Predicting outcomes from cognitive behavioural therapy for social anxiety disorder: A Bayesian network analysis

Area of Science:

• Psychiatry
• Clinical Psychology
• Data Science

Background:

• Predicting treatment response in social anxiety disorder (SAD) is crucial for optimizing cognitive behavioural therapy (CBT).
• Identifying specific pre-treatment symptom predictors can enhance therapeutic outcomes for SAD patients.

Purpose of the Study:

• To identify pre-treatment social anxiety symptoms associated with post-treatment outcomes in patients receiving CBT for SAD.
• To develop a predictive model for treatment response based on symptom profiles.

Main Methods:

• A retrospective cohort study of 1315 patients with SAD treated with CBT.
• A Bayesian network model was developed using a training sample (n=658) and validated on a test sample (n=657).
• Variable selection utilized 10-fold cross-validation, and predictive accuracy was assessed using the area under the receiver operating characteristic curve (AUC).

Main Results:

• The predictive model showed moderate out-of-sample accuracy (AUC = 0.58), with a 6.76% shrinkage from training performance (AUC = 0.67).
• Four key interrelated SAD symptoms emerged as significant predictors: avoiding public speaking, fear of embarrassment, avoiding criticism, and fear of being observed.
• These identified symptoms were measured using the Social Phobia Inventory (SPIN).

Conclusions:

• Specific social anxiety symptoms identified at baseline are associated with treatment response to CBT.
• Targeting these key symptoms during CBT assessment and intervention may improve treatment efficiency and effectiveness for social anxiety disorder.
• The findings support a more personalized approach to CBT for SAD based on symptom profiles.