Abstract
BACKGROUND: Depression and anxiety is a major public health problem, in Sweden and internationally. Internet-based interventions are increasingly acknowledged as promising approaches for individuals with varying degrees of mental health problems. We present findings from the implementation of myCompass, a fully automated self-guided intervention of Australian origin, in a Swedish context. METHODS: We (i) share our experience of the E-health study platform (i.e., regarding security aspects, functionality) to which the myCompass intervention was linked, and (ii) report findings from the empirical evaluation of myCompass (i.e., prerequisites, execution, study outcomes), in a community sample of individuals (N = 837) reporting mild-to-moderate levels of depression, anxiety and stress. Outcomes were calculated with repeated measures ANOVA and linear mixed models. RESULTS: The E-health study platform proved to be an efficient tool enabling randomization, informed consent and evaluation to be administered in a fully automated manner. The study rendered substantial interest initially with 1207 individuals enrolling, however it failed to maintain engagement of those enrolled with only few participants logging in more than once or twice following registration. A smaller subgroup of "active users" (n = 35) had a markedly higher activity in the program, however their treatment results were not significantly better than those of the control group. CONCLUSION: Based on the large number of dropouts and also modest use of the intervention overall, only tentative speculations can be made regarding its effectiveness in a Swedish context. The number of individuals remaining active in the intervention is much more limited that the number of individuals initially signing up. Moreover, the transportation of interventions across countries and cultures may need more careful consideration, and pilot-trials before attempting large-scale trials are recommended. TRIAL REGISTRATION: MyCompass was retrospectively registered at ClinicalTrials.gov . NCT03659630 September 3rd 2018, and was given the protocol ID 2015/1268-31/2 + 2016/88.