Zero Self-Harm - a Mobile Phone Application to Reduce Non-suicidal Self-injury

The purpose of the trial is to investigate whether the introduction of the Zero Self-harm app to individuals with NSSI in Denmark, can 1) reduce the frequency of NSSI-episodes, 2) prevent escalation of NSSI and 3) reduce suicide ideation, and depressive symptoms, compared to treatment as usual.

Hypothesis: The hypothesis is that Zero Self-Harm will reduce the frequency of NSSI-episodes by enhancing participants skills to avoid impulsive urges resulting in NSSI. Thereby, the app will help deal with negative affect and trait impulsivity, core features in the etiology of NSSI, and further decrease associated symptoms, e.g. suicide ideation and depression. We expect that the Zero Self-Harm app will be able to reduce the mean monthly frequency of NSSI in the intervention group by at least 3 episodes. Thus, we hypothesize that the Zero Self-Harm app will be superior to the control condition in 1) reducing the frequency of NSSI-episodes, 2) preventing escalation of NSSI and 3) reducing suicide ideation, and depressive symptoms, compared to treatment as usual.

Methods and design: The trial is designed as a 2-arm, parallel-group, multicentre, pragmatic, randomized clinical superiority trial. Based on the power calculation listed below, a total of 280 participants, 140 in each arm, should be included. One group will receive the Zero Self-Harm app by downloading a specific link sent to them after randomization, and the control group will be allocated to a wait-list and allowed to download the app after they have finished the last questionnaire at 6 months.

Recruitment: Recruitment will be broad and target both individuals with shorter duration and milder symptoms of NSSI, possibly not in contact with mental health facilities, and individuals with more severe and chronic NSSI, already in contact with mental health facilities in Denmark. Recruitment will be done through non-profit organizations, service centres in municipalities, schools for adolescents above 18, outpatient clinics for psychiatric disorders, and psychiatric and somatic emergency departments in Denmark. Recruitment will be conducted through non-profit organizations' websites and social media platforms, as well as providing information about the project through presentations at schools, posters at relevant recruitment facilities and encouraging personnel to refer potential participants when meeting people with NSSI.

Enrolment and randomization: Enrolment will be done by the research assistant when contacted by potential participants via email or telephone. Participants will be encouraged not to send personal information through email and only over the phone, to ensure confidentiality. Potential participants will be given oral and written information about the study, time for consideration and the possibility for an appointment with a next of kin before deciding to enrol in the project. Consent will be obtained by sending an e-mail from REDCap to the trusted Danish secure platform e-boks.

After filling out the baseline questionnaire, participants will be randomly assigned to either the waitlist control group or the Zero Self-harm app, with a 1:1 allocation, using a computer-generated sequence randomization generator in REDCap. The randomization will be stratified by sex and number of previous NSSI-episodes (based on the baseline score of the DSHI) to avoid overrepresentation of patients with a high number og NSSI in one treatment group.

Blinding: Due to the nature of the intervention, neither participants nor the research assistant can be blinded in this trial, and neither primary nor explorative outcomes can be blinded. An employee outside the research team will extract data from REDCap at study completion to two separate excel-sheets, and group allocation will be coded with A and B, to ensure blinding of the research assistant while analysing data, drawing conclusions and drafting the first manuscript. Blinding will be lifted once drafting of the first manuscript has been done.

Intervention: When randomized to the Zero Self-Harm app the participants will receive an email with a link to download the app and an introduction via an instruction manual, in addition to instruction videos implemented in the app. This will ensure the navigation and knowledge of the technicalities of the app. Participants in the waitlist control group will receive no study intervention but will be offered the possibility to download the Zero Self-Harm app after they have finished the last questionnaire at 6 months. This will be stressed at the initial appointment as well as after the collection of data.

Data management: Above-mentioned outcome measures are collected at baseline and 3- and 6-month follow-up through self-administered internet-based questionnaires from REDCap. Participants will receive an e-mail with a link to the online survey, where they can log on to a secure data portal, REDCap, with a personal trial id-number and a two step-code. To improve the response rate, two email reminders are sent out automatically, and if participants still have not answered the questionnaire, they will receive a phone call from the research assistant encouraging them to answer the questionnaire. When data collection is finished, all data from the participants will be imported to a local secured drive with limited access.

Further, there will be a continuous data collection with information about the usage of the app, and register-based data will be collected through the Danish National Hospital Register and the Cause of Death Register. App use and register-based outcomes, such as mortality and admission to psychiatric and/or somatic emergency departments, will be assessed at 3 and 6 months.

To be aware of potential harms of the app, the 20 first participants receiving the Zero Self-harm app will receive a call after they have completed the last questionnaire at 6-months, and interviewed about their experiences of the app. These interviews will be analysed for possible harms, and the results will be discussed in the research group which will then decide on whether or not to continue the trial.

Power analysis: Primary outcome is the mean difference in the monthly frequency of NSSI after 6 months. We expect that participants in the intervention group will have a mean score that is 3.5 points lower compared to participants in the TAU group at follow-up after 6 months. The mean score is based on clinical experience and estimates from a previous study investigating the interventions ability to reduce monthly NSSI episodes (45). Based on previous publications, we expect a post-intervention standard deviation of 9. If the true difference in the experimental and control means is 3.5, we will need to include 140 participants for each group to a total of 280 participants to be able to reject the null hypothesis, that the population means of the experimental and control groups are equal with a probability of 0,1. The Type I error probability associated with this test is 0.05.

Statistical analysis: Analysis of the primary outcome will be conducted according to the intention-to-treat principle: All participants will be included in the final analysis according to group assignment regardless of adherence to treatment. This study has multiple assessment points and for the primary analysis, we will use repeated measurements in a mixed model with unstructured variance. This procedure can handle baseline differences and missing data. For participants with missing data in two or more data points, we will identify potential differences to participants with full data and include these variables as potential confounders in secondary analysis .

Statistical analysis will be conducted in SPSS, version 22.0. Confidence intervals will be presented as well as significance level. All tests will be two-tailed and p values below 0.5 will be considered significant and interpreted with respect to hierarchy of hypothesis recognizing that all outcomes, apart from the primary, are exploratory .