Diagnosis is a clinician-based determination involving the integration of data from a variety of sources, often including a clinical interview, behavioral observations, information from collateral sources, narrowband self-report measures (e.g., the Beck Anxiety Inventory; Beck & Steer, 1993), performance-based measures of personality (e.g., the Rorschach Inkblot Method), and broadband instruments such as the PAI, MCMI-III, and NEO-PI-3. Although broadband measures are only one of several sources of data at the clinician’s fingertips, they provide a rich basis of empirically derived information from which to develop diagnostic hypotheses.

When a client presents for treatment reporting symptoms of anxiety, a comprehensive evaluation of his or her personality and global levels of psychopathology can shed considerable light on many variables directly relevant to diagnosis, especially when the data obtained converges with information gathered from other sources. To illustrate, a client¹ of the author (Dr. Rogove) sought treatment for what he described as, “terrible anxiety.” As part of the initial evaluation, he was administered the PAI, MCMI-III, and NEO-PI-3. All of the validity scales were within normal limits. On the PAI, he obtained elevations on the Anxiety (ANX), Anxiety-Related Disorders (ARD), Depression (DEP), Stress (STR), and Nonsupport (NON) scales; on the MCMI-III, the Anxiety (A) and Major Depression (CC) scales were elevated along with a moderate elevation on the Compulsive (C) scale; and on the NEO-PI-3, he scored in the high range on the Neuroticism (N) and Con-scientiousness (C) domain scales, in the low range on the Agreeableness (A) domain scale, in the high range on the Angry Hostility (N2) and Depression (N3) facet scales, and in the very high range on the Anxiety (N1) and Vulnerability (N6) facet scales. No further scale elevations were obtained, including those that would have otherwise suggested the presence of a substance-related disorder, thought disorder, somatoform disorder, severe personality disorder, or heightened risk of suicide or violence.

In the absence of other information, the findings obtained from these instruments indicated several things about the client. First, the client’s scores on the PAI and the MCMI-III suggested that, at the time of the evaluation, he was experiencing considerable symptoms of anxiety and depression. The elevation that he obtained on the Stress (STR) scale of the PAI further indicated that recent stressors might have played a role in his condition, while the elevation on the Nonsupport (NON) scale suggested that he perceived little in the way of a social support network. Third, his profile on the NEO-PI-3 and the elevation he obtained on the Compulsiveness (C) scale of the MCMI-III further suggested that the issues that brought him in for treatment were superimposed on personality characterized by: (a) a general tendency to experience negative emotions (especially worry, fear, and other anxiety-related states), (b) a propensity for having difficulty managing stressful situations, (c) a generally disagreeable interpersonal style, and (d) obsessive–compulsive traits, including perfectionism, rigidity, and remarkable adherence to social conventions. These findings converged with data from other sources and provided empirical support for the author’s working hypothesis that the client had generalized anxiety disorder and major depressive disorder, and they resulted in empirically based hypotheses about variables maintaining his current condition and the best course of treatment. He was referred for a psychiatric consultation, a functional analysis was conducted, and cognitive–behavioral treatment was initiated to address his depression.

Treatment for anxiety is less likely to be effective when the condition(s) being treated is severe, and the client reports experiencing substantial distress (Newman et al., 2006). This is especially true in the case of social anxiety disorder (e.g., Otto, Pollack, Gould, Worthington, McArdle, & Rosenbaum, 2000; Safran, Alden, & Davidson, 1980; Scholing & Emmelkamp, 1999), PTSD (e.g., Ford, Fisher, & Larson, 1997; Johnson & Lubin, 1997; Perconte & Griger, 1991; Taylor, Fedoroff, Koch, Thordarson, Fecteau, & Nicki, 2001), PD (e.g., Emmelkamp & Kuipers, 1979), and OCD, though at least two studies have shown this to also be the case with GAD (e.g., Butler, 1993; Butler & Anastasiades, 1988). Baseline symptom severity also reliably predicts treatment outcome in child and adolescent samples undergoing anxiety treatment (e.g., Piacentini, Bergman, Jacobs, McCracken, & Kretchman, 2002; Silverman, Pina, & Viswesvaran, 2008).

As reported by Woody and Ollendick (2006), both Axis I and Axis II comorbidity also reliably predicts poorer treatment outcome in individuals being treated for anxiety. Further, personality disorder traits may well have differential effects on the treatment of some anxiety disorders (Scholing & Emmelkamp, 1999; cf., Newman et al., 2006). These findings are consistent with studies indicating that clients with more diagnoses and more severe conditions are at risk for profiting less from psychotherapy in general (for a review, see Clarkin & Levy, 2004). Low levels of symptom severity and distress may also impede therapeutic progress in certain clients, presumably because low levels of distress result in poor motivation and decreased responsibility to change (McCrae, Harwood, & Kelly, 2011; Sanderson & Clarkin, 2002).

An extensive history of research has identified moderators of the stress response that, if targeted, may buffer against the harmful psychological effects of overwhelming stress and have beneficial effects on the outcome of treatment (c.f., Feldman, Eisenberg, Gambini-Suarez, & Nassau, 2007; Lehrer, Woolfolk, & Sime, 2007; Quick, Quick, & Nelson, 1997; Smith, 2007). Moderators that may have value in the context of anxiety treatment include interpersonal conflict (Berman, Weems, Silverman, & Kurtines, 2000; Brent & Birmaher, 2006; Chambless & Steketee, 1999; Durham, Allan, & Hackett, 1997; McGrady, 2007; Tarrier, Sommerfield, Pilgrim, & Faragher, 2000), perceived social support (Newman et al., 2006), social problem solving (Chang, D’Zurilla, & Sanna, 2004; D’Zurilla & Nezu, 1999; Nezu, Wilkins, & Nezu, 2004), anger and hostility (Meichenbaum, 2007; Spielberger, 1991), cognitive appraisal (Lazarus, 1999; Meichenbaum, 2007; Pretzer & Beck, 2007), perception of control (Barlow, 2002), pessimistic attributional style (Barlow, 2002; Luten, Ralph, & Mineka, 1997), and, in children, parenting style (Barlow, 2002; Chorpita & Barlow, 1998). Thus, it behooves ­clinicians to evaluate a client’s level of distress, identify potential moderators of the stress response, and adjust interventions accordingly.