Beck and Clark's 1997 information processing model of panic was presented a
nd evaluated. In general, studies using the Stroop task have shown panic pa
tients have a cognitive bias toward negative and personally relevant inform
ation. Several studies have also shown that panic patients tend to have mor
e catastrophic thoughts and are more likely to misinterpret bodily sensatio
ns. Further, cognitive-behavioral therapy is more effective than drug thera
pies. The limitations of the cognitive model and cognitive-behavioral thera
py were discussed, and an extension of current models was proposed. The ext
ension of current models combines cognitive-behavioral and psychodynamic co
ncepts and techniques and suggests that exposure in vivo used for agoraphob
ic situations and interoceptive exposure used for panic should also be appl
ied to the underlying themes of panic disorder.