Limitations of Reliability and Validity

Limitations of Reliability and Validity

Despite the obvious role of quantification in neuropsychologi-cal testing, interpretation of test data ultimately depends on the knowledge base, training and skill of the clinician. Neuropsy-chological test scores are an indirect measure of the status of the brain, as contrasted with direct measures of structure by magnetic resonance imaging (MRI) or function by positron emission tomography (PET), functional MRI, or EEG. In a psy-chiatric setting, where problems of motivation, effort, coopera-tion and stage of the illness are ubiquitous, analysis of neuropsy-chological data must go beyond the level of performance deficits because many studies have shown performance to be especially affected by functional (emotional) factors. Process analyses ori-ented to focal syndromes and focused on the relative efficiency of the two sides of the body and hemispace may enhance predic-tive validity.

The patient’s clinical state may change, and repeated test-ing when the patient’s clinical status is optimal often clarifies the nature of the diagnosis. Selective deficits found in the con-text of otherwise good performance when patients are tested in their best state can be considered most valid. Neuropsychologists must also take into account the effect of medication on neuropsy-chological function and distinguish medication effects from the patient’s adaptive ability. Different medications are likely to produce different effects. For example, Trimble and Thompson (1986) have demonstrated that, for epileptic patients and normal subjects, anticonvulsants have negative effects on most measures of neuropsychological testing. On the other hand, Cassens and colleagues (1990) have demonstrated that (traditional) antipsy-chotic medications have negligible or mildly positive effects on most measures of neuropsychological testing in chronic schizo-phrenia, with the exception of a negative effect on motor per-formance. Spohn and Strauss (1989) have indicated that typical antipsychotic medications tend to improve attentional perform-ance, such as on versions of the CPT.

Structured Clinical Instruments and Rating Scales

Structured instruments and rating scales have been developed primarily for research purposes. They allow investigators to compare findings in different studies by ensuring that similar data and criteria have been used to establish diagnoses and to measure the presence and severity of psychiatric symptoms and their response to treatment. Many types of mental health profes-sionals and, in some cases, nonclinicians can be trained to ad-minister these rating scales.

Although most practicing clinicians do not commonly use structured instruments to assess or follow-up patients, a small number of rating scales have come to be used routinely in clini-cal practice. For example, the Abnormal Involuntary Movement Scale (Figure 19.3) is often used to monitor patients receiving antipsychotic medication for the presence of tardive dyskinesia, and the Global Assessment of Functioning Scale (Figure 19.4), which is a slight modification of the Global Assessment Scale, is now used in Axis V in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV-TR).

Table 19.11 shows some of the most commonly cited struc-tured instruments and rating scales. Hundreds of other special-ized scales are also in use to assess such diverse areas as person-ality disorder, aggressive behavior, sexual practices, stressful life events and quality of life.

Table 19.12 gives an example of how these different rat-ing scales approach the assessment of two symptoms: guilt, a purely subjective state of mind, and suicide risk, an incli-nation that is assessed using both subjective and behavioral components.