Paper
Symptom Report and Diagnostic Properties of the Neuropsychiatric Questionnaire
- Authors:
- C. Thomas Gualtieri
- Abstract
- Background: A problem in psychiatry is the reliance on subjective data, especially self-reported symptoms. In this paper, we investigate the diagnostic validity of the Neuropsychiatric Questionnaire (NPQ) in a large group of patients from a private neuropsychiatric practice. Method: Patients over age 18 with generalized anxiety disorder (GAD), major depressive disorder (MDD), attention deficit disorder (ADHD), bipolar disorder (BPAD), and normal subjects were included in the analysis (N=1127). Inter-group comparisons were made by MANOVA, controlling for age, gender, education and computer familiarity with the criterion for significance at P < 0.01. Analysis of pairwise group differences was by one-way ANOVA with Bonferroni correction. Effect sizes were measured by Cohen’s d. The NPQ consists of 20 symptoms scale and four factors: cognitive, mania, somatic and anxiety-depression, which combine to make up a symptom load scale (SLS). Results: Patient groups differed in the makeup of relative contribution of the four factors to the SLS. When compared to normal subjects (N=45), those in the with (GAD, MDD, ADHD, BPAD), When the four diagnostic groups were compared in pairwise fashion to normal subjects (N=45), patient-reported symptoms occurred in the expected directions, but the effect sizes for the cited differences were, on average, small to moderate. In the discriminant validity analysis, the lowest Wilks’ Lambda was 0.622 and patients were correctly classified by the NPQ ranging from 65-79% of the time. Conclusion: The NPQ, while being a useful tool for the clinician, deserves a diagnostic weight no higher than 24%. The clinical history, family history and examination deserve more weight. This tool may be more appropriate for tracking symptoms over time than for initial diagnosis. The average lambda score in Table 4 is 0.76. If that metric is at all meaningful, then what it means is that patient self-report contributes 24% to diagnostic discrimination and 76% comes from other sources.
- Keywords
- Neuropsych Questionnaire; Self-reported Symptoms; Diagnosis
- StartPage
- 67
- EndPage
- 73
- Doi