Nature Review. 2: 515-517.