Background

Many people with social anxiety disorder (SAD) feel ashamed or stigmatized, which stops them from seeking help. Traditional therapies can be expensive and time-consuming. Therefore, we need quick, affordable treatments that can reach more people. It’s important to find out which patients will benefit most from these treatments.

Objective

This study aimed to use advanced methods to identify which patients with SAD would benefit from a 14-day self-guided mindfulness program compared to a self-monitoring app.

Methods

We studied 191 participants with SAD. They were randomly assigned to either the mindfulness program or the self-monitoring app. Participants reported their symptoms and personal information at the start, after treatment, and one month later. We used machine learning to analyze 17 factors that could predict who would benefit more from the mindfulness program.

Results

The machine learning models performed better than traditional methods. The best models identified key factors that indicated which participants were likely to improve after using the mindfulness program. Important factors included:

Higher mindfulness skills
Lower severity of SAD
Having a university education
Not using psychotropic medications
Higher generalized anxiety severity
Having a diagnosed depression or anxiety disorder
Being of Chinese ethnicity

These factors helped predict who would benefit from the mindfulness program.

Conclusions

The findings suggest that we can better identify patients who will respond well to scalable treatments for SAD. By focusing on individual strengths and weaknesses, we can improve treatment selection. Using a “prescriptive predictor calculator” could help clinicians decide the best treatment for each patient. Those likely to benefit from the mindfulness program could use it while waiting for more intensive therapy.

Next Steps

To implement these findings in clinics:

Define Measurable Outcomes: Set clear goals for using the mindfulness program.
Select AI Tools: Choose AI solutions that fit specific clinical needs.
Implement Step by Step: Start with a pilot project and track results using AI solutions.