An Internet-based emotion regulation intervention versus no intervention for non-suicidal self-injury in adolescents: a statistical analysis plan for a feasibility randomised clinical trial

Background: Non-suicidal self-injury (NSSI) has a lifetime prevalence of 17% in adolescents in the general population and up to 74% in adolescents with psychiatric disorders. NSSI is one of the most important predictors of later suicidal behaviour and death by suicide. The TEENS feasibility trial was initiated to assess the feasibility and safety of Internet-based Emotion Regulation Individual Therapy for Adolescents (ERITA) as an add-on to treatment as usual in 13-17-year-old patients with NSSI referred to the Child and Adolescent Mental Health Services. Methods: The TEENS feasibility trial is a randomised clinical trial with a parallel-group design. The trial intervention is an 11-week online therapy which is tested as an add-on to treatment as usual versus treatment as usual. The primary feasibility outcomes are the fraction of participants who (1) completed 12 weeks of follow-up interview or assessment, (2) consented to inclusion and randomisation out of all eligible participants, and (3) were compliant with the experimental intervention, assessed as completion of at least six out of eleven modules in the programme. Since this is a feasibility trial, we did not predefine a required sample size. The exploratory clinical outcome, the frequency of NSSI episodes, assessed using Deliberate Self-Harm Inventory - Youth version (DSHI-Y), at the end of intervention, is planned to be the future primary outcome in a larger pragmatic definitive randomised clinical trial. After completion of the feasibility trial, blinded data will be analysed by two independent statisticians blinded to the intervention, where 'A' and 'B' refer to the two groups. A third party will compare these reports, and discrepancies will be discussed. The statistical report with the analyses chosen for the manuscript is being tracked using a version control system, and both statistical reports will be published as a supplementary material. Based on the final statistical report, two blinded conclusions will be drawn by the steering group. Discussion: We present a pre-defined statistical analysis plan for the TEENS feasibility trial, which limits bias, p-hacking, data-driven interpretations. This statistical analysis plan is accompanied by a pre-programmed version-controlled statistical report with simulated data, which increases transparency and reproducibility.