Brief Online Help-seeking Barrier Reduction Intervention

Randomized Controlled Trial of an Online Machine Learning-Driven Risk Assessment and Intervention Platform for Increasing the Use of Crisis Services

Objective: Mental illness is a leading cause of disease burden; however, many barriers prevent people from seeking mental health services. Technological innovations may improve the ability to reach under-served populations by overcoming many existing barriers. The investigators evaluated a brief, automated risk assessment and intervention platform designed to increase the use of crisis resources provided to individuals who were online and in crisis. Hypothesis: The investigators hypothesized that individuals assigned to the intervention condition would report using crisis resources at higher rates than individuals in the control condition. Method: Participants, users of the digital mental health app Koko, were randomly assigned to treatment or control conditions upon accessing the app and were included in the study after their posts were identified by machine learning classifiers as signaling a current mental health crisis. Participants in the treatment condition received a brief Barrier Reduction Intervention (BRI) designed to increase the use of crisis service referrals provided on the app. Participants were followed-up several hours later to assess the use of crisis services.

The BRI was designed to overcome common concerns and misconceptions (i.e., barriers) related to using crisis services. It works by first asking the user about what potential barriers may keep them from using the crisis service referrals, and then, based on the user's response, by providing information intended to help the user overcome the potential barrier(s) they selected. By exploring the menu of barriers, users could read brief messages designed to dispel common misconceptions or concerns related to each barrier. For example, a common concern among Koko users was that calls to lifelines invariably result in visits by the police or other emergency services. Users who feared this possibility could tap on the associated button and learn that active rescues such as these are extremely rare, and occur in less than one percent of all cases. Whenever possible, we used language throughout the intervention to help validate the experiences of the users.

The number participants indicating at follow-up that they used the crisis resources provided to them (e.g., called the suicide crisis hotline)

Number of Participants in the Treatment Versus Control Conditions Reporting Their Experience Using Koko Was "Good"

The number of participants in the treatment versus control conditions reporting that their experience on the Koko digital platform was "good" versus "bad" using a two-option response question.

Inclusion Criteria: Participants identified as experiencing a mental health crisis by a hybrid human-machine computation system evaluating semantic content of posts made on digital platforms. Exclusion Criteria: Participants who were not identified as experiencing a mental health crisis.

Jaroszewski AC, Morris RR, Nock MK. Randomized controlled trial of an online machine learning-driven risk assessment and intervention platform for increasing the use of crisis services. J Consult Clin Psychol. 2019 Apr;87(4):370-379. doi: 10.1037/ccp0000389.