اثر بخشی درمان فردی دستی و کتاب محور برای اختلال اضطراب اجتماعی

Abstract Social anxiety disorder is a prevalent and impairing disorder for which viable cognitive-behavioral therapies exist. However, these treatments have not been easily packaged for dissemination and may be underutilized as a result. The current study reports on the findings of a randomized controlled trial of a manualized and workbook-driven individual cognitive-behavioral treatment for social anxiety disorder (Hope, Heimberg, Juster, & Turk, 2000; Hope, Heimberg, & Turk, 2006). This treatment package was derived from an empirically supported group treatment for social anxiety disorder and intended for broad dissemination, but it has not previously been subjected to empirical examination on its own. As a first step in that examination, 38 clients seeking treatment for social anxiety disorder at either the Adult Anxiety Clinic of Temple University or the Anxiety Disorders Clinic of the University of Nebraska-Lincoln were randomly assigned to receive either immediate treatment with this cognitive-behavioral treatment package or treatment delayed for 20 weeks. Evaluation at the posttreatment/postdelay period revealed substantially greater improvements among immediate treatment clients on interviewer-rated and self-report measures of social anxiety and impairment. Three-month follow-up assessment revealed maintenance of gains. Clinical implications and directions for future research are discussed.

Results Attrition Of the 22 clients assigned to the delayed treatment condition, 3 dropped during the wait period and did not provide postdelay data. Of the 16 clients assigned to the immediate treatment condition, all but 1 completed treatment and provided posttreatment data. Both intent-to-treat (ITT) and completer analyses were performed. For the ITT analyses, participants’ last observation was carried forward. For participants in the delayed treatment condition, this meant carrying forward their pretreatment evaluation to the postdelay time point. For participants in the immediate treatment condition, this meant carrying forward the pretreatment evaluation to the posttreatment time point. Completer and ITT analyses yielded virtually identical findings. Therefore, only the completer analyses are reported here. ITT analyses are available on request. Between-Group Analyses There were no significant differences between the IT and DT groups on any measure at the baseline assessment. Main hypotheses were evaluated using univariate analysis of covariance with treatment condition as the independent variable (immediate treatment versus delayed treatment) and the posttreatment or postdelay measure of interest as the dependent variable. The baseline score (pretreatment/predelay) on the measure of interest was used as a covariate. These analyses were first conducted with site (Temple vs. Nebraska) as an additional independent variable. No main effects of site and only one site-by-treatment-condition interaction was significant. Across analyses, inclusion of site effects resulted in larger main effects for condition. Therefore, we report (with the one exception) the more straightforward analyses without site. The site-by-treatment-condition analyses are available on request. Interviewer-Rated Measures Table 1 displays means, standard deviations, and effect sizes for all interviewer-rated and self-report measures. At the 20-week assessment, participants in the IT condition received significantly lower ADIS CSR ratings than participants in the DT condition, F(1, 32) = 13.29, p < .001. This main effect was qualified by the one significant site-by-treatment interaction, F (1, 32) = 6.27, p < .02. Posttreatment/postdelay means adjusted for baseline demonstrated a tendency for greater improvement on this rating at the Nebraska site. However, the 95% confidence intervals around these means overlapped, suggesting that specific pairwise comparisons were not significant. Clients who completed the IT condition received significantly lower scores on the LSAS at the 20-week assessment than clients who completed the DT condition. Of the 31 clients for whom the CGI-I was completed, 11/15 (73.3%) participants in the IT condition were classified as responders on the CGI-I, compared to 1 (6.3%) client of 16 in the DT condition, χ2(1, N = 31) = 14.69, p < .001. Table 1. Pretreatment scores and posttreatment/postdelay scores on study measures Measure Estimated Marginal Mean at Pretreatment (nonadjusted mean and standard deviation in parentheses) Treatment condition Posttreatment or Postdelay F Cohen’s d Interviewer Rated Measures ADIS Clinician Severity Rating for social anxiety disorder 5.61 (5.61, SD = 0.95) Immediate Delayed 3.92 (SE = 0.29) 5.24 (SE = 0.27) 10.90⁎⁎ 1.21 Liebowitz Social Anxiety Scale 67.71 (68.19, SD = 22.95) Immediate Delayed 46.81 (SE = 5.23) 62.98 (SE = 4.71) 4.86⁎ 0.83 Self-Report Measures: Social Anxiety Brief Fear of Negative Evaluation Scale (8-item) 32.52 (32.47, SD = 6.42) Immediate Delayed 21.30 (SE = 1.47) 31.23 (SE = 1.33) 23.80⁎⁎⁎ 1.84 Social Phobia Scale 32.90 (32.89, SD = 15.29) Immediate Delayed 14.64 (SE = 2.52) 32.71 (SE = 2.29) 27.70⁎⁎⁎ 1.99 Social Interaction Anxiety Scale 42.70 (43.16, SD = 12.60) Immediate Delayed 27.63 (SE = 1.77) 45.89 (SE = 1.65) 55.94⁎⁎⁎ 2.88 Self-Report Measures: Disability and Quality of Life Sheehan Disability Scale 14.26 (14.11, SD = 7.06) Immediate Delayed 8.66 (SE = 1.10) 14.26 (SE = 1.06) 13.17⁎⁎ 1.37 Quality of Life Inventory 0.53 (0.34, SD = 1.67) Immediate Delayed 1.19 (SE = 0.33) 0.63 (SE = 0.33) 1.50 0.51 Note. ADIS = Anxiety Disorders Interview Schedule; SD = Standard Deviation; SE = Standard Error of Measurement. F-test is for the main effect for condition (immediate vs. delayed treatment) in a one-way analysis of covariance. See text for report of the significant site-by-treatment-condition interaction in preliminary analysis for the ADIS Clinician’s Severity Rating. Cohen’s d is calculated as the differences in means between clients receiving immediate and delayed treatment divided by the pooled SD; a d of 0.20 is considered small, 0.50 is considered moderate; and 0.80 is considered large ( Cohen, 1988). ⁎ p < .05. ⁎⁎ p < .01. ⁎⁎⁎ p < .001. Table options At posttreatment assessments, ADIS modules for which clients met diagnosis at pretreatment were readministered and CSR ratings were made. Of the 11 clients who met criteria for at least one additional Axis I disorder at baseline and who completed the posttreatment/postdelay assessment, 4 (36.36%) no longer met criteria for their additional diagnoses by the latter assessment, despite the fact that therapy only targeted the symptoms of social anxiety disorder. Three of these 4 clients had an additional diagnosis of generalized anxiety disorder; 1 of these clients also had co-occurring dysthymia. The other client had a specific phobia. Interestingly, the client with both generalized anxiety disorder and dysthymia lost both of these diagnoses by the end of treatment. Self-Report Measures Three social anxiety self-report measures were administered in the study: the BFNE, the SIAS, and the SPS. At the 20-week assessment, clients in the IT condition scored significantly lower on all of these measures than clients who completed the DT condition (see Table 1). Two measures were administered to assess for disorder-related impairment, the SDS and the QOLI. At the 20-week assessment, clients who were assigned to the IT condition reported significantly less disability (SDS) than clients who completed the DT condition (see Table 1). Clients who completed the IT condition also reported slightly better quality of life than those in the DT condition; this trend resulted in a moderate effect size (all other analyses reported above resulted in large effect sizes per the criteria outlined by Cohen, 1988; see Table 1), but did not reach the level of statistical significance. Within-Group Analyses Paired-sample t-tests were conducted separately for IT and DT participants to examine the degree to which they demonstrated significant within-group change. The IT group demonstrated significant pretreatment to posttreatment change on six of the seven outcome variables, the lone exception being the QOLI. Within-group effect sizes (d) ranged from 1.03 to 5.22, with a median of 2.52 and a mean of 2.98. The DT group demonstrated significant change on only one measure, the SIAS, and that test revealed a significant increase in social interaction anxiety. Five of seven effect sizes were small (< 0.36), and three of seven denoted modest change in the direction of increased anxiety and impairment. Details of these analyses are available from the authors. Follow-up Analyses Clients underwent another assessment 3 months posttreatment. Only 12 clients (35% of the completer sample) completed the follow-up independent evaluation. Clients who completed the follow-up evaluation were compared to clients who did not complete the evaluation on pre- and posttreatment ADIS clinician’s severity ratings and on pre- and posttreatment LSAS scores. The two groups did not differ significantly on any of these measures. A greater proportion of clients who completed the follow-up evaluation had originally been assigned to the IT group (8/12, 66.6%). However, these analyses are performed on clients who had completed treatment, either immediate or delayed, who were combined for these analyses. Because some of these clients did not return self-report measures at follow-up, the present analyses focus only on the ADIS clinician’s severity rating and the LSAS (interviewer-rated measures). Paired sample t-tests were run comparing scores at pretreatment and follow-up. For both the ADIS rating and the LSAS, clients showed significant improvements from pretreatment to follow-up (both ps < .001). However, t-tests comparing posttreatment to follow-up scores were not significant, suggesting maintenance of gains but no further change during the follow-up interval.