No traditional treatments have been consistently shown to reduce nonsuicidal self-injury (NSSI) compared to an active control group. Regardless of their efficacy, these treatments have limited potential to impact national and international rates of NSSI because they are not scalable. In other words, because of logistical barriers (e.g., ~100:1 ratio of people in need to mental healthcare professionals), cost barriers (e.g., money, time, effort), and autonomy barriers (e.g., desire to handle one's own problems), traditional treatments struggle to reach many people who engage in NSSI. In the series of studies described below, we endeavored to take the initial steps toward developing an effective NSSI intervention that has the potential to overcome these barriers.
Based on our prior laboratory and longitudinal studies, we created an app-based treatment that we call Therapeutic Evaluative Conditioning (TEC) for the reduction of NSSI. Given that self-cutting is a particularly common and severe form of NSSI, this version of TEC primarily targeted self-cutting. Across two one-month, web-based randomized controlled trials (Ns = 114 and 131 participants, all severe/recent self-injurers) we found individuals with access to the active TEC app displayed moderate reductions in NSSI frequency - especially self-cutting frequency (37% and 40% reductions) - compared to the individuals with access to a control version of the TEC app. These significant effects persisted after controlling for several covariates, including prior month self-cutting, self-prediction of future NSSI, desire to stop NSSI, treatment history, emotion reactivity, and internalizing symptoms. Participants were allowed to play the app as often (or as little) as they desired. Both studies detected a significant dosage effect, with greater use of the active TEC app being associated with greater reductions in self-cutting frequency (there were no dosage effects for the control version of TEC). Notably, even participants who received a "high dosage" of active TEC only played the app for about 10 minutes per week. However, exploratory analyses revealed that these treatment effects were not maintained during the post-treatment month, suggesting that effects depend on regular TEC use.
Somewhat surprisingly, this NSSI-focused version of TEC also seemed to have effects on suicide-related phenomena. Specifically, Study 1 detected a significant reduction in suicide plans (45% reduction), though this effect was not significant in Study 2 (21% reduction). There were also reductions in the frequency of suicidal behaviors in Study 1 (47% reduction) and Study 2 (33% reduction). Among those who accessed active TEC at least once (as noted above, participants were not forced to access TEC), reductions were even greater (79% and 70% reductions). To maximize reliability of results for these rare behaviors (n = 64 combined for Studies 1 and 2), these data were combined with suicidal behavior data from Study 3 (which was more suicide-focused) for statistical analyses (click here for more information). However, there were no significant effects on suicide ideation in any study.
Although TEC for the reduction of NSSI requires much further testing and could benefit from many improvements, these initial studies suggest that may be a promising potential large-scale treatment for NSSI. TEC itself is described in greater detail below.
Recommended citation for this information: Franklin, J.C., Fox, K.R., Franklin, C.R., Kleiman, E.M., Ribeiro, J.D., Jaroszewski, A.C., Hooley, J.M., & Nock, M.K. (in press). A brief mobile app reduces nonsuicidal and suicidal self-injury: Evidence from three randomized controlled trials. Journal of Consulting and Clinical Psychology.
Note: Two co-authors have a patent pending for the general design/function of TEC (JCF, CRF) and one owns an app distribution company (JCF). TEC was developed/programmed by CRF.
What is evaluative conditioning?
Answer: A form of Pavlovian conditioning that includes an extra evaluation component (i.e., whether the conditioning increases or decreases liking of a given stimulus)
For example, if these two stimuli were repeatedly paired together, you would eventually start to like the first stimulus more
Likewise, if these two stimuli were repeatedly paired, you would eventually start to dislike the first stimulus more
There is no standard evaluative conditioning paradigm (see Hoffman et al., 2010 in Psychological Bulletin for more information). However, most evaluative conditioning paradigms involve passive viewing of simultaneous or sequential images/words and most paradigms are administered one time, usually in a laboratory setting. We wanted to create a more powerful and fun paradigm. As a result, TEC is a brief (1-2 minute) mobile app that includes several gamification elements designed to (a) motivate the individual to play TEC multiple times [ideally hundreds or thousands] and (b) increase the strength of conditioning. Essentially, TEC is a matching game that increases in difficulty across trials; faster and more accurate matching is rewarded with points, achievement badges, etc.
What are the treatment targets of TEC-NSSI?
Answer: This version of TEC has two primary treatment targets -- (1) diminished aversion toward NSSI stimuli and (2) heightened self-aversion.
As described in greater detail here, our prior laboratory work has shown that people who engage in NSSI tend to regard NSSI-related stimuli (e.g., blood, wounds, knife cutting the skin) as neutral or even pleasant. This is especially true for people who have engaged in NSSI recently or frequently, and diminished aversion to NSSI stimuli is a strong predictor of future NSSI. These findings are consistent with the idea of self-injury/suicide capability (Franklin et al., 2014a,b; Joiner, 2005; Joiner, Ribeiro, Silva, 2012). In short, for most people these stimuli evoke an instinctive response that motivates them not to engage in NSSI. In the absence of this barrier it is easier to engage in NSSI. TEC is designed to rebuild this barrier by increasing the aversion to NSSI stimuli. To accomplish this, TEC repeatedly pairs NSSI stimuli with unpleasant stimuli.
Based primarily on Dr. Jill Hooley's work, TEC also targeted negative association with the self. Although most people have a positive self-association, Dr. Hooley's work has shown that many people who engage in NSSI display a negative association with the self (e.g., high levels of self-criticism, self-disgust, etc.). This factor appears to play a strong role in motivating NSSI (especially initial NSSI episodes) and plays a key role in allowing individuals to overcome the physical pain involved in NSSI (see Hooley & St. Germain, 2013). In an attempt to restore a positive self-association, TEC repeatedly pairs self-related words with pleasant stimuli.
Although these initial studies suggest that TEC is promising, many improvements should be made to increase TEC's potentcy. For example, we are currently working on new versions of TEC that include personalization algorithms, additional gamification elements, and a wider range of language options. We are also working on ways to integrate TEC with large-scale risk detection strategies in an effort to connect those in need with TEC. We hope that future work ultimately goes far beyond these preliminary studies with TEC to create a wide range of highly effective and accessible apps (or other novel-format interventions) that can impact NSSI on a large scale.