قابلیت اطمینان و اعتبار ویژگی های ابعادی اختلال اضطراب فراگیر

The reliability and validity of the dimensional features of generalized anxiety disorder (GAD) were examined in a diverse sample of 508 outpatients with anxiety and mood disorders who underwent two independent administrations of the Anxiety Disorders Interview Schedule for DSM-IV: Lifetime version (ADIS-IV-L; Di Nardo, Brown, & Barlow, 1994). Inter-rater reliability was higher in the full sample than in patients with current GAD. Additionally, the presence of a mood disorder weakened inter-rater reliability. We also explored the unique contribution of excessiveness and uncontrollability of worry to various clinical outcomes and found that excessiveness predicted anxiety, depression, and stress self-report measures, and uncontrollability predicted clinical severity and number of diagnoses. Findings are discussed with regard to their implications for the classification of GAD (e.g., utility of dimension-based assessment to improve the classification of psychological disorders).

Classification of anxiety and mood disorders has been an inexact science, reflected by the modest reliability of many diagnostic categories (e.g., Brown, Di Nardo, Lehman, & Campbell, 2001). Generalized anxiety disorder (GAD) is a diagnostic category that has undergone substantial revisions (Brown, Barlow, & Liebowitz, 1994). GAD is classified in the 5th edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) ( American Psychiatric Association, 2013) as chronic (lasting at least six months), excessive anxiety and worry about a number of events or activities that is difficult to control, and is associated with at least three of six symptoms of tension/negative affect with some present more days than not for at least six months. For GAD to be assigned, the worries and associated symptoms must cause clinically significant distress or impairment in social, occupational, or other important areas of functioning. Additionally, a diagnosis of GAD requires that the anxiety and worry do not occur exclusively during the course of a mood or psychotic disorder. When GAD first appeared in DSM-III ( American Psychiatric Association, 1980), it was a residual category, diagnosed only if a patient did not meet criteria for any other anxiety or mood disorder. This definition was associated with low inter-rater reliability (κ = .47; Di Nardo, O’Brien, Barlow, Waddell, & Blanchard, 1983). The reformulation of GAD in DSM-III-R ( American Psychiatric Association, 1987) failed to considerably improve the reliability of the disorder, as was shown by large-scale studies entailing administration of two independent structured interviews (κs for current GAD were .27 in Mannuzza, Fyer, Martin, & Gallops, 1989; .53 in Di Nardo, Moras, Barlow, & Rapee, 1993; and .56 in Williams et al., 1992). Evidence of low reliability and high comorbidity of GAD with other disorders (comorbidity rates exceeding 80%; see Brown & Barlow, 1992) led researchers to question whether there was sufficient discriminant validity to retain GAD as a diagnostic category in DSM-IV ( Brown et al., 1994). The diagnostic criteria were revised substantially in DSM-IV ( American Psychiatric Association, 1994) in an effort to define the boundary of GAD with other anxiety disorders, mood disorders, adjustment disorders, and nonpathological worry. Revisions to DSM-IV included the requirement that worry must be perceived as uncontrollable (based on evidence that uncontrollability of worry distinguishes GAD worry from normal worry; Abel and Borkovec, 1995 and Borkovec and Roemer, 1994). Another substantial change in DSM-IV was the reduction of associated symptoms from 18 to 6. Symptoms of autonomic arousal (e.g., accelerated heart rate, shortness of breath) were eliminated, while symptoms of tension and negative affect (e.g., muscle tension, irritability) were retained. Although this change was partly data-driven (e.g., Brown, Marten, & Barlow, 1995), researchers were concerned that this revision would further obfuscate the boundary between GAD and mood disorders ( Clark & Watson, 1991). Nonetheless, the revisions to DSM-IV GAD diagnostic criteria were associated with increased diagnostic reliability (κ = .67 in Brown et al., 2001), compared to DSM-III-R (κ = .53 in Di Nardo et al., 1993). In addition to examining the diagnostic reliability of the various DSM-IV anxiety and mood disorders, Brown et al. (2001) evaluated the factors most commonly involved in diagnostic disagreements; e.g., difference in patient report, threshold disagreements (e.g., difficulties applying cutoffs for presence or absence of a disorder based on sufficient distress or impairment), change in clinical status, interviewer error, and diagnosis subsumed under comorbid condition. With regard to the diagnosis of GAD, difference in patient report was the most common source of disagreement (55%). A reliable diagnosis of GAD calls for consistent self-report of many subjective features, their onset, and their duration in relation to other conditions (e.g., mood disorders). Inconsistency in patient reports could be indicative of vagueness of these diagnostic features and patients’ difficulty distinguishing them from other disorders ( Brown et al., 2001). This study also found that GAD diagnostic disagreements involved mood disorders in 47% of cases, which is consistent with prior evidence that boundary issues with mood disorders pose a larger problem for GAD than do other anxiety disorders (e.g., Brown, Chorpita, & Barlow, 1998). In addition to work on reliability at the diagnostic level, researchers have begun to explore the reliability of the dimensional features of GAD. Gordon and Heimberg (2011) examined the reliability of GAD features in a sample of 129 patients with a principal diagnosis of GAD. GAD features were assessed using the Anxiety Disorders Interview Schedule for DSM-IV: Lifetime version (ADIS-IV-L; Di Nardo, Brown, & Barlow, 1994). As estimated by intraclass correlations (ICCs), the dimensions of excessiveness of worry, uncontrollability of worry, and interference due to worry were found to have fair to good inter-rater agreement (i.e., ICCs = .60, .59, and .62, respectively). Agreement was poor for distress due to worry (ICC = .30). Agreement for the associated symptoms ranged from poor to good (range of ICCs = .22–.65), and varied by symptom (i.e., good for fatigue, fair for irritability, muscle tension, sleep disturbance, and concentration difficulties, and poor for restlessness) ( Gordon & Heimberg, 2011).