Many individuals with eating disorder (ED) symptoms do not receive appropriate mental health care due to shame, fear of stigmatization or the belief that one’s symptoms are not severe enough. The combination of E-health and peer support by expert patients (in Dutch: ‘ervaringsdeskundigen’ also called experts by experience) may help to decrease the time between onset of symptoms and seeking help and may furthermore be useful for individuals who have completed treatment and who are at risk for relapse. Internet-based interventions are easily accessible and can reduce social barriers for seeking help. The self-evident credibility of expert patients may improve the self-efficacy of individuals with EDs and improve patient satisfaction.
In a previous randomized controlled trial (RCT), our group found that an Internet-based intervention (called ‘Featback’) with and without therapist support was (cost-)effective in comparison to a waiting list. Featback consists of psychoeducation and a fully automated monitoring- and feedback system. Weekly, participants receive an e-mail containing a link to a short monitoring questionnaire tapping ED symptoms. After completion, tailored feedback messages are automatically generated according to a pre-defined algorithm and send to the participants. However, more high-quality studies are required to determine for whom and under which circumstances the intervention is effective. Furthermore, the effectiveness of expert patient support requires further investigation.
The present study will be a RCT comparing 4 conditions: 1) Internet-based intervention ‘Featback’, 2) online support from an expert patient, 3) Featback supplemented with online support from an expert patient, and 4) a waiting list control condition. This study has two aims: 1) To investigate the effectiveness and cost-effectiveness of 8 weeks of Featback, in comparison to Featback with online support from an expert patient, online support from an expert patient without Featback, and a waiting list control condition. 2) To investigate moderators of intervention response: what works for whom? Eligible participants are aged = 16 years, have access to the Internet, have written and oral fluency in Dutch language, and have self-reported ED symptoms.
After the end of the study, participants who are randomized to the waiting list will be offered Featback plus weekly support by an expert patient. Participants in all four intervention conditions will be free to undergo any other type of intervention or treatment during the study (i.e., usual care). There will be monthly supervision from an experienced clinical psychologist and an expert patient as a matter of routine professional and ethical care, and to reinforce adherence to the protocol. Participants will be assessed at baseline, post-intervention (i.e. after 8 weeks) and at 3-, 6-, 9-, and 12-month follow-up. The primary outcome measure is ED psychopathology (Eating Disorder Examination Questionnaire: EDEQ and the Short Evaluation of Eating Disorders: SEED). Secondary outcome measures are generic (EQ-5D) and ED-related quality of life (ED-QOL), symptoms of depression and anxiety (Patient Health Questionnaire: PHQ-4), social support (SSL) and self-efficacy (GSE). Costs will be evaluated by the Trimbos/iMTA questionnaire for Costs associated with Psychiatric Illness (TiC-P).
We conducted an a priori statistical power analysis (G*Power) to determine the optimal sample size having 95% power to detect a small effect size of 0.25 (two-tailed a=0.05) between the four study conditions using 6 repeated assessments. The effect size was based on data from our previous randomized controlled trial (Aardoom et al., 2016a). Adjusting for an anticipated dropout rate of 40% (based on previous data), a minimum of 73 participants per condition (N=292) will be needed
Longitudinal linear mixed-effects model analyses will be conducted according to an intent to treat principle. Potential moderators will be investigated using model-based recursive partitioning methods (Zeileis et al., 2008). Model-based recursive partitioning can be used to detect treatment-subgroup interactions: subgroups of individuals with different (i.e. more or less favorable) responses to one or more interventions, as compared with other individuals. The economic evaluation will be performed from a societal perspective. The uncertainty around the mean costs and effects per participant will be estimated using bootstrapping techniques. The results will be represented in cost-effectiveness acceptability curves.