(PS11-14) Association Between Quality of CBT Skills Practice Through a Smartphone-based Ecological Momentary Intervention and Short-term Change in Suicidal Thoughts or Related Negative Affect States

Abstract Text: Practicing emotion regulation skills outside of formal therapy sessions is a key tenet of CBT treatment. During periods of elevated suicide risk, such as the month after an inpatient psychiatric hospitalization, successfully engaging in skills practice may be crucial yet difficult. Smartphones have promise for providing immediate, accessible support during this critical care transition by delivering brief interventions like reminders to practice skills and guidance through them. Although a growing body of research focuses on frequency of engagement with smartphone-based mental health tools, little is understood about how the quality of engagement impacts efficacy. This project aims to explore whether the quality of CBT skills practice in a smartphone-based ecological momentary intervention (EMI) is associated with short-term changes in momentary suicidal thinking and negative affect states. 23 adults admitted for suicidal thoughts and behaviors to an inpatient psychiatric unit participated in an open trial of a brief, adjunctive CBT-based emotion management intervention. While in the hospital, participants received up to 3 brief sessions focused on 3 skills: Anchoring in the Present (AP), Changing Emotional Behaviors (CEB), and Thinking Flexibly (TF). For the month after discharge, participants received real-time skills practice exercises up to 3x/day on their smartphone. Participants were guided through skills practice over a series of screens, involving entering responses to key steps via open text fields (e.g., current thoughts, physical feelings, and behaviors for AP) and lists of response options (e.g., potential alternative actions to try for CEB). Before and after each EMI, participants rated from 0-10 their momentary levels of suicidal urge, suicide intent, agitation, and anger. A structured coding template was developed and used to rate the quality (completeness and relevance) of each skills practice on a 3-point scale ranging from 0 (missing or limited) to 2 (highly relevant and complete). For example, for TF, responses of identifying a negative thought and alternate interpretation were rated. All practices were double-coded and consensus dialogue was used to resolve discrepancies. In total, 945 skills practices were completed (N_AP=329, N_CEB=241, N_TF=375). Multi-level models examined whether quality ratings were associated with pre- to post-EMI change in momentary suicidal urge, suicide intent, agitation, and anger. Findings indicate that, when combining across all three skills, quality was not associated with proximal change in these states (ps >0.05). However, models exploring the three skills separately found that higher quality of CEB was associated with greater proximal reductions in suicidal intent (p=0.003). Higher quality of both TF and CEB was associated with greater reductions in anger (TF:p=0.006; CEB:p=0.011). These results suggest that the degree to which the quality of in-the-moment skills practice is associated with short-term skill effectiveness may differ as a function of both CBT skill and distress type. We must continue to better understand the nuances of engagement with smartphone-based interventions and CBT skills practice to bolster the precision and impact of these tools.