Using Text Messaging to Deliver Positive Affect Skills: Examining Naturalistic Rate and Predictors of Engagement in Suicidal Adolescents

Shirley Yen, Ph.D., Katherine Tezanos, B.A., Adam Chuong, B.A., Megan Ranney, M.D., Joel Solomon, M.D., Christopher Kahler, Ph.D., & Anthony Spirito, Ph.D. 

Introduction Delivery of therapeutic content via SMS has the potential to reach populations that experience barriers to in-person treatment. This is particularly true for teens, as they report high usage of text messaging. SMS interventions have been developed to target a number of medical and behavioral health outcomes in teens, such as diabetes management1, skin cancer prevention2, substance use3-5, asthma6, and weight management7. Fewer programs have targeted adolescents with psychiatric/emotional difficulties. 

The Skills to Enhance Positivity Program (STEP) is an adjunctive intervention that combines in-person positive affect skills sessions delivered to suicidal adolescents during inpatient hospitalization, with a one-month remote delivery which includes daily text messages and weekly phone calls. The purposes of the daily text messages are two-fold: 1) emotional mood monitoring; and 2) reminder of skills to increase positive affect. The overall objective of STEP is to increase attention to positivity, which should decrease suicidal ideation. 

Despite the recent proliferation of interventions delivered via SMS, data on whether therapeutic messages are read are lacking, and few studies have examined naturalistic response rate to text messages. The present study seeks to provide such data in a clinical sample of adolescents hospitalized for suicide risk. Specifically, we examine naturalistic response rate to mood monitoring questionnaires, predictors of responsiveness, and qualitative feedback about the text messaging component of the intervention. 

Methods Twelve participants have completed the open phase trial of the STEP program, an adjunctive intervention that combines in-person positive affect skills sessions delivered to suicidal adolescents during inpatient hospitalization, with a one-month remote delivery which includes daily text messages and weekly phone calls. Text messaging begins the day after discharge, delivered daily at a time of participants’ choosing. If they are re-hospitalized during the 1 month phase, text messages are suspended until discharge. Thus, all participants receive at least 30 days of messages. There are 6 mood monitoring questions, rated on a Likert scale from 1 to 5 (1 – “not at all,” 5 – “extremely”): 

How content, serene, peaceful, do you feel RIGHT NOW? 

How angry, irritated, annoyed, do you feel RIGHT NOW? 

How glad, happy, joyful, do you feel RIGHT NOW? 

How scared, fearful, afraid, do you feel RIGHT NOW? 

How grateful, appreciative, thankful, do you feel RIGHT NOW? 

How sad, downhearted, unhappy, do you feel RIGHT NOW? 

Responsiveness is operationalized by daily response rate to mood monitoring questions. There is also a question to select a skill from a category of their choice (mindfulness, gratitude, savoring); a skill (e.g. “write 3 good things in your journal”) will be delivered accordingly. Participants were not compensated for any responses to text messages; thus data reflects naturalistic response tendencies. Demographic and clinical variables of interest were assessed in baseline interviews and extracted from medical charts. 

Participant attitudes towards text messaging


"Send text messages longer than a month to make sure the practice of positive emotions becomes habitual“ 

"I believe this study helped me a great deal and taught me how to just let things go and try to not let things stress me out. Thank you!“


"It may be helpful for parents to receive sample text and for the texts to slowly decrease rather than abruptly end.“ 

"The daily text-messages kept my child focused on his recovery and he knew he had an outside source available to him, if he needed it“ 

"I think the text messages should continue through 3-months post hospitalization. Some days that may be the only support the child sees and it's VERY important for her to be continually reminded she’s supported and not alone." 

Results At present, we have administered the intervention to 12 of the targeted 20 participants in an open pilot trial; more data are forthcoming. Preliminary results indicate that on average, participants responded on 72.37% of days (range: 6.66 -100). While the sample size is small, some differences in response rates have emerged. Participants with less educated parents had higher response rates vs. those with college educated parents (t=2.383, p=.04). Furthermore, the vast majority of participants reported that the text messaging was either somewhat or a great deal helpful. Most participants reported that the frequency of text messages and duration of the program were appropriate. 

Summary and Conclusions Clinical characteristics at baseline (depression, hopelessness, past suicide attempts, treatment utilization) did not predict response rate. Interestingly, two participants who have not been responsive to phone outreach have been highly responsive (95-100%) to text messaging. This demonstrates that SMS may be acceptable and preferable to telephone outreach for some participants, and acceptable to those from varying socioeconomic backgrounds. This data is preliminary and sample size is small but demonstrates high acceptability in a high-risk difficult-to-reach clinical population.