Consultations Interactions Coding Scheme: Predictive validity with session-by-session outcome and sudden gains

Early interaction-style and –type can affect outcomes in cognitive behavioural therapy (CBT) for anxiety disorders. However, it is unclear specifically what interactions are helpful or hindering in CBT for severe health anxiety (SHA). This study aimed to identify interaction-styles and types that predict session-by-session outcomes and sudden gains in CBT for SHA. Each interaction in the first and second sessions of 49 recipients of CBT for SHA was categorised and rated on the basis of patient activation: an individual’s confidence and perceived ability to manage their own health. Multilevel modelling was used to assess early interactions’ predictive validity with session-by-session self-reported wellbeing. The 11 participants experiencing sudden gains had each interaction in the session directly prior to the gain similarly categorised and rated. The results were then compared to ratings of the preceding session. The greater the proportion of first and second sessions taken up with problem description interactions the smaller the improvement in wellbeing over treatment course. Furthermore, there was a significant reduction in problem description interactions at sessions directly prior to sudden gains, as compared to preceding control sessions. Problem descriptions are needed early in CBT for SHA, but they can reduce effectiveness if they dominate initial discussions.