Apps have been identified as a good way to reach out to a broad population to influence health behavioural changes such as diet and exercise. Not only are apps advantageous in terms of reach, but they are also more cost-effective than conventional interventions.
While a number of studies have evaluated the design, features, functionality, feasibility of eHealth and mHealth, there has been little specifically on apps, whether usage of these apps results in the intended behavioral changes (ie, efficacy) and whether the effect is different in children versus adult users. The systematic review by Schoeppe et al aimed to address the current gaps in the literature, with two reviewers independently reviewing 30 published articles describing 27 studies which met the inclusion criteria.
The results of this systematic review showed that apps from 19 of the 27 studies significantly improved behavioural and related health outcomes. Convenience and “on the go” accessibility of smartphone apps may partially be responsible for the lower attrition rate, and therefore have higher efficacy compared with other intervention delivery modes such as counselling. Interestingly, apps with multi-component interventions (vs apps with standalone interventions) and an intervention duration of over 8 weeks, seemed to be more likely to produce significant improvements in health outcomes.
One of the limitations of current literature is the generalizability of the findings outside of the adult population. Only 4 of the 27 studies looked at children or adolescents. This doesn’t really make sense as young people use phones and apps more often and would likely be attracted to and adhere to interventions on apps better than adults.
Finally, based on the included studies, the authors of this paper suggested a number of recommendations for future studies on apps to improve the quality and value of studies.
- Test the efficacy of app features and behavioural change techniques
- Compare the efficacy of stand-alone vs multi-component apps vs other delivery modes
- Use large sample sizes based on power calculations
- Tailor app interventions to specific population groups (eg, women, young people) in whom usage and adoption of app technology is high (I would compare these groups with the lower usage and adoption groups, or stratify with these groups during analysis if the sample size is large enough)
- Use both statistical and self-reported measures to understand why users use or discontinue use of the app
- Explore how to optimize user engagement and retention
- Identify factors that increase user engagement and retention
- Investigate the relationship between user engagement and intervention efficacy, while taking sociodemographic and psychosocial factors into consideration
These recommendations are not exactly new, and the demand for data on the efficacy of apps have been voiced before in both scientific literature and general mobile health news. If we really want to showcase the efficacy and benefits of mobile interventions with apps, compared with traditional modes of intervention delivery, we need to bring studies and reporting up a notch to convince users, fellow medical practitioners, and policy makers.