Research project Long-term effects of treatments for self-harming behavior - statistical method development +analysis
Long-term effects on adolescent suicide attempts, diagnoses, drug use and school attendance after self-harm treatment - statistical method development and analysis of register data based on a RCT
Self-injury is a serious health problem that often manifests in early adolescence, with approximately 20% of Swedish high school students affected. Self-injury is strongly correlated with suicide, which is the leading cause of death among individuals aged 15-24. It is crucial to implement effective treatment methods early for adolescents engaging in self-injury, to prevent suffering, suicide, and societal costs.
Project description
The project analyzes the long-term effects of two treatment methods for self-harming adolescents, differences between the methods and treatment as usual, and for which groups each treatment method works best. The focus is on long-term effects related to suicide risk, comorbidity, medication use, and school attendance.
The goal is to plan and conduct a register study to provide knowledge about the long-term effects of the two treatment methods internet-delivered ERITA (Emotion Regulation Individual Therapy for Adolescents) and DBT-A (Dialectical Behavior Therapy for Adolescents). Data from the register study is also combined with data from a previously conducted randomized clinical trial of ERITA.
A significant advantage of register studies is that they facilitate long-term follow-ups, minimize the burden on patients and their families, and can be conducted at lower costs for society. However, drawing conclusions about causal effects through observational studies is generally associated with greater statistical challenges compared to randomized experiments. Accurately analyzing observational data and estimating treatment effects, require careful control of assumptions within the framework for causal inference.
An important aim of the project is statistical method development throughout all stages of the study, including study design, matching against controls, data analysis, bias adjustment, and sensitivity analyses. The project also aims to develop analysis methods tailored to the combination of register data and data from a randomized clinical trial.
Project members
Project managers
Ellinor Fackle Fornius
Universitetslektor
Jessica Franzén
Universitetslektor
