Diagnostic Dilemmas in Classifying Personality Disorders

  • Poor correspondence between DSM-IV-TR diagnostic categories and typical presentations
  • Different structured interviews show poor agreement
  • Statistical analysis of PD criteria and traits fail to replicate DSM concepts
  • Investigations consistently fail to support the categorical representations of personality phenotypes
  • (Livesley, 2003)

Problems with DSM-IV-TR

  • Limited Clinical Utility
    • Conditions of the patients do not match the diagnostic concepts very closely
    • Specific diagnoses have a limited value for planning treatment or predicting outcome
    • Most diagnoses are global constructs
      • Pharmacological treatments tend to target specific dimensions rather than global diagnoses
      • Psychosocial interventions are directed toward specific behaviors
  • Lack of Exclusiveness and Exhaustiveness
    • Multiple diagnoses are the norm instead of being mutually exclusive
      • The most exclusive category is Obsessive-Compulsive PD, but it still has about 70% of all cases meeting the criteria for a second PD
    • According to some studies, the PDNOS is the most common diagnosis
      • This suggests that the system does not reflect common presentations
  • Psychometric Limitations
    • Agreement across different interviews is modest
    • Construct validity is an even greater problem
      • Internal validity
        • Internal consistency improved with the DSM-IV-R
        • Coefficient alpha falls below 0.7 for Histrionic, Dependent, and Schizotypal personality disorders
      • External validity
        • Convergent validity
          • Different measurement leading to the same diagnosis shows only modest agreement among different measures
        • Discriminante validity
          • Diagnoses are not distinct from each other
        • Predictive validity
          • Little evidence that diagnoses predict important variables related to etiology and outcome
  • Lack of Empirical Support for Diagnostic Concepts
    • Multivariate studies of personality characteristics consistently fail to generate factors that resemble DSM diagnoses
  • Atheoretical approach
    • It is atheoretical when it comes to etiology
    • Fails to offer a rationale for selecting diagnoses and criteria
      • Arbitrary selections that are drawn from diverse sources
        • Classical phenomenology
        • Traditional psychoanalytic theory
        • Spectrum disorders
        • Object relatons theory
        • Psychoanalytic thinking
        • Social learning concepts
  • Use of Categorical Diagnoses
    • Clinicians have to make arbitrary decisions, which leads to poor diagnostic agreement
    • This accounts for great diagnostic overlap, prevalence of the diagnosis NOS and limited validity

(Livesley, 2003)

Failures of the Categorical Model

  • Excessive Diagnostic Co-Occurrence
    • DSM-IV-TR routinely fails to indicate the presence of a specific pathology and suggest a specific treatment
    • Diagnostic comorbidity is so extensive that some argue for abandoning the term comorbidity in favor of a term that is more simply descriptive
    • Much of the PD diagnostic co-occurrence is readily explained if the DSM-IV-TR PDs are understood as maladaptive variants of general personality structure
  • Inadequate Coverage
    • PDNOS is one of the most frequently used Axis II diagnosis in clinical practice
    • Not entirely clear how clinicians are using PDNOS within their practice, but it is suggested that clinicians are not finding the existing diagnostic categories to be adequate in their coverage of PD symptomatology
    • Efforts to demarcate a limited number of specific categories to identify homogeneous and distinct groups, yet also provide adequate coverage, will likely continue to be problematic and frustrating
  • Arbitrary and Unstable Boundaries with Normal Psychological Functioning
    • DSM-IV-TR provides specific and explicit rules for distinguishing between the presence versus absence of each of the individual diagnostic categories but the schizotypal and borderline diagnoses are the only two for which a published rationale has ever been provided
    • No explanation, rationale, or even supportive discussion has ever been attempted for the diagnostic thresholds for the Avoidant, Schizoid, Paranoid, Histrionic, Narcissistic, Dependent, or Obsessive-Compulsive PDs
    • There have been many revisions, deletions, and additions to the criterion sets that the current diagnostic thresholds no longer relate well to the original thresholds
      • These unanticipated and substantial shifts in prevalence rates across revisions to the DSM are problematic to scientific theory and public health decisions
      • Seemingly minor changes to criterion sets result in substantial changes in prevalence rates
  • Heterogeneity Among Persons with the Same Diagnosis
    • DSM-III-R switched to polythetic criterion sets in which only a subset of diagnostic criteria are required
    • Polythetic criterion sets do not resolve the problems associated with the heterogeneity among persons sharing the same diagnosis
    • Polythetic criterion sets are simply an acknowledgement of the existence of this problematic heterogeneity
  • Inadequate Scientific Base
    • The only PD whose literature is clearly alive and growing is that of Borderline PD
    • There has been little comparable research on the etiology, course, pathology, or treatment of the Paranoid, Schizoid, Histrionic, Avoidant, Passive-Aggressive, or Obsessive-Compulsive PDs

(Widiger & Trull, 2007)

Dimensional Model of Classification

  • Five Factor Model (FFM)
    • FFM was developed originally through empirical studies of trait terms within existing languages
    • Lexical paradigm is guided by the compelling hypothesis that what is of most importance, interest, or meaning to persons is encoded within the language
    • Most important domains of personality functioning are those with the greatest number of terms to describe and differentiate various manifestations and nuance, and the structure of personality is evident in the empirical relationship among these trait terms
    • Initial lexical studies were conducted with the English language, and found a 5-factor Structure
      • Extraversion
      • Agreeableness
      • Conscientiousness
      • Emotional Instability
      • Openness
    • Disagreement about the single best term to describe each domain
      • Difficult to identify a single term to adequately characterize the entire range of personality functioning included within a large domain
    • Empirical support for the construct validity of the FFM as a dimensional model of personality structure is extensive
    • Heritability
      • Behavior genetic research has generally supported the validity of the domains and facets of the FFM and even the FFM structural model
      • Yamagata et al. concluded that the results support the view that the FFM reflects a genetic structure that is universal
      • Behavior genetic studies of individual PDs have been confined to Borderline, Antisocial, and Schizotypal PDs
        • Research concerning the seven other PDs have been so sparse that reviews of the heritability of these PDs have in fact based many of their conclusions on the behavior genetic research of normal personality traits, implicitly assuming that these PDs are in fact maladaptive variants of general personality structure
    • Universality
      • Etic studies
        • They use constructs and measures from one culture imported into another, determining whether the importation reproduces the nomological net of predictions previously obtained in other cultures
      • Emic studies
        • They use constructs and measures that are indigenous to a particular culture, determining whether a particular model of personality structure is evident from the perspective of that culture
      • FFM lexical studies would be considered emic studies
      • Virtually o systematic emic studies of PDs
      • Criticism of the emic lexical paradigm is that it might simply be studying folk concepts that lack any validity beyond the belief systems of a particular culture
      • There have been a few etic studies of the PD nomenclature of the DSM-IV-TR
        • Some have applied the PDs within an individual culture that is different from the predominant Western society in which the manual was largely created
          • It appears to be only one systematic multinational study, in which the DSM-III-R PD criterion sets were assessed in 14 mental health centers located in 11 different countries of North America, Europe, Africa, and Asia
      • The etic cross-cultural support for the FFM personality structure is extensive
        • Results show that the 5-dimensional structure was highly robust across major regions of the world, including: North America, South America, Western Europe, Eastern Europe, Southern Europe, the Middle East, Africa, Oceania, South-Southeast Asia, and East Asia
    • Childhood Antecedents
      • Remarkably little research examining the childhood and adolescent antecedents of the DSM-IV-TR PDs, with perhaps the exceptions of Antisocial, Borderline, and Schizotypal studies
      • Child and adolescent temperaments are probably among the best candidates for general broadband developmental antecedents for adult PDs
      • Limited amount of research relating empirically the temperaments of childhood with adult personality traits, but Shiner (1998) suggest that many of the apparently disparate temperaments being studied do appear to be well organized within 4 of the 5 broad domains of the FFM (extroversion, neuroticism, conscientiousness, and agreeableness)
        • Missing from Shiner’s theoretical model of childhood temperament was an openness dimension, which could reflect that preschool teachers do not generally distinguish curiosity and creativity from conscientiousness
    • Temporal stability
      • Fundamental to the concept of personality is temporal stability
      • Empirical support for the temporal stability of PDs has been elusive
      • Apparent failure of longitudinal studies to verify the temporal stability of PDs
      • Temporal stability has been well documented for general personality structure

(Widiger & Trull, 2007)

Five Factor Model of Personality Disorder

  • Integration of the psychiatric PD nomenclature with psychological models of general personality structure would go far in buttressing the weak construct validity of the DSM-IV-TR diagnostic categories
  • Primary concerns are obtaining a consensus structure, implementation, and clinical utility
  • Consensus structure
    • 18 alternative proposals for a dimensional model of PD
    • Proposals are so disparate that no consensus is likely to emerge
    • The FFM has been used effectively in many prior studies and reviews as a basis for comparing, contrasting, and integrating seemingly diverse sets of personality scales
    • Strengths of the Big Five taxonomy is that it can capture, at a broad level of abstraction, the commonalities among most of the existing systems of personality traits, thus providing an integrative descriptive model for research
    • One alternative proposal for DSM-V
      • Simply convert each diagnostic category to a 5-point Liker scale
      • One could then use these scales to provide profile descriptions of a patient
      • Limitation of this proposal is that dimensions consisting of the existing categories would be grossly overlapping
    • Two predominant dimensional models of the DSM-IV-TR PD symptomatology
      • 18 scales of the Dimensional Assessment of Personality Pathology
      • 12 scales of the Schedule for Non-adaptive and Adaptive Personality
      • They were both constructed by factor analyzing PD diagnostic criteria and symptoms to yield more distinctive scales of maladaptive personality traits
      • They would both provide profile descriptions that would be more differentiating and less susceptible to construct and scale overlap than 5-point Likert scales of existing diagnostic categories
      • Limitation of both of these scales as a sole replacement for the DSM-IV-TR diagnostic categories would be an absence of an explicit coordination with general personality structure
      • The ideal solution is likely to be a common integrative representation that includes the important contributions and potential advantages of each respective model
  • Implementation
    • Second concern is how a dimensional model of general personality structure would in fact be implemented in clinical practice
    • Dimensional classification is better suited for myriad clinical decisions that than the existing diagnostic categories because it can include different cutoff points for different clinical decisions
    • FFM description
      • 4 step procedure for an FFM diagnosis of PD
        1. Obtain a hierarchical and multifactorial description of an individual’s general personality structure in terms of the 5 domains and 30 facets of the FFM, providing a reasonably comprehensive description of the person’s adaptive and maladaptive personality traits
          • Recommend that clinicians use both a self-reprot inventory and a semi-structured interview because multiple methods provide more valid assessments of PD
        2. Identify social and occupational impairments and distress associated with extreme scores on the FFM personality traits
        3. Determine whether the dysfunction and distress reach a clinically significant level of impairment that would warrant a diagnosis of PD
          • An important area of future research will be studies relating the GAF to maladaptive personality functioning in order to develop precise cutoff points for specific clinical decisions
        4. Quantitative matching of the individual’s FFM personality profile to prototypic profiles of diagnostic constructs
          • Provided for clinicians and researchers who wish to continue to provide or study single diagnostic constructs
          • Clinicians and researchers can develop FFM profiles for PD constructs not included within DSM-IV-TR
          • Prototypal matching serves primarily to indicate the extent to which any single construct fails to provide a fully accurate or precise description of the individual person
  • Clinical Utility
    • Maser, Kaelber, and Weise (1991) indicated that the section of the DSM with which most were dissatisfied was the section of the personality disorders
    • Likely sources of frustration for clinicians
      • Heterogeneity of diagnostic membership
      • Lack of precision in description
      • Excessive diagnostic co-occurrence
      • Failure to lead to a specific diagnosis
      • Reliance on the personality disorder NOS wastebasket diagnosis
      • Unstable and arbitrary diagnostic boundaries
    • There have been no adequate empirical studies on the treatment of the Avoidant, Schizoid, Paranoid, Histrionic, Narcissistic, Obsessive-Compulsive, or Dependent PDs
    • PDs are among the more difficult disorders to treat
      • Treatment rarely invovles a comprehensive or complete cure of the PD and does not appear to focus on the entire personality structure
    • An integrated dimensional model of PD would consist precisely of the dimensions of maladaptive personality functioning that are currently the focus of clinical attention
    • Limitation of the FFM
      • Some of the lower order facet scales focus primarily on the normal variants of personality functioning that are themselves unlikely to be the focus of clinical interventions

(Widiger & Trull, 2007)