This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Formative Research, is properly cited. The complete bibliographic information, a link to the original publication on http://formative.jmir.org, as well as this copyright and license information must be included.
New opportunities to create and evaluate population-based selective prevention programs for suicidal behavior are emerging in health care settings. Standard depression severity measures recorded in electronic medical records (EMRs) can be used to identify patients at risk for suicide and suicide attempt, and promising interventions for reducing the risk of suicide attempt in at-risk populations can be adapted for web-based delivery in health care.
This study aims to evaluate a pilot of a psychoeducational program, focused on developing emotion regulation techniques via a web-based dialectical behavior therapy (DBT) skills site, including four DBT skills, and supported by secure message coaching, including elements of caring messages.
Patients were eligible based on the EMR-documented responses to the Patient Health Questionnaire indicating suicidal thoughts. We measured feasibility via the proportion of invitees who opened program invitations, visited the web-based consent form page, and consented; acceptability via qualitative feedback from participants about the DBT program; and engagement via the proportion of invitees who began DBT skills as well as the number of website visits for DBT skills and the degree of site engagement.
A total of 60 patients were invited to participate. Overall, 93% (56/60) of the patients opened the invitation and 43% (26/60) consented to participate. DBT skills website users visited the home page on an average of 5.3 times (SD 6.0). Procedures resulted in no complaints and some participant feedback emphasizing the usefulness of DBT skills.
This study supports the potential of using responses to patient health questionnaires in EMRs to identify a high-risk population and offer key elements of caring messages and DBT adapted for a low-intensity intervention. A randomized trial evaluating the effectiveness of this program is now underway (ClinicalTrials.gov: NCT02326883).
Significant momentum supports the work of suicide prevention in health care settings as well as the inclusion of people who have personal experience with suicidal thoughts and behaviors, including systems changes initiatives such as Zero Suicide and new national patient care guidelines such as Recommended Standard Care for People With Suicide Risk: Making Health Care Suicide Safe [
Selective prevention—otherwise known as secondary or indicated prevention—focused on outreach, support, and skills development would allow for a more collaborative approach. Owing to several conceptual, methodological, logistical, and ethical challenges, selective prevention for suicide remains to be relatively unexplored. Effective selective prevention for suicide requires three things: (1) accurate methods for identifying those at risk before a suicide attempt, (2) effective interventions suitable for large-scale delivery, and (3) acceptable and efficient models for population-based delivery.
The Patient Health Questionnaire (PHQ) [
A promising low-intensity intervention is caring messages, comprising brief, nondemanding, unsolicited caring messages sent by a health care provider over time. Although caring messages studies are not without limitations, data suggest that they may reduce the incidence of suicide in psychiatric patients [
DBT, another intervention, is a high-intensity treatment demonstrated to reduce suicidal behaviors by teaching individuals how to effectively manage intense emotions and tolerate distress [
For the delivery of such interventions, the population-level impact requires a model that is affordable and scalable. Web-based interventions have the potential to reach large numbers of high-risk individuals and have been shown to be effective for support inside and outside medical settings [
Meanwhile, EMR systems can be used not only to support the delivery of web-based interventions but also to identify those at risk and to support patient-provider secure messaging used for intervention. EMRs with secure messaging have the potential to significantly increase the efficiency and dissemination of suicide prevention and suicide attempts. Secure messaging has been successfully used to deliver low-intensity interventions [
This pilot study aims to assess the feasibility, acceptability, and engagement of a brief suicide prevention intervention in preparation for a full-scale randomized trial. We evaluated this by measuring patient response to an invitation to a web-based, population-based suicide attempt and suicide prevention intervention, intended to be easily adapted into large health care systems, which included elements of DBT and caring messages.
We evaluated the feasibility, acceptability, and engagement of a brief web-based suicide prevention intervention using qualitative and quantitative methods. The intervention included EMR-mediated secure messaging linked to a web-based DBT skills platform. Participants included outpatients who were receiving care within Kaiser Permanente Washington, a health system serving approximately 700,000 people in Washington State, which routinely administers the PHQ to all patients receiving care for mental health conditions [
Over the course of 3.5 months, we used an automated EMR data query program to pull 6 samples, identify 60 individuals, and conduct outreach via secure messaging accessible on a mobile app or website. We used this sample size, which could be considered large for a pilot, to gain confidence in our estimates for the feasibility of a large clinical trial. We sent secure message invitations to the DBT skills program approximately every other week to between 5 and 20 eligible participants for each sample (starting at the top of the list) to evaluate ideal caseload management. Invitations were sent 1 to 2 weeks after the visit, if an elevated PHQ item 9 was involved. Patient care continued to be the responsibility of the behavioral health and primary care providers in the Kaiser Permanente Washington system who had administered the PHQ at the recent visit. In addition, a suicide risk protocol involving follow-up by study clinicians (author 1 or 3) for those who indicated imminent risk over secure message was in place, although the protocol was not triggered during the study. The study procedures were approved by the organization’s institutional review board, which included a waiver of research consent for patient identification and intervention invitation; however, this restricted the analysis of additional data (eg, descriptive data of eligible participants). The waiver of consent for patient identification and intervention invitation was granted because this pilot and subsequent trial involved no more than minimal risk and because the purpose of the pilot was to evaluate the feasibility and acceptability of procedures planned for the randomized trial of outreach. Evaluating only patients who actively consented to receive intervention would have yielded a result of questionable validity and generalizability.
The invitation messages, and the messages that followed, can be conceptualized as having two elements: a caring message (a brief and unsolicited expression of care) and support for a web-based DBT skills intervention (
The web-based DBT skills site included four DBT skills (mindfulness, opposite action, mindfulness of current emotion, and paced breathing). This content focused on training in specific skills to manage upsetting or painful emotions that can precipitate suicidal thoughts and behavior. Content was based on a brief DBT skills face-to-face intervention developed and pilot-tested by the first author [
Dear [First Name]
Sometimes a little extra help can be just what someone needs to get through tough times.
A new web-based program called Now Matters Now was designed to give you that help when you need it. The program uses real people to teach specific coping skills, like mindfulness and paced breathing.
We invite you to try Now Matters Now today. Learn more at <Link to Intervention Website>
We care about you.
My job is to encourage you to use the program over the next three months and to practice the skills you are learning. If you haven’t visited the program in a while, I’ll send you a message to remind you. If you have visited, I’ll check with you about what you find helpful.
I’ll communicate with you through messages, and in order to message you, I will access your medical record.
Sending care,
TEAM NOW MATTERS NOW—Introducing the real people who use these skills and helped develop this program
MINDFULNESS—A way to be present for what is most important in your life
OPPOSITE ACTION—Living a full and healthy life, despite what negative or unhelpful emotions and thoughts are telling you to do
MINDFULNESS OF CURRENT EMOTION—Observing, honoring, and moving through strong emotions without being controlled by them
PACED BREATHING—A different and scientific approach to the old saying “Just Breathe”
GIVE US FEEDBACK—Did you have technical problems with this program? Do you want to give us feedback about the program, good or bad?
The DBT skills intervention was supported by an interventionist or program
Hi [First Name],
I’m writing to check-in from the Now Matters Now program. Do any of the skills look interesting to you to try?
Paced Breathing is a breathing technique that can calm the mind and body naturally. Some people like to use Paced Breathing when they’re feeling nervous or anxious. It can also be a helpful way to clear the mind if you’re having a difficult time falling asleep.
The key to Paced Breathing is to have your exhale be longer than your inhale. It can be a good way to center yourself and makes it easier to then practice other skills.
If you try it, let me know how it goes.
To learn more, please click the link below:
[link to Consent “form” video]
Take care,
We evaluated feasibility by measuring the proportion of invited patients who opened the program invitation, the proportion of patients who visited the consent page, and the proportion of patients who consented. We qualitatively evaluated acceptability by using secure message feedback that the coach received from patients invited to participate and summarized relevant themes [
In total, 60 patients were invited to participate (
Study flow.
The coach received no complaints from those invited to participate in the program. The feedback the coach received via secure messaging mainly included descriptions of how participants had used the program, specifically their helpful experiences practicing each of the four program skills and the challenges they experienced. One participant described a helpful experience in trying mindfulness skills, which involved bringing attention and awareness to their thoughts:
I did the first lesson last night—I was able to use it today when I started having negative thoughts that wouldn’t stop. I cleared my mind and started looking at the trees, buildings and other things around me and noticing the little details of things. It actually calmed me down and gave me a sense of control.
Another participant described the experience of practicing paced breathing:
I am doing my best to make the Paced Breathing a daily practice. I forget sometimes until I am “over the edge” and am in deep panic/terror, but I can still even then use the breathing to help get me feeling more calm.
Another participant described a powerful experience using an opposite action, which involved choosing to do the opposite of ineffective emotional urges:
Yesterday I began the module called “Opposite Action” and printed it out after I completed it. Doing it was empowering but also very emotional for me.
Finally, a participant described a plan to use mindfulness of current emotion, which involved learning how to bring attention and awareness to the bodily sensations of emotions:
I used [a DBT skill] about 5 times today when I was out walking—it keeps my emotions under control, so I don’t start crying. I’m going to try [the]“mindful emotion” thing tomorrow. I appreciate [my health care organization] offering something like this—I like having concrete tools I can use.
In terms of challenges, participants described difficulties in using their skills effectively while experiencing intense emotions. For example, one participant described:
One inescapable problem I have found is that once the line that gets crossed where “there is no hope” of being able to turn around the anxiety, I have little ability to do anything constructive to stop the terror from escalating ... if I am not too “far gone” the skills do help to lower the physical and mental/emotional distress.
In total, 92% (24/26) of the participants that consented had visited the landing page of DBT skills. These DBT skills site users logged in on an average of 5.3 times to the home page (SD 6.0, range 1-26). We examined a subsample of those consenters to look at engagement in the specific DBT modules (eg, opposite action). Among the first 20 participants enrolled, 19 (95%) read the invite. Of these 19 participants, 12 (63%) visited the website and 8 (42%) agreed to participate. Of those who consented, participant engagement in the different skills ranged from 38% (3/8) for paced breathing to 88% (7/8) for opposite action (
Dialectical behavior therapy skill engagement for subsample (N=8).
Program visits | Opposite action | Mindfulness | Mindfulness of current emotion | Paced breathing |
Visited at least once, n (%) | 7 (88) | 4 (50) | 6 (75) | 3 (38) |
Average visit, range (min) | 23 (1-54) | 34 (2-65) | 13 (1-57) | 18 (3-30) |
Visited at least twice, n (%) | 4 (50) | 3 (38) | 1 (13) | 3 (38) |
Visited 3 times or more, n (%) | 2 (25) | 3 (38) | 1 (13) | 1 (13) |
This pilot study demonstrated the feasibility, acceptability, and engagement of a population-based intervention involving active elements of caring messages and DBT. Specifically, we demonstrated that by using an accurate method for identifying those at risk for suicide and suicide attempt—in this case, PHQ item 9—we can adapt existing interventions to be suitable for delivery on a large scale and that these models are promising for intervention delivery.
The number of patients who received the initial message that involved elements of caring messages was high (56/60, 93%). This provides support for the delivery of caring messages through this type of intervention (delivered over time through secure messaging with the support of a coach or care manager). The rate of patients who actually reached the DBT skills site content (26/60, 43%) indicates a need for strategies to reduce barriers to DBT skills site participation, such as simplifying the invitation and consent process for patients. The steepest drops were between viewing the message and visiting the consent page (17/56, 30% of those remaining were lost), and between visiting the consent page and consenting to the DBT skills intervention (13/39, 33% of those remaining were lost;
A limitation of our study was that the data analysis was limited to anonymized information regarding program participation. Health records data to identify potential participants were used under a waiver of informed consent, so we did not retain demographic, diagnostic, or other clinical information, including the total PHQ score. Although patients were identified as eligible based on their high severity of suicidal thoughts and all eligible patients were enrolled in the study, it could be that there were more and less severe patients within this group and that DBT skills intervention participation varied depending on the severity. Furthermore, participation among those invited may have been reduced by the need for a research consent process. We suspect that engagement would have been higher without this step between secure message invitation and the DBT skills site. In addition, we do not yet know whether the combination of elements of caring messages with brief DBT skills support will be effective, given that caring messages are intended to be nondirective. Finally, it is possible that those who received the invitation to the intervention felt singled out in some way, having had a visit recently where they disclosed suicidal thoughts. Indeed, qualitative research indicates that those who receive the PHQ-9 have significant concerns about how the information regarding the ninth item and suicide will be used [
As noted by one participant, once one’s emotions reach a certain threshold, the DBT skills become more difficult to access. One limitation of this intervention is that the DBT skills content was not framed in terms of skills for day-to-day stressors versus skills for acute crises. Future interventions may benefit from tailoring additional skills to match the spectrum of emotional stress/crises [
This project addresses a number of the most important conceptual, methodological, logistical, and ethical challenges that hamper major advancement in the field of suicide prevention. Qualitative and quantitative data from this study were used to inform the design of a full-scale effectiveness trial (ClinicalTrials.gov: NCT02326883). This research, led by four health care systems within the Mental Health Research Network, is underway to determine its effectiveness [
The videos that were developed as part of this project that make up most of the DBT skills content are publicly available to both the researchers and the public as one piece of NowMattersNow.org [
Consent “form” video.
Program landing page video.
cognitive behavioral therapy
dialectical behavior therapy
electronic medical record
Patient Health Questionnaire
This study was funded by the National Institute of Mental Health (5R34MH097836 and 1UH2AT007755) and the American Foundation for Suicide Prevention. The authors also thank the
None declared.