Behavior Change Approaches in Digital Technology–Based Physical Rehabilitation Interventions Following Stroke: Scoping Review

Abstract

Background: Digital health technologies (DHTs) are increasingly used in physical stroke rehabilitation to support individuals to successfully engage with the frequent, intensive, and lengthy activities required to optimise recovery. Despite this, little is known about behaviour change within these interventions.
Objective: This scoping review aimed to identify if, and how behaviour change approaches (i.e., theories, models and frameworks, and techniques to influence behaviour) are incorporated within physical stroke rehabilitation interventions that include a DHT.
Methods: Databases (Embase, Medline, PyscINFO, CINAHL, Cochrane Library and AMED) were searched using keywords relating to behaviour change, DHT, physical rehabilitation and stroke. Results were independently screened by 2 reviewers. Sources were included if they reported a completed primary research study in which a behaviour change approach could be identified within a physical stroke rehabilitation intervention that included a DHT. Data including the study design, DHT utilised, and behaviour change approaches were charted. Specific behaviour change techniques were coded to the behaviour change technique taxonomy version 1 (BCTTv1).
Results: From a total of 1973 identified sources, 103 studies (5%) were included for data charting. The most common reason for exclusion at full text screening, was the absence of an explicit approach to behaviour change (165/245, 67%). Almost half of the included studies (45/103, 44%) were described as pilot or feasibility studies. Virtual reality (VR) was the most frequently identified DHT type (58/103, 56%) and almost two-thirds of studies focused on upper limb rehabilitation (65/103, 63%). Only a limited number of studies (18/103, 17%) included a theory, model, or framework for behaviour change. The most frequently used BCTTv1 clusters were feedback and monitoring (88/103, 85%), reward and threat (56, 54%), goals and planning (33, 32%), and shaping knowledge (33, 32%). Relationships between feedback and monitoring, and reward and threat were identified using a relationship map with prominent use of both these clusters in interventions which included VR.
Conclusions: Despite an assumption that DHTs can promote engagement in rehabilitation, this scoping review demonstrates that very few studies of physical stroke rehabilitation which include a DHT, overtly utilised any form of behaviour change approach. From those studies that did consider behaviour change, most did not report robust underpinning theory. Future development and research need to explicitly articulate how including DHTs within an intervention may support the behaviour change required for optimal engagement in physical rehabilitation following stroke, as well as establish their effectiveness. This understanding is likely to support the realisation of the transformative potential of DHTs in stroke rehabilitation.