Do digital health behaviour change interventions lead to an increase in self- efficacy? A systematic review and meta-analysis. Katie Newby1. Grace Teah1.
Dr Katie Newby Centre for Advances in Behavioural Science
Katie Newby1 Grace Teah1 Stefanie Williams1 Richard Cooke2 Katherine Brown1 Bradley Salisbury-Finch1 Kristina Curtis1 Kayleigh Kwah1 Emmie Fulton1 Xinru Li3 Elise Dusseldorp3 Joanne Parsons1 Coventry University, UK Aston University, UK Leiden University, The Netherlands
Do digital health behaviour change interventions lead to an increase in selfefficacy?
A systematic review and meta-analysis
• Health psychology has accumulated an extensive toolbox of BCTs in recent years
Introduction
• Meta-analyses have been used to identify which work best to change determinants of health behaviour • One important determinant of health behaviour is selfefficacy • But can digital interventions change self-efficacy? And if so, which BCTs work best?
Aims:
Study aims: 1. To examine the overall effect of digital behaviour change interventions on selfefficacy 2. To examine whether the overall effect of digital behaviour change interventions on selfefficacy varies as a function of the behaviour being addressed 3. To identify which BCTs are most effective in increasing self-efficacy
Method
1
2
3
4
Inclusion/exclusion criteria
Search and screening
Data extraction and coding
Statistical methods
Effect size info (pre and 1st measure post)
Means/SD, pre-post test
Coded: design and sample characteristics, BCTs
Moderation analysis
Automated digital intervention that aimed to have + impact on SE for one of five health behaviours RCT or quasi
CINAHL, Medline, PsycINFO, Scopus, Web of Science Forwards/backwards Requests Dual screening
CMA software Hedges’ g
Results
Results Characteristic Health Behaviour
Sexual Behaviour Smoking Healthy Eating Physical Activity Alcohol Use
Number of studies 2 4 3 7 0
• 6839 records after duplicates removed
Study design
• 16 studies reporting on 17 interventions
RCT Quasi-experimental
15 1
Study Country
High income Middle income Low income
15 1 0
Modes of delivery
IVR Text Email Website DVD
2 4 5 11 2
• 20 separate comparisons
What is the overall effect of digital behaviour change interventions on self-efficacy? Study interventions were successful at increasing selfefficacy g= 0.165, CI 95% 0.034 to 0.296, n= 4788 k= 20, p= 0.014, I² = 76.280
Does the overall effect of digital behaviour change interventions on self-efficacy vary as a function of the behaviour being addressed?
Smoking and sexual behaviour Interventions targeting smoking (g= 0.435, CI 95% 0.250 to 0.619, k= 4, p=