Comparing inductive and deductive analysis techniques to understand health service implementation problems: a case study of childhood vaccination barriers

Background
Effective implementation requires a comprehensive understanding of individual, organisational and system determinants. This study aimed to compare inductive and deductive analysis techniques to understand a complex implementation issue. We used childhood vaccination as a case study, an issue with wide-ranging barriers contributing to low-vaccine uptake internationally.

Methods
The study is based on the Behaviour Change Wheel framework, which was derived from several levels of theory: the 3 components of the COM-B framework (capability, opportunity and motivation) can be mapped to the 14 domains of the Theoretical Domains Framework (TDF), which is based on 84 underlying constructs. We first conducted a review of systematic reviews of parent-level barriers to childhood vaccination. Subsequently we (1) inductively coded these barriers into a data-driven framework, using thematic analysis, and (2) deductively mapped the barriers to COM-B and TDF domains and constructs. These processes were undertaken by two authors independently, and discrepancies were resolved through discussion. Inductive and deductive results were compared.

Results
The inductive process coded 583 descriptions of barriers identified from the literature into a framework of 74 barriers in 7 categories. The initial definitions used to map the barriers to deductive domains/constructs led to 89% agreement at the domain level. Resolving discrepancies required further definitions at the construct level. Of the 14 TDF domains, 10 were clearly identified in the data from the barrier reviews. Some domains were not specific enough to differentiate between types of barriers (e.g. Environmental Context and Resources), while other domains were not represented in the review data (e.g. Behavioural Regulation).

Conclusions
Using both inductive and deductive analysis techniques can help achieve a more comprehensive understanding of barriers to health service implementation. The inductive categories represented the review data in a clearer way than the theoretical domains, with better differentiation; but the missing deductive domains were useful as a way to identify additional issues to investigate further. Both analysis techniques resulted in a comprehensive list of barriers to vaccination that would not have been achieved using either approach alone. We recommend a hybrid approach combining TDF with broader frameworks, for future researchers conducting evidence syntheses.