(PS11-66) Intensity of Suicidal Thoughts, but Not Negative Affect, Is Negatively Associated with Skills Practice Quality in Emotion-focused Ecological Momentary Intervention

Cognitive Behavioral Therapy (CBT) is efficacious in the treatment of a variety of emotional disorders. Its efficacy is presumed to depend, in part, on the regular practice of skills outside of therapy sessions. Smartphone apps and related mHealth tools have great promise in helping to bolster outside-of-session skills use. Many such tools include sending prompts or notifications that guide users through skills practice exercises; however, little is known about the specific moments when individuals are most likely to engage – and importantly, engage effectively – in smartphone-based skills practice. Identifying the contexts when prompting the use of CBT skills is most likely to result in effective practice is critical to the development of just-in-time interventions.

We leverage data from a recent study of an ecological momentary intervention (EMI) that aims to promote in-the-moment CBT emotion management skills use after psychiatric inpatient hospitalization for suicide risk to examine the associations between affective context and skills practice quality. A total of 23 adults received up to three brief individual CBT sessions (focused on mindful emotion awareness, cognitive flexibility, and changing emotion-driven behaviors) while in the hospital. Then, participants received up to three randomly timed prompts to complete guided in-app exercises practicing each skill learned during the four weeks after discharge. Each exercise began with a brief ecological momentary assessment (EMA) survey of negative affect states and momentary suicidal thoughts. We developed a structured coding framework to evaluate the quality (i.e., completeness and relevance) of each completed exercise (945 total) on a scale of 0 (missing or limited) to 2 (highly relevant and complete). For example, responses to Changing Emotional Behaviors exercises were rated for whether the participant identified at least one unhelpful behavior tied to a strong emotion and then, an adaptive alternative action. All skills practices were double-coded and discrepancies resolved by consensus.

Descriptive analyses showed significant within-person variance in skills practice quality (ICC = .245). Multi-level models were used to test the associations between the intensity of three specific momentary negative affect states (agitation, fatigue, hopelessness) and suicidal thoughts (suicide urge, suicide intent) captured with EMA immediately prior to each skills practice and quality of the time-matched skills practice exercise. Greater intensity of suicide urge and intent were both associated with poorer skills practice quality (p < .05). Intensity of negative affect, was not associated with skills practice quality. These findings, although preliminary, suggest that it may be more difficult to engage in high-quality app-based CBT emotion management skills practice when suicidal thoughts are more intense, whereas quality of skills practice may be less sensitive to changing intensity of negative affect states. Results underscore the importance of working toward CBT-based just-in-time interventions that flexibly adapt to individuals’ specific internal and external contexts.