Using logic models to capture complexity in systematic reviews

Res Synth Methods. 2011 Mar;2(1):33-42. doi: 10.1002/jrsm.32. Epub 2011 Jun 10.

Abstract

Logic models have long been used to understand complex programs to improve social and health outcomes. They illustrate how a program is designed to achieve its intended outcomes. They also can be used to describe connections between determinants of outcomes, for example, low high-school graduation rates or spiraling obesity rates, thus aiding the development of interventions that target causal factors. However, these models have not often been used in systematic reviews. This paper argues that logic models can be valuable in the systematic review process. First, they can aid in the conceptualization of the review focus and illustrate hypothesized causal links, identify effect mediators or moderators, specify intermediate outcomes and potential harms, and justify a priori subgroup analyses when differential effects are anticipated. Second, logic models can be used to direct the review process more specifically. They can help justify narrowing the scope of a review, identify the most relevant inclusion criteria, guide the literature search, and clarify interpretation of results when drawing policy-relevant conclusions about review findings. We present examples that explain how logic models have been used and how they can be applied at different stages in a systematic review. Copyright © 2011 John Wiley & Sons, Ltd.

Keywords: logic models; population health; program evaluation; social welfare; systematic reviews.