Adherence in mobile and web-based technologies for people with psychosis

Abstract

Background: In order to advance and build on the recent boom in mobile and web-based interventions for individuals with psychosis, a better understanding of current levels of adherence and predictors of adherence to mobile and internet interventions is required. 

Method: This paper systematically reviews rates of adherence, dropout and approaches to analyzing predictors of adherence to newly developed mobile and web-based interventions for people with psychosis. A systematic review of randomized controlled trials, feasibility trials and observational trials is presented. We also examine three theoretically proposed predictors of adherence; level of social support present in the trial, level of service user involvement and type of study. 

Results: All the included studies (n=17) reported a measure of adherence and a rate of dropout. The studies varied in terms of their further analysis of adherence; five studies conducted statistical analyses to determine predictors of dropout, five studies conducted analyses on specific predictors of adherence to the intervention, four administered post-trial feedback questionnaires to assess continued use of the intervention, and two studies evaluated different types of interventions with the aim affecting adherence. Overall the percentage of participants adhering to interventions ranged from 60% to 100% with a mean of 79.5%. There was preliminary support for the three theoretically proposed predictors of adherence; adherence was slightly higher in RCT studies (compared to observational studies), in studies with higher levels of social support and in studies with higher levels of servicer user involvement. 

Conclusion: Adherence to mobile and web-based interventions is robust regardless of service-user (e.g. symptoms severity) and intervention (e.g. type of technological interface) specific factors. Future studies should consider reporting a universal measure of adherence such as percent of adherence and should aim to conduct complex analyses on predictors that may impact on adherence including social presence, service user involvement and the type of study.

Date of Award	1 Oct 2016
Original language	English
Awarding Institution	King's College London
Supervisor	Til Wykes (Supervisor) & Clare Reeder (Supervisor)