Extended better than brief online CBT sessions for depression

Brief internet-based cognitive-behavior therapy (CBT) interventions are not as effective as extended sessions for the treatment of depression, say Australian researchers.

They add that longer programs are hindered by high dropout rates , but note that low continuation and high dropout do not necessarily denote a poor intervention.

Helen Christensen and colleagues from The Australian National University in Canberra, carried out an online randomized controlled trial in which 2794 individuals with elevated scores on the Goldberg Depression Scale received one of six versions of a CBT website.

The different versions were based on various components of brief CBT, extended CBT, behavior strategies, stress management, and problem solving.

In all, 20.4% of the participants completed the assigned intervention. Brief interventions delivered over the internet led to an immediate small-to-moderate reduction in depressive symptoms.

However, not all brief interventions were effective. A single module brief CBT was not sufficient by itself to reduce depressive symptoms, nor was the combination of brief CBT plus problem solving.

In contrast, extended CBT sessions proved to be effective for the treatment of depression, suggesting that these longer forms of CBT are an important component in reducing depressive symptoms.

The researchers note that most patients who completed baseline assessment failed to complete even one module of therapy.

"From a clinical perspective, this study suggests that internet users would benefit from tracking and monitoring," say Christensen et al.

They call for additional supportive resources to maximize outcomes for mental health individuals wanting to undertake full CBT programs on the internet.

"Methods currently being trailed in conjunction with web applications include the use of telephone calls and email reminders for diabetes and depression," the team reports.

"These appear to both increase adherence to the program and improve health outcomes."

Psychol Med 2006; 36: 1737-1746