A follow-up of two panic disorder-agoraphobic study populations. The role of recruitment biases.

A self-referred and recruited panic disorder-agoraphobic study population are compared in terms of their symptomatic presentation and treatment outcome. The recruited group was found to be more symptomatically severe, more phobic, less character disordered, and more likely to stay in treatment. Implications for the potential methodological biases in recruiting study populations are explored.