Usage and weekly attrition from a smartphone based health behaviour intervention for adolescents: Pilot study (Preprint)
BACKGROUND Majority of adolescents own smartphones though only 8% of them use health apps. Attrition rates from adolescent mHealth interventions treating mental health problems, such as anxiety and depression, are an issue with high degree of variation and attrition rates. Attrition in mHealth interventions targeting adolescent populations is frequently presented in two-point fashion, from initiation of intervention to end of treatment, lacking more time-specific information on usage and times of attrition. Self-efficacy could provide avenue to lower attrition rates though a better understanding of the relationship between mental health factors and time specific attrition rates is needed. OBJECTIVE The study aimed at obtaining time-specific attrition rates among adolescents in an mHealth intervention and describe the intervention’s usage and feasibility in relation to adolescent self-efficacy levels, emotional- and physical health. METHODS A single-centre, randomized controlled elementary school pilot trial was undertaken with 41 adolescents. Outcome measures were assessed at baseline and after 6-weeks while in-app activity and attrition rates were continually assessed throughout intervention period. Primary outcome was attrition based on time and type of in-app health behaviour usage and feasibility of mHealth app. Secondary outcome measures were self-efficacy levels, depressive- and anxiety symptoms as well as BMI-SDS and sleep. Analyses of group mean variances with adjusted alpha levels through Bonferroni corrections were used to assess main outcome effects. RESULTS Attrition from initiation of intervention to 6-weeks follow-up was 35%. Attrition started in 3rd week of intervention and was related to daily time of app usage (Rt = 0.43, p <.05.). Average weekly in-app health exercises decreased significantly from 1st week of intervention (M = 55.25, SD = 10.96) to the next (M = 13.63, SD = 2.94). However, usage increased by 22% between week 2 and last week of intervention (M = 16.69, SD = 8.37). Usability measures revealed satisfactory scores or 78.09 (SD = 9.82) without gender differences (P=.85). Self-reported daily physical activity increased by 19.61% in intervention group while dropping by 26.21% among controls. Self-efficacy levels were increased by 8.23% in invention arm compared to 3.03% decrease in control group. CONCLUSIONS This pilot study demonstrated the feasibility and usability of a mHealth intervention among adolescent participants. Indications were towards beneficial effects on physical and mental health that warrant further research. Focus on time-specific attrition measures alongside daily times of usage and ways to increase participants’ self-efficacy levels appear to be a promising avenue for research on mHealth interventions for adolescent populations with the aim to ultimately lower attrition rates. CLINICALTRIAL