Predicting outcomes and sudden gains from initial in-session interactions during remote cognitive-behavioural therapy for severe health anxiety.

There has been a dramatic increase in remote psychotherapy since the onset of the COVID-19 crisis. There is also expected to be an increase in mental health problems in the wake of the COVID-19 pandemic. An increase in severe health anxiety (SHA) is particularly anticipated, for which cognitive-behavioural therapy (CBT) is a frontline treatment. However, it is unclear what interaction types are associated with outcome-improvement in remote-CBT (rCBT) for SHA. This study aimed to identify interaction types that predict outcomes and sudden gains in rCBT for SHA using initial therapy session content. Forty-eight participants in rCBT for SHA had interactions at their first sessions categorized and rated in terms of patient activation: an individual's confidence and ability to manage their health. Multilevel modelling assessed whether early interaction types predicted session-by-session wellbeing. For participants experiencing sudden gains (n = 12) interactions at the session directly prior to the gain were similarly categorized and rated. The scores were then compared with ratings for the preceding session. A smaller proportion of early sessions was taken up with problem descriptions among those with greater outcome improvements. There was also a significant reduction in the proportion of the session spent describing problems in the session directly prior to a sudden gain, as compared with the previous session. Conversely, clients with better outcomes made more positive evaluations of themselves and therapy, noticed more positive changes and made more contributions to structuring interactions at initial sessions. Specific early interaction types predict session-by-session outcomes and precede sudden gains in rCBT for SHA.