Both lively supervision and taped supervision for CCT were held weekly. Two trained psychiatrists who had no other contact with participants evaluated all assessments. The inter-rater reliability was high enough for the study (r > 0.95, P < 0.001). OCD symptoms were recorded using the Y-BOCS Symptom
Checklist (Y-BOCS-SC; Goodman et al. 1989). Statistical analysis Analysis of covariance (ANCOVA) with repeated measured and baseline data Inhibitors,research,lifescience,medical control and Tukey Honest Significant Difference (HSD) post hoc were performed to test the effects of treatment, time, and interaction on OCD Y-BOCS-SR score and GAF score, using the SAS (ver 9.1)’. The Tukey HSD is the most widely used post hoc test in psychological and the behavioral sciences (http://www.une.edu.au/WebStat/unit_materials/c7_anova/oneway_post_hoc.htm). Chi-square was performed to analyze the response rate and clinical remission rate. When this requirement was not met for a 2 × 2 table, a Fisher Exact Probability Test was performed. Linear regression analyses were performed to
measure Inhibitors,research,lifescience,medical the correlation between the reduction in Y-BOCS-SR score rated by psychiatrists and the improvement of the OCD symptoms rated by Inhibitors,research,lifescience,medical patient clinical trial self-report to confirm the accuracy of the rating. Multiple linear regression was performed to investigate correlative factors that influenced the efficacy of treatment. An intent-to-treat (ITT) analysis using the last observation carried forward (LOCF) was conducted to examine all participants who received treatment for any Inhibitors,research,lifescience,medical time period. Results Evaluation reliability on the severity of OCD symptoms Reliability was detected using linear regressive analysis between the variables of reduction percentage of Y-BOCS-SR score and the reduction percentage of OCD symptoms at the four time-points after treatment. The correlation coefficients were greater than Inhibitors,research,lifescience,medical 0.98 (P < 0.001) (Fig. 3). Figure 3 Reliability analysis on the severity of OCD symptoms. Changes in the severity of OCD symptoms ANCOVA analysis showed that the reduction in Y-BOCS-SR total scores was significantly greater overall, dependant
on the treatment and treatment period (F = Mephenoxalone 3.66, df1 = 14, df2 = 328, P < 0.001). The repeated measures analysis of variance showed that the interaction of treatment and treatment period resulted in a reduction in Y-BOCS-SR total scores (P < 0.001). The ANOVA post hoc tests showed that the Y-BOCS-SR scores were not different among the three groups (P > 0.05). Compared with baseline, the Y-BOCS-SR score was significantly reduced only in the PCCT group at month 1 (P < 0.001). The score was significantly reduced in the PCCT group (P < 0.001) and the PCBT group (P < 0.05) at months 3, 6, and 12. In the pharmacotherapy group, there was a trend of significant reduction at month 6 (P = 0.059) and a significant reduction at month 12 (P < 0.05), but no differences were found at months 1 and 3 (P > 0.05) (Table 2).