Systematic searching for theory to inform systematic reviews: is it feasible? Is it desirable?

BackgroundIn recognising the potential value of theory in understanding how interventions work comes a challenge - how to make identification of theory less haphazard? ObjectivesTo explore the feasibility of systematic identification of theory. MethodWe searched PubMed for published reviews (1998-2012) that had explicitly sought to identify theory. Systematic searching may be characterised by a structured question, methodological filters and anitemised search procedure. We constructed a template (BeHEMoTh - Behaviour of interest; Health context; Exclusions; Models or Theories) for use when systematically identifying theory. The authors tested the template within two systematic reviews. ResultsOf 34 systematic reviews, only 12 reviews (35%) reported a method for identifying theory. Nineteen did not specify how they identified studies containing theory. Data were unavailable for three reviews. Candidate terms include concept(s)/conceptual, framework(s), model(s), and theory/theories/theoretical. Information professionals must overcome inadequate reporting and the use of theory out of context. The review team faces an additional concern in lack of theory fidelity'. ConclusionsBased on experience with two systematic reviews, the BeHEMoTh template and procedure offers a feasible and useful approach for identification of theory. Applications include realist synthesis, framework synthesis or review of complex interventions. The procedure requires rigorous evaluation.