Identifying critical intervention, contextual and implementation features in systematic reviews: Intervention component analysis ten years on

Abstract

Background

It is now widely recognised that in addition to providing robust evidence about intervention effectiveness, systematic reviews need to provide decision-makers with information about critical intervention, contextual and implementation features that support successful outcomes. Our 2015 paper introduced Intervention Component Analysis (ICA), a method comprised of three key approaches for examining trials within an effectiveness synthesis: a) an inductive approach to coding trial features; b) extraction of trialists perspectives about the strengths and limitations of features and experiences of implementation; and c) an assessment of identified features in relation to outcomes to assess which appear to be important. In this paper we reflect on how ICA has since been employed by ourselves and others to demonstrate the variety of ways it has been applied and to support further developments of the method.

Methods

In March 2025 we searched Google Scholar for papers citing the 2015 ICA paper. We extracted information on: the year of publication; whether ICA was employed as an analysis method or simply cited; which aspects of ICA were employed; whether ICA was employed in combination with other synthesis methods; and any innovations to or refinements of the method. We also conducted several in-depth case studies to explore variation in application and illustrate the benefits of ICA.

Findings

We identified 95 papers citing the 2015 paper of which 44 reported using or drawing on ICA in their analysis and a further 4 protocols reported a plan to use ICA; the remainder (n = 47) cited but did not use ICA. Of the 48 that used or planned to use ICA, most (n = 38) used or planned to use it in combination with another method such as a meta-analysis or qualitative comparative analysis whilst 10 used ICA as the sole method of analysis. We identified several innovative applications of ICA including the use of existing frameworks or logic models alongside the inductive coding method.

Discussion

Use of ICA allows systematic reviewers to better understand ‘how’ interventions work. ICA has been found by ourselves and others to be both useful and flexible – able to be tailored to both large and small reviews, suitable for use as a stand-alone tool or alongside existing theories or taxonomies, useful as a tool for understanding intervention variation only or as a tool for explaining variation in outcomes. Potential future avenues for development include the use of subgroup analysis and meta-regression to test the theories generated by ICA.