(PS8-D81) Implementation in Action: An Evaluation of the Association Between Therapist Initiated App-based Messages and Engagement in a Behavioral Telehealth Program with an Adjunctive App for Depression

Background: Although effective psychotherapy interventions for depression exist, engagement continues to be a challenge with dropout rates as high as 19-47% (Berryhill et al., 2019; Pentaraki, 2018). Technology, including telehealth and digital tools, has the potential to improve engagement in evidence-based care by eliminating barriers to treatment and enhancing connection between the therapist and patient. The introduction of adjunctive digital tools into care may create new pathways for communication, yet little is known about the impact of therapist messaging outside of session on patient engagement in real-world settings. The purpose of this study was to evaluate: 1) the association between therapist initiated messages and participant digital app utilization and 2) the association between program completion and the number of therapist initiated messages among participants in a behavioral telehealth program for depression with an adjunctive digital application. 

Methods: We analyzed previously collected de-identified data from a consecutive cohort of 179 participants (70.4% female, 55.3% non-white race/ethnicity, mean age 36.7) who initiated treatment in a protocolized, ~8-session, cognitive behavioral teletherapy intervention for depression led by a licensed therapist and augmented with an app designed to support homework completion, skill acquisition, and client-therapist communication. As part of this intervention, therapists could send participants a brief (~4-5 sentences) personalized message between sessions to reinforce program engagement, encourage skill practice, and build rapport. Two sample t-tests were utilized to examine differences in the number of activities completed among participants who received at least one message and those who did not, as well as differences in the number of messages sent between program completers and non-completers. Likelihood of program completion among participants who received at least one message versus those who did not was assessed using chi-square statistics.

Results: Over the course of the program, therapists sent approximately 3.6 (SD=2.8) messages and participants completed 14.0 (SD=14.6) digital activities and 6.8 (SD=3.0) therapy sessions. Overall, 58.7% (n=105) participants completed the full program. Participants who had at least one therapist message sent to them completed significantly more activities per session (mean=2.2; SD=1.9) than participants who had no messages (mean=0.9; SD=1.2; t=-3.5; p=.0006). The number of messages sent per session was significantly higher among program completers (mean=0.6; SD=0.3) than non-completers (mean=0.4; SD=0.4; t=-3.1; p=.003) and participants who had at least one therapist message sent to them had a 12 times greater likelihood of completing the program than participants who had no messages (OR=12.1; 95% CI [4.0, 36.8]; 𝝌2=27.0; p< .0001). 

Conclusions: These data suggest that brief personalized therapist messages sent through an adjunctive digital app may be associated with greater intervention engagement. Next research steps include understanding the factors that inform therapists’ decisions to send messages and whether message content moderates the observed associations.