Reliability of the 35% carbon dioxide panic provocation challenge

The objective of this study was to determine the test-retest reliability of the 35% carbon dioxide (CO2) panic provocation challenge. Thirty patients with panic disorder were included in this study. Twenty-four patients were challenged twice, with 1 week between the two challenges. Six patients dropped out after the first test. The 35% CO2 challenge appeared to have a good test-retest reliability; both on induced subjective anxiety, measured on a Visual Analogue Scale for Anxiety (VAS-A), and induced panic symptoms, measured with a Panic Symptom List. Assessing the state of anxiety immediately after the challenge gave the most reliable results. Calculating increase in anxiety from the pre- and post-scores on the VAS-A rendered less reliable scores. This study completes a series of studies in which the criteria for an ideal model of panic are tested for the 35% CO2 challenge. Apart from an absolute specificity for panic disorder, the challenge meets these criteria.