Theory use in process evaluations alongside randomised trials targeting healthcare professional behaviour change: systematic review

Abstract

Background: Process evaluations (PEs) can help explain the effects of interventions aiming to change healthcare professional (HCP) behaviour by examining the mechanisms through which they operate. Though using theory has been encouraged to better understand these mechanisms, the extent to which theory has been used in this context is unknown. This review aimed to investigate if and how theory has been used in PEs conducted alongside randomized controlled trials (RCTs) of HCP behaviour change interventions. Methods: Eligible studies were (1) primary research/protocol articles (2) describing PEs of (3) interventions evaluated in RCTs (4) targeting HCPs’ behaviour. Searches were conducted on 15/6/2016 in MEDLINE, Embase, PsycINFO, CENTRAL, CINAHL. Study quality was appraised using the Mixed Methods Appraisal Tool. Data extraction included which theories were cited and the extent to which theories were used (i.e. informed by, applied, tested, or built/created) in PEs. Findings: One-hundred-and-six studies were included. Thirty (28%) cited theory. Normalisation process theory (n=12), theory of planned behaviour (n=9), diffusion of innovations (n=6) were most frequently cited. Extent of theory use varied: seven (6.5%) were informed by theory, 16 (15%) applied theory, seven (6.5%) tested theory, and none built/created theory. Discussion: Theory use has been infrequent and insubstantial in this context. Studies that tested theory demonstrate the potential to lend predictive, rather than just explanatory, value to PEs. Thus, there is scope for greater theory use in PEs to develop a cumulative science of behaviour change as they provide an applied setting in which to test and build theories. PROSPERO:CRD42016042789.