Online CBT for Panic Disorder: Does It Work as Well as In-Person Treatment?

Panic Disorder Is a Good Test Case

Panic disorder is treated with well-established CBT protocols -- interoceptive exposure, breathing retraining, cognitive restructuring around catastrophic misinterpretation of physical sensations. Because the treatment itself is so standardized, it's a clean test of whether the delivery format (in-person individual, in-person group, or internet-delivered) changes the outcome. Two separate lines of research have tested exactly this.

The Network Meta-Analysis: 74 Trials, One Important Split

A 2023 systematic review and network meta-analysis in Psychological Medicine pooled 74 randomized controlled trials with 6,699 participants comparing CBT delivery formats for panic disorder. The finding: face-to-face group therapy, face-to-face individual therapy, and guided self-help (a therapist actively involved, even if remotely) were all superior to treatment-as-usual and did not differ meaningfully from each other in effectiveness. The critical split in the data: unguided self-help -- programs with no real therapist involvement -- was not found to be significantly effective. Discontinuation rates did not differ significantly across the effective formats. In plain terms: the format doesn't matter much, as long as a real therapist is actually involved in the process.

A Direct Head-to-Head Trial

A randomized trial published in BMC Psychiatry put this to a direct test: 113 patients with panic disorder in psychiatric care were randomized to either internet-delivered CBT or in-person group CBT. Both groups improved substantially (Cohen's d = 1.73 for internet, d = 1.63 for group, on the Panic Disorder Severity Scale), and the between-group difference at the end of treatment was d = 0.00 (95% CI: -0.41 to 0.41) -- no statistically significant difference, using a mixed-models analysis. One notable secondary finding: the internet-delivered version required an average of 35.4 minutes of therapist time per patient, compared with roughly 6 hours for the group format, without a corresponding drop in outcome.

The Distinction That Actually Matters

Taken together, these two sources point to the same conclusion, and it isn't simply "online works as well as in-person." It's more specific: guided online CBT -- where a licensed therapist is actively part of the process -- performs comparably to face-to-face care for panic disorder, while unsupervised self-help programs do not have the same evidence behind them. The format (screen vs. room) matters far less than whether a real therapist remains involved.

Practical Takeaway

• For panic disorder, therapist-guided internet CBT and in-person CBT (individual or group) show comparable effectiveness across a large pooled evidence base
• Purely self-directed, unguided programs do not have the same evidence of effectiveness -- the guided element is what the data actually supports, not "online" by itself
• A direct trial found internet-delivered CBT needed far less therapist time than group therapy for an equivalent outcome, which is part of why guided online programs can be offered at lower cost
• When evaluating any online CBT program for panic disorder, the presence of ongoing licensed-therapist involvement is the feature the evidence actually points to, not the platform itself