Studying the Clinician: Judgment Research and Psychological Assessment

This scholarly book reviews the validity of clinical judgments, as shown by more than 1,000 controlled studies of assessments of personality and psychopathology, psychodiagnosis, case formulation, behavior prediction, treatment planning, and neuropsychological assessment. Although written primarily for psychologists, Studying the Clinician is certainly relevant to psychiatric practice. The difficulties in evaluating judgment validity are clearly expounded. In behavioral predictions and treatment planning, validity analyses are possible because outcome yields a predictive test. However, the correctness problem is vexing for personality, psychopathology, diagnostics, and causal judgments.

Clinicians can make reliable, valid judgments—if they are careful about the information used and that the type of judgment is likely to be valid. Most clinicians realize that the prediction of violence and suicide is difficult, but unfortunately the same wariness does not apply to describing defense mechanisms, making causal judgments, or allowing for the impact of mood state on personality evaluations. Interrater reliability was good for participants in the DSM-III and ICD-10 field trials; however, the raters were familiar with and adhered to the diagnostic criteria. The evidence indicates that clinicians do not adhere closely to diagnostic criteria.

Structured patient interviews are often inaccurate or incomplete. In a large sample of subjects studied prospectively since birth, retrospective reports agreed poorly with data collected prospectively. There is support, however, for Spitzer’s LEAD (longitudinal, expert, and all data) standard. The book is implicitly critical of epidemiologic studies whose subject diagnoses are based solely on verbal questionnaires.

Clinical psychopharmacological practice is also problematic because a surprisingly low proportion of treated patients with major depressive disorder receive antidepressant medicine. Among those who do receive medication, subtherapeutic doses and inappropriately short treatment periods are common. Comorbidity is frequently missed; for example, substance abuse is overlooked in psychiatric patients, and, conversely, anxiety and depressive disorders are not detected in substance abusers and mentally retarded patients.

The section on neuropsychological assessment is of particular interest for psychiatrists who find it difficult to assess such reports critically. The evidence is reassuring for standardized batteries such as the Halstead-Reitan and Luria Nebraska in terms of interrater reliability, impairment detection, and lateralization localization. However, only 14% of neuropsychologists consistently administer such batteries. Reliability has not been described for “flexible” approaches. In addition, a false positive rate of 21% should evoke caution in test interpretation. The accuracy of batteries has been shown to be substantially superior to that of commonly used tests for brain impairment, such as the WAIS, Bender-Gestalt, and MMPI.

The popular computerized interpretations of the MMPI are questioned. The rules that make up automated assessments are frequently commercial property; however, a study given access to program rules concluded that use of these reports requires “great care by qualified personnel as one of several types of evidence.”

The overwhelming diagnostic message is the importance of systematic adherence to explicit inclusion and exclusion criteria. Those who consider the DSM approach diagnostically inadequate have not, as yet, developed valid alternatives. To sum up, this unique book documents how clinical practice could be substantially improved, even given our current knowledge base.