Background: Disentangling the interplay between experience-based intuition and theory-informed implementation is crucial for identifying the direct contribution theory can make for generating behaviour changes needed for successful evidence translation. In the context of 'clinicogenomics', a complex and rapidly evolving field demanding swift practice change, we aimed to (a) describe a combined clinician intuition- and theory-driven method for identifying determinants of and strategies for implementing clinicogenomics, and (b) articulate a structured approach to standardise hypothesised behavioural pathways and make potential underlying theory explicit.
Methods: Interview data from 16 non-genetic medical specialists using genomics in practice identified three target behaviour areas across the testing process: (1) identifying patients, (2) test ordering and reporting, (3) communicating results. The Theoretical Domains Framework (TDF) was used to group barriers and facilitators to performing these actions. Barriers were grouped by distinct TDF domains, with 'overarching' TDF themes identified for overlapping barriers. Clinician intuitively-derived implementation strategies were matched with corresponding barriers, and retrospectively coded against behaviour change techniques (BCTs). Where no intuitive strategies were provided, theory-driven strategies were generated. An algorithm was developed and applied to articulate how implementation strategies address barriers to influence behaviour change.
Results: Across all target behaviour areas, 32 identified barriers were coded across seven distinct TDF domains and eight overarching TDF themes. Within the 29 intuitive strategies, 21 BCTs were represented and used on 49 occasions to address 23 barriers. On 10 (20%) of these occasions, existing empirical links were found between BCTs and corresponding distinct TDF-coded barriers. Twenty additional theory-driven implementation strategies (using 19 BCTs on 31 occasions) were developed to address nine remaining barriers.
Conclusion: Clinicians naturally generate their own solutions when implementing clinical interventions, and in this clinicogenomics example these intuitive strategies aligned with theoretical recommendations 20% of the time. We have matched intuitive strategies with theory-driven BCTs to make potential underlying theory explicit through proposed structured hypothesised causal pathways. Transparency and efficiency are enhanced, providing a novel method to identify determinants of implementation. Operationalising this approach to support the design of implementation strategies may optimise practice change in response to rapidly evolving scientific advances requiring swift translation into healthcare.
Keywords: Algorithm; Clinical genomics; Clinical practice change; Implementation science; Mechanism of action.
© 2023. The Author(s).