Effectiveness of a Self-management Program After Traumatic Injury

This study is designed as a pragmatic randomized controlled trial using a mixed-methods design with 6-months follow-up to determine the effectiveness of a newly developed self-management support program, entitled the SEMPO. Participants will be included 3-4 months post-injury and followed up over a 12 months period.

Study setting and participants:

The study will be conducted at Oslo University Hospital (OUH). Patients will be recruited at the Trauma Centre at Oslo University Hospital, which is the Level 1 trauma centre for the southeast region of Norway and serves a population base consisting of more than half of the Norwegian population. Patients who fulfil the following inclusion criteria will be asked to participate:

• Adults residing in the southeast region of Norway who are aged between 18 and 72 years.
• Admitted to OUH directly or after transfer from local hospitals within 72 hours of injury.
• At least a two-day hospital stay.
• Traumatic injury corresponding to a New Injury Severity Scale score (NISS) >9.
• Patients reporting injury-related symptoms, functional impairments, and/or difficulties with daily activities at the discharge from OUH. According to preliminary analyses in ongoing study on unmet needs after trauma (funded by South-Eastern Norway Regional Health Authority), these patients are in risk of developing long-lasting disabilities.

Patients will be excluded in case of:

• Cognitive function corresponding to a Mini Mental Status score (MMS) < 20 points.
• Psychiatric diseases that require treatment.
• Drug/alcohol dependence that require treatment.
• Complete spinal cord injury and isolated thoracic or abdominal injury.
• Insufficient command of Norwegian.

Patients who consent to participate will be randomised into a randomisation arm (RA) or a self-selection arm (SA). Participants in the RA will then be further randomised to intervention or control group. In the self-selection arm, the participants can choose whether they will be placed in the intervention or control group. The self-selection arm is an explorative part of the study to explore if taking into account patient treatment preferences may maximise adherence, reduce potential attrition bias and influence the intervention results. Data will be collected through interview/patient-reported outcome measures (PROMs) at the following time points: baseline assessment (T1), after completion of the intervention (T2), 3-months after completion of the intervention (T3), and 6-months after the intervention (T4), with similar time-points for those in the control group.

The primary outcome is self-efficacy assessed with the Trauma Coping Self-Efficacy Scale. Secondary outcomes include symptom burden, physical, cognitive and emotional functioning. Additional outcomes include resilience, health-related quality of life, communication with health professionals, health literacy, global functioning and disability, and evaluation of changes. The participants will complete all outcome measures. In addition, participants will nominate their target problem areas with their own words and will be asked to select one personal outcome preference from the predefined outcomes.

Sample size calculation and statistical analyses:

Sample size was estimated based on the primary outcome. Results from a study on mental eHealth interventions following trauma were used. A power calculation was conducted with a standard deviation of 1.0 and a mean group difference of 0.6. Given an 80% power and an alpha of 0.05, 90 participants should be included to observe significant treatment related improvement. With an assumed attrition rate of 20%, 110 participants will be included (i.e. 55 participants in each group). The sample size calculation was done based on the randomization arm (RA), and we aim to achieve the same power in the self-selection (SA) arm. However, the SA is considered an exploratory part of this study to evaluate the influence of patients' treatment preferences.

To investigate the effectiveness of the intervention on self-efficacy, symptom burden and disability, and patient-impression of change, mixed-effect models will be used to account for repeated measurements by patients. Time and time-by-treatment interaction will be used as fixed effects in these models. The linear mixed model will give estimated mean values for all time points (T1, T2, T3, and T4), changes from baseline, and between -group differences from baseline with 95% confidence intervals. The analyses will be based on an "intention to treat" principle. In addition, clinical benefit will be estimated in quality-adjusted life years (QALYs) using standardized conversion tools to convert health benefits into an index of HRQL, as measured by EQ-5D. Simulated cost consequences of the program will be considered from health and societal perspectives.